MAM 37: 00-prelims – page i – Press

THE SOCIETY FOR

MEDIEVAL ARCHAEOLOGY
MONOGRAPH 37

Series Editors
Christopher Gerrard
and
Gabor Thomas

MARITIME SOCIETIES OF THE

VIKING AND MEDIEVAL WORLD
MAM 37: 00-prelims – page ii – Press

THE SOCIETY FOR MEDIEVAL ARCHAEOLOGY

MONOGRAPHS
1 G Bersu and D M Wilson (1966) Three Viking Graves in the Isle of Man
2 F W B Charles (1967) Medieval Cruck-building and its Derivatives
3 P A Rahtz (1969) Excavations at King John’s Hunting Lodge, Writtle, Essex, 1955–57
4 A L Meaney and S C Hawkes (1970) Two Anglo-Saxon Cemeteries at Winnall, Winchester, Hampshire
5 H E J Le Patourel (1973) The Moated Sites of Yorkshire
6 G T M Beresford (1975) The Medieval Clay-land Village: Excavations at Goltho and Barton Blount
7 H Clarke and A Carter (1977) Excavations in King’s Lynn, 1963–1970
8 J G Hurst (general ed) Wharram. A Study of Settlement on the Yorkshire Wolds, vol I: D D Andrews and
G Milne (eds) (1979) Domestic Settlement I: Areas 10 and 6
9 C M Mahany, A Burchard and G Simpson (1982) Excavations at Stamford, Lincolnshire, 1963–69
10 P Mayes and K Scott (1984) Pottery Kilns at Chilvers Coton, Nuneaton
11 J G Hurst and P A Rahtz (general eds) Wharram. A Study of Settlement on the Yorkshire Wolds, vol III: R D Bell,
M W Beresford and others (1987) The Church of St Martin
12 D Austin (1989) The Deserted Medieval Village of Thrislington, Co Durham: Excavations 1973–1974
13 V L Yanin, E N Nosov, A S Khoroshev, A N Sorokin, E A Rybina, V L Povetkin and P G Gaidukov (1992) The
Archaeology of Novgorod, Russia
14 K Parfitt and B Brugmann (1997) The Anglo-Saxon Cemetery on Mill Hill, Deal, Kent
15 D Gaimster and P Stamper (eds) (1997) The Age of Transition: the Archaeology of English Culture 1400–1600
16 D A Hinton (2000) A Smith in Lindsey: the Anglo-Saxon Grave at Tattershall Thorpe, Lincolnshire
17 S Lucy and A Reynolds (eds) (2002) Burial in Early Medieval England and Wales
18 S T Driscoll (2002) Excavations at Glasgow Cathedral 1988–1997
19 P Mayes (2003) Excavations at a Templar Preceptory. South Witham, Lincolnshire 1965–67
20 J Hines, A Lane and M Redknap (eds) (2004) Land, Sea and Home
21 W D Klemperer and N Boothroyd (2004) Excavations at Hulton Abbey, Staffordshire 1987–1994
22 K Giles and C Dyer (eds) (2005) Town and Country in the Middle Ages: Contrasts, Contacts and Interconnec-
tions, 1100–1500
23 S M Foster and M Cross (eds) (2005) Able Minds and Practised Hands: Scotland’s Early Medieval Sculpture in
the 21st Century
24 A Saunders (2006) Excavations at Launceston Castle, Cornwall
25 C Gerrard with M Aston (2007) The Shapwick Project, Somerset. A Rural Landscape Explored
26 C Caple (2007) Excavations at Dryslwyn Castle 1980–95
27 C Scull (2009) Early Medieval (Late 5th–Early 8th Centuries ad) Cemeteries at Boss Hall and Buttermarket,
Ipswich, Suffolk
28 G Beresford (2009) Caldecote. The Development and Desertion of a Hertfordshire Village
29 N Edwards (2009) The Archaeology of the Early Medieval Celtic Churches
30 R Gilchrist and A Reynolds (eds) (2009) Reflections: 50 Years of Medieval Archaeology, 1957–2007
31 I H Goodall (2011) Ironwork in Medieval Britain: an Archaeological Study
32 R Barrowman (2011) The Chapel and Burial Ground on St Ninian’s Isle, Shetland: Excavations Past and Present
33 J Hines and A Bayliss (eds) (2013) Anglo-Saxon Graves and Grave Goods of the 6th and 7th Centuries ad:
A Chronological Framework
34 S Rátkai (2013) Wigmore Castle, North Herefordshire: Excavations 1996 and 1998
35 N Christie and O Creighton with M Edgeworth and H Hamerow (2013) Transforming Townscapes. From Burh
to Borough: The Archaeology of Wallingford, ad 800–1400
36 Rosemary Cramp (2014) The Hirsel Excavations

The Society for Medieval Archaeology Monographs are available from

Oxbow Books, 10 Hythe Bridge Street, Oxford ox1 2ew, UK
or in North America from
Casemate Academic, 20 Main Street, Oakville CT 06779, USA
For further information, including prices, or to order online, visit
www.oxbowbooks.com
MAM 37: 00-prelims – page iii – Press

MARITIME SOCIETIES OF THE

VIKING AND MEDIEVAL WORLD

Edited by
James H Barrett
and
Sarah Jane Gibbon
 First published 2015 by Maney Publishing

Published 2016 by Routledge

2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
711 Third Avenue, New York, NY 10017 USA

Routledge is an imprint of the Taylor & Francis Group, an informa business

All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system or transmitted in any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise without the written permission of the
copyright holder.
# 2015 The Society for Medieval Archaeology and Authors

For users in the UK, the Copyright Licensing Agency, [email protected], is mandated to
grant permission to make copies. For Australia, please see www.copyright.com.au or
www.copyright.org.au for more information. For North America and the rest of the
world permission is granted by the copyright holder for libraries and others registered
with the Copyright Clearance Center (CCC), www.copyright.com

Statements in the volume reflect the views of the authors, and not necessarily those of the
Society, editors or publisher.

isbn 978-1-909662-79-7
issn 0583–9106

front cover
Maggi Hambling’s Wall of water XIII, war, 2013, oil on canvas, 198.16226.1 cm,
first exhibited at Maggi Hambling: Walls of Water at the National Gallery, London
(2014–2015) and then at Maggi Hambling: War Requiem & Aftermath at Somerset
House, London (2015)

ISBN: 978-1-315-63075-5(eISBN)
MAM 37: 00-prelims – page v – Press

CONTENTS
page
Preface and Acknowledgements . . . . . . ix
List of Contributors . . . . . . . . xi
chapter 1
Maritime Societies and the Transformation of the Viking Age and
Medieval World
By James H Barrett . . . . . . . . 1
chapter 2
Sails and the Cognitive Roles of Viking Age Ships
By Christer Westerdahl . . . . . . . 14
chapter 3
Trade and Trust in the Baltic Sea Area During the Viking Age
By Ingrid Gustin . . . . . . . . . 25
chapter 4
Bound for the Eastern Baltic: Trade and Centres ad 800–1200
By Marika Mä gi . . . . . . . . . 41
chapter 5
Between East and West: Economy and Society on the Island of Gotland
By Dan Carlsson . . . . . . . . . 62
chapter 6
Viking Age Bornholm: An Island on the Crossways
By Magdalena Naum . . . . . . . . 69
chapter 7
Trading Hubs or Political Centres of Power? Maritime Focal Sites in Early
Sweden
By Stefan Brink . . . . . . . . . 88
chapter 8
Accessibility and Vulnerability: Maritime Defence and Political Allegiance
on the Vikbolandet Peninsula, Östergötland, Sweden
By Martin Rundkvist . . . . . . . . 99
chapter 9
Dorestad as a Fluviatile Society
By Annemarieke Willemsen . . . . . . . 108
chapter 10
Maritime Environment and Social Identities in Medieval Coastal Flanders:
The Management of Water and Environment and its Consequences for the
Local Community and the Landscape
By Dries Tys . . . . . . . . . . 122
MAM 37: 00-prelims – page vi – Press

chapter 11
The Maritime Cultural Landscape of Early Medieval Northumbria: Small
Landing Places and the Emergence of Coastal Urbanism
By Pieterjan Deckers . . . . . . . . 138
chapter 12
Post-Substantivist Production and Trade: Specialized Sites for Trade and
Craft Production in Scandinavia ad c700–1000
By Dagfinn Skre . . . . . . . . . 156
chapter 13
Late Iron Age Boat Rituals and Ritual Boats in Norway
By SÆbj Ø rg Walaker Nordeide . . . . . . 171
chapter 14
Bergen ad 1020/30–1170: Between Plans and Reality
By Gitte Hansen . . . . . . . . . 182
chapter 15
Steatite Vessels and the Viking Diaspora: Migrants, Travellers and Cultural
Change in Early Medieval Britain and Ireland
By S Ø ren Michael SindbÆk . . . . . . . 198
chapter 16
Status and Identity in Norse Settlements: A Case Study from Orkney
By David Griffiths . . . . . . . . 219
chapter 17
The Viking Occupation of the Hebrides: Evidence from the Excavations at
Bornais, South Uist
By Niall M Sharples, Claire Ingrem, Peter Marshall,
Jacqui Mulville, Adrienne Powell and Kelly Reed . . 237
chapter 18
Disentangling Trade: Combs in the North and Irish Seas in the Long
Viking Age
By Steven P Ashby . . . . . . . . . 259
chapter 19
Dealing with Deer: Norse Responses to Scottish Isles Cervids
By Jacqui Mulville . . . . . . . . 279
chapter 20
‘Warrior Graves’? The Weapon Burial Rite in Viking Age Britain and
Ireland
By Stephen H Harrison . . . . . . . 299
chapter 21
The Threatening Wave: Norse Poetry and the Scottish Isles
By Judith Jesch . . . . . . . . . 320
MAM 37: 00-prelims – page vii – Press

chapter 22
Sea Kings, Maritime Kingdoms and the Tides of Change: Man and the
Isles and Medieval European Change, ad c1100–1265
By R Andrew McDonald . . . . . . . 333
chapter 23
The Sea Power of the Western Isles of Scotland in the Late Medieval
Period
By David H Caldwell . . . . . . . . 350
chapter 24
Coastal Communities and Diaspora Identities in Viking Age Ireland
By Clare Downham . . . . . . . . 369
Index . . . . . . . . . . . 384
MAM 37: 00-prelims – page ix – Press

PREFACE AND ACKNOWLEDGEMENTS

The idea behind this volume emerged from a conference of the same name, jointly
organized by the Orkney Heritage Society and the McDonald Institute for Archaeo-
logical Research, held in Kirkwall in May/June of 2008. The meeting initially aimed
to set the semi-independent island polities of Viking Age and medieval Scotland (such
as the Earldom of Orkney and the Lordship of the Isles) in a comparative context
that went beyond the rural North Atlantic region to which they are often naturally
ascribed. It was an opportunity to consider the intersections and separations of rural
and urban worlds, of small islands and the coastal or riparian settlements of ‘main-
lands’, around the Irish, North and Baltic Seas. We aimed to explore a pivotal time in
the creation of the social, economic and political landscapes of Europe — when
small-scale maritime polities had a disproportionate impact on the course of world
history. The chronology of the project was set to include the Viking Age (ad c790–
1050) and early Middle Ages (defined as ad c1050–1200), with a few later exceptions
that were clearly relevant to earlier developments. The result was an illuminating
experience, drawing attention to the complex relationship between long-range inter-
connections and distinct regional identities characteristic of maritime communities.
The participants expressed enthusiasm for a resulting publication, and written con-
tributions began to come in. The original line-up was then supplemented with invited
papers that filled important missing niches (regarding, for example, the key Viking
Age port of Dorestad and the island societies of the Baltic). To accommodate this
broadened coverage within the publisher’s limits we have excluded our own contri-
butions regarding Atlantic Scotland, but they are available elsewhere and were
equally informed by the event. The book’s geographical range remains far from com-
prehensive. Nevertheless, we hope its audiences will find much of value as they seek
to situate their own worlds of study in the context of numerous examples which were
diverse and changing, yet exhibited significant resonances and often direct inter-
connections.
Many people and organizations have contributed to the volume’s completion.
Acknowledgements for each chapter are provided in endnotes where relevant. Credits
for the images used are on page iv and in the figure captions. Linda Fisher, Suzanne
Needs-Howarth and Lembi Lõugas assisted with copy-editing. Funding for the initial
conference was provided by the British Academy, Historic Scotland, the McDonald
Institute for Archaeological Research, Orkney Islands Council and the Society of
Antiquaries of Scotland. It was organized by the editors, in cooperation with Anne
Brundle (Orkney Museum), Julie Gibson (Orkney College, University of the High-
lands and Islands) and Kate Towsey. The Orkney Heritage Society, Orkney College
and the then Friends of Orkney Archaeological Trust (now Orkney Archaeology
Society) provided much logistical support during the event. Christopher Gerrard,
the Society for Medieval Archaeology and Maney kindly welcomed its publication
in the Society’s monograph series — and maintained the right balance of pressure and
patience during its completion. Steady early progress was slowed by maternity,
MAM 37: 00-prelims – page x – Press

x maritime societies of the viking and medieval world

teaching and administrative responsibilities, before the project received additional

invaluable support from the McDonald Institute for Archaeological Research in
2013. This made it possible for Dora Kemp (devoted sailor and production editor) to
begin to format the edited chapters into the publisher’s house style. When she tragi-
cally fell ill and passed away in 2014, it was possible for us to maintain the momen-
tum to completion (thanks to research leave supported by the Leverhulme Trust).
This volume is thus in part a tribute to Dora, and to Anne Brundle who left us in
2011, while also being a contribution to an understanding of (often small-scale) com-
munities that were brought into being by their relationships with water — and set
waves in motion that altered distant shores.
James Barrett
McDonald Institute for Archaeological Research, University of Cambridge
Sarah Jane Gibbon
Archaeology Institute, University of the Highlands and Islands
MAM 37: 00-prelims – page xi – Press

LIST OF CONTRIBUTORS

Steven P Ashby
Department of Archaeology, University of York, York, England, UK
James H Barrett
McDonald Institute for Archaeological Research, Department of Archaeology and
Anthropology, University of Cambridge, England, UK
Stefan Brink
Centre for Scandinavian Studies, University of Aberdeen, Scotland, UK
David H Caldwell
Society of Antiquaries of Scotland, Edinburgh, Scotland, UK; previously National
Museums Scotland, Edinburgh, Scotland, UK
Dan Carlsson
Arendus Ltd, Visby, Gotland, Sweden
Pieterjan Deckers
Medieval Archaeology, Department of Archaeology and Art Studies, Vrije
Universiteit Brussel, Belgium
Clare Downham
Institute of Irish Studies, University of Liverpool, England, UK
Sarah Jane Gibbon
Archaeology Institute, University of the Highlands and Islands, Kirkwall, Orkney,
Scotland, UK
David Griffiths
Oxford University Department for Continuing Education, University of Oxford,
England, UK
Ingrid Gustin
Department of Archaeology and Ancient History, Lund University, Sweden
Gitte Hansen
Department of Cultural History, University Museum, University of Bergen,
Norway
Stephen H Harrison
Archaeology, School of Humanities, University of Glasgow, Scotland, UK
Claire Ingrem
Freelance zooarchaeologist, Lyme Regis, England, UK
Judith Jesch
Centre for the Study of the Viking Age, School of English, University of
Nottingham, England, UK
MAM 37: 00-prelims – page xii – Press

xii maritime societies of the viking and medieval world

R Andrew McDonald
Department of History, Brock University, St Catharines, Canada
Marika Mägi
Centre of Medieval Studies, Tallinn University, Estonia
Peter Marshall
Chronologies, Sheffield, England, UK
Jacqui Mulville
School of History, Archaeology and Religion, Cardiff University, Wales, UK
Magdalena Naum
Faculty of Humanities, Oulu University, Finland
Adrienne Powell
School of History, Archaeology and Religion, Cardiff University, Wales, UK
SÆbj ø rg Walaker Nordeide
Department of Cultural History, University Museum of Bergen, University of
Bergen, Norway
Kelly Reed
Department of Engineering Science, University of Oxford, England, UK
Martin Rundkvist
Department of Historical, Philosophical and Religious Studies, Umeå University,
Sweden
Niall M Sharples
School of History, Archaeology and Religion, Cardiff University, Wales, UK
S ø ren Michael SindbÆk
Section for Archaeology, Department of Culture and Society, Aarhus University,
Denmark
Dagfinn Skre
Museum of Cultural History, University of Oslo, Norway
Dries Tys
Medieval Archaeology, Department of Archaeology and Art Studies, Vrije
Universiteit Brussel, Belgium
Christer Westerdahl
Institute of Historical Studies, Norwegian University of Science and Technology
(NTNU), Trondheim, Norway
Annemarieke Willemsen
Medieval Department, National Museum of Antiquities, Leiden, The Netherlands
MAM 37: 01-barrett – Press

chapter 1
MARITIME SOCIETIES AND THE
TRANSFORMATION OF THE VIKING
AGE AND MEDIEVAL WORLD

By James H Barrett

introduction
This book is about maritime societies, groups at scales ranging from a settlement
to a polity, who drew their identity and livelihood from relations with water. It
focuses on the Baltic, North and Irish Seas, and the waters that linked them, in the
Viking Age (ad c790–1050) and, with a few later exceptions, the early Middle Ages
(ad c1050–1200). Thus it explores long-range interaction in the northern European
world prior to the era of the German Hansa, and provides a complement to past
research on the Norse settlements of the North Atlantic (eg Arneborg et al 2009–10;
Barrett 2003; Dugmore et al 2005; Vésteinsson et al 2011; Sigurdsson 2008; Zori and
Byock 2014). The book crosscuts the traditional chronological boundary between the
Viking Age and the Middle Ages, and provides a comparative context for maritime
communities ranging from, for example, the Isle of Man in the Irish Sea to Gotland
in the Baltic. Concurrently, it touches on the relationship between estate centres,
landing places, towns and the sea in the more terrestrially oriented societies that sur-
rounded northern Europe’s main spheres of maritime interaction. Many studies of
Viking Age and medieval maritime activity focus on warfare (eg Price 2002), trade
(eg Sindbæk 2007) or identity (eg Abrams 2012) — recognizing that they are inter-
dependent at various scales. A key purpose of the present book is to illuminate this
interrelationship, exploring the dialectic between long-range interconnections and
distinctive expressions of local identity.
To achieve the desired comparative perspective, the book includes contributions
that draw on diverse academic traditions, ranging from the excavation-project
approach of much research in Atlantic Scotland (Shetland, Orkney, the Hebrides and
adjacent areas of the Scottish mainland) to the use of archaeological cultures as units
of analysis in eastern Europe. It is predominately an archaeological project, but
draws no arbitrary lines between studies of historical archaeology, history and litera-
ture, given that the methodological differences within each subject can be as great as
those between them.
MAM 37: 01-barrett – Press

2 maritime societies of the viking and medieval world

In introducing the volume, this chapter aims to explore key themes to which
many of the contributors speak. To what degree were maritime societies isolated or
interconnected? Did they emulate or reject the practices and identities of their neigh-
bours? How and why did these choices change through time? To what degree did
increasing urbanism, commercialization and trade impact on local cultures, economies
and landscapes/seascapes? How did fluid maritime societies react to increasingly
centralized power during the consolidation of medieval kingdoms and principal-
ities? These themes are at the heart of our understanding of the transition from the
Viking Age to the Middle Ages in Europe. They also run deeper, emerging from and
contributing to wider debates regarding the dialectical nature of maritime societies
and, more holistically still, the ebb and flow of globalization over the longue durée.

worlds within worlds

The use of ‘world’ in the title of the book is intended to convey three interrelated
scales of analysis. Firstly, some of the maritime polities of the Viking Age and medi-
eval north were microcosms unto themselves, marked out at times in their histories
by distinctive suites of material culture, burial traditions, literature and/or social
relations with external powers. Examples include the Islands of Gotland (Carlsson
Chapter 5; Gustin Chapter 3) and Bornholm (Naum Chapter 6) in the Baltic Sea —
and the Earldom of Orkney (Griffiths Chapter 16; Jesch Chapter 21), the Kingdom of
Man (McDonald Chapter 22), the Kingdom of Dublin (Downham Chapter 24) and
the Lordship of the (Hebridean) Isles (Caldwell Chapter 23) in the west. Secondly,
these societies operated within a spatially attenuated world-system, a geographically
discontinuous political economy that integrated parts of the Irish, North and Baltic
Seas in exchanges of people, objects, wealth and organized violence. An 11th-century
earl of Orkney (Rognvald Brusason) might spend time in exile in Novgorod
(Crawford 2004), Baltic amber might be worked in 10th-century Dublin (Wallace
1987, 215–216) and Islamic silver might pass through Gotland or Bornholm to help
pay an army of conquest at Torksey in 9th-century England (Blackburn 2011; Hadley
and Richards 2013). Within Europe, the plunder-fuelled expansions and reorganiza-
tions of the Carolingian Empire provided a model of conquest, trade and redistri-
bution of wealth that was emulated by this northern system (Hernæs 1997; Reuter
1985; Thomas 2012) — and connected to it, by centres like Dorestad and later
Hedeby (Theuws 2004; Willemsen Chapter 9). Lastly, efflorescences of mobility in
the Viking Age and medieval north were also expressions of wider pan-Eurasian
currents, Lieberman’s (2009, 135) ‘phase of demographic, agrarian, and commercial
vitality’ and ‘supralocal contacts’ that unfolded on a hemispherical scale (cf Sindbæk
and Trakadas 2014). Given the interrelationships between these scales, studies that
begin with the small worlds of polities like Orkney, Man or Gotland are central to
understanding the local dynamics that ultimately created much deeper and more
widespread phenomena (which themselves were only ever experienced in specific
places) (cf Barrett 2012).
The ebb and flow of these different worlds can be studied from the top down,
drawing on overlapping corpora of theory that attempt to clarify facets of
MAM 37: 01-barrett – Press

the transformation of the viking age and medieval world 3

globalization. In the last two decades, aspects of world-systems, post-colonial, social

network, actor-network and diaspora theory have all played a role in research on the
Viking Age and medieval north (eg Abrams 2012; Barrett 2012; Barrett et al 2000;
Glørstad 2014; Mehler 2013; Sindbæk 2007; 2012a; 2012b; Svanberg 2003). Con-
versely, it can be understood from the bottom up, exploring the empirical evidence
from particular case studies while avoiding naı̈ve positivism. Both approaches have
merit. Given that recent phenomena — such as electronic currency markets, instant
communication and migration of wage labour to follow rapid flows of capital —
ultimately underpin much relevant theory (eg Appadurai 1996; Ferguson and
Mansbach 2012; Palla et al 2005), there is a legitimate argument for focused case
studies built around bounded examination of primary sources. Many contributions
to this volume put such an empirical approach to good effect — sometimes rejecting
theoretically inspired conclusions that were once orthodoxies. Conversely, a method
drawing predominantly on comparative models can be equally rewarding. It illu-
minates how the relationship between source material and meaning can be very
complex — particularly in the context of creating and maintaining new identities.
Thus Downham (Chapter 24) explores how Dublin’s Hiberno-Scandinavian com-
munity might have emerged in the context of a trade diaspora, drawing on examples
from eastern Africa and Malaysia. The two approaches meet in constructive ways.
Skre (Chapter 12), for example, develops his argument for profit-based trade in the
Viking Age from his knowledge of the archaeological parallels for Norway’s first
town at Kaupang — set within the context of an informed survey of the formalist-
substantivist debate in economic anthropology. The book is thus a fusion of the
traditional and topical. It reflects the hybridity of medieval archaeology itself,
drawing on both the particularistic methods of history and archaeology’s links to the
social sciences where analogical reasoning is the norm.

mobility, trade and identity

It is no longer necessary to debate whether influential Viking Age Scandinavian
migrations (to east, south and west) happened, as was the case in the second half of
the 20th century (eg Sawyer 1962; Wormald 1982). Nor is it controversial that they
were multicausal, culturally diverse and unfolded in multiple ways at very different
scales. In terms of time, the 11th-century empire building of Cnut the Great (Bolton
2009) is hardly comparable with the two fleets of 60 ships said to have been operating
in Ireland in 837 (MacAirt and MacNiocaill 1983, 295), nor with the three ships that
landed at Portland in southern England at the very beginning of the Viking Age
around the end of the 8th century (Campbell 1962, 26–27; Whitelock et al 1961, 35).
In terms of space, royal Scandinavian exiles at the Carolingian court (Coupland 1998)
did not represent the same phenomenon as conquest of Anglo-Saxon kingdoms
(Brooks 1979) or rural settlement by farmer-fishermen who rubbed shoulders with
local Pictish peasants in Orkney (Barrett 2004; 2012; Griffiths Chapter 16; Mont-
gomery et al 2014). Moreover, all of these examples differed from the cultural
exchanges between Scandinavian, Slavic and eastern Baltic communities that are
evident from finds of distinctive metalwork and/or pottery in Gotland (Carlsson
MAM 37: 01-barrett – Press

4 maritime societies of the viking and medieval world

Chapter 5; Gustin Chapters 3) and Bornholm (Naum Chapter 6). Even the latter
islands had different political and social histories, with Gotland having maintained
its distinctive role in inter-regional trade into the 12th century and beyond (Myrberg
2010).
Much that was once controversial is thus now commonplace. Yet the Viking Age
and its medieval echoes do entail a paradox at the heart of ongoing debate. How can a
phenomenon as widespread as the Scandinavian diaspora, with some elements of
astounding homogeneity (in language, art styles and some artefact types, such as oval
brooches), concurrently entail a diversity of extremely local expressions of distinc-
tive, sometimes hybrid, identities (Abrams 2012; Sindbæk 2012a)? At different scales,
this regionalism might be expressed with Gotlandic animal-head brooches (Carlsson
Chapter 5), ‘Dublin-type’ shield bosses (Harrison Chapter 20) or Norwegian soap-
stone cooking pots (Sindbæk Chapter 15). One possibility is that mobility was limited
and the wider Viking Age is at least in part a modern historiographical construct
(Svanberg 2003). The alternative, drawing on comparative study of globalization
processes more holistically, is that local reactions to enhanced interconnectedness are
exactly what one should expect, so much so that they have inspired use of the neo-
logism ‘glocalization’ among social scientists (Ferguson and Mansbach 2012, 138).
Resolution of the paradox lies in the observation that historical efflorescences of
mobility should be concurrently marked by elements of sweeping homogeneity and
highly idiosyncratic local traditions. Moreover, the social tensions created by
external influences can lead to the most connected places having the most distinctive
regional reactions. Thus Dublin’s hybrid Hiberno-Scandinavian mercantile commu-
nity was still evident at the time of the Anglo-Norman invasions of the late 12th
century (Downham Chapter 24) and Gotland’s unique rural society of trade middle-
men/women adopted its own vocabulary of material culture symbols — before,
during and after the Viking Age (Carlsson Chapter 5; Gustin Chapter 3; Myrberg
2010).
This issue brings together two themes that have been central to studies of the
Viking Age: trade and identity. The division of the north into (Christian-influenced)
coin and (pagan-influenced) bullion economies has long been recognized as the result
of an ideological rather than a trade boundary (Kilger 2008; Steuer 1987; Theuws
2004). Towns like Dorestad existed to facilitate large-scale exchange of goods
between these systems, and grew astonishingly wealthy in the process (Willemsen
Chapter 9). Smaller ports like York or Kaupang, and a plethora of estate centres with
their associated landing places, did the same on different scales (Deckers Chapter 11;
Skre Chapter 12). In certain times at certain places even peasant settlements on the
coastal fringe could benefit from the flow of goods between north and south, east and
west (Loveluck 2013; Loveluck and Tys 2006; Tys Chapter 10). Within this intercon-
nected world, different maritime communities made different choices at different
times regarding which model to emulate. Thus silver as bullion was used in Kaupang
while coinage was briefly minted at contemporary Hedeby, despite both towns
probably being Danish royal foundations (Skre 2008; Chapter 12; Wiechmann 2007).
Even late in the Viking Age, after coinage had been adopted in Dublin c995
(Blackburn 2008; Woods 2013) and (briefly) the Isle of Man c1025 (Bornholdt-Collins
MAM 37: 01-barrett – Press

the transformation of the viking age and medieval world 5

1999), the Earldom of Orkney probably continued to employ an anachronistic form

of weight-adjusted silver ring-money as a means of exchange (Graham-Campbell
1995; Critch forthcoming). This cannot have been an accident of remoteness.
Orcadians were deeply involved in the politics of the Irish and North Seas, often as
warriors for hire, in the 11th century — most notably at the Battle of Clontarf in
Ireland (1014) and the Battle of Stamford Bridge in England (1066) (Barrett 2007).
As the Viking Age became the Middle Ages, growing political authority in many
parts of the north led to the increasing restriction of long-range trade to centres
that became medieval towns (eg Loveluck 2013, 302–303; see below). Whether this
was after an economic downturn in the 9th and 10th centuries (attributable to the
disruptions of Viking Age raiding, changing ship types, river silting and/or internal
economic changes) is a matter of geographical perspective (eg Lebecq 2007; Sindbæk
2012b). New mercantile communities like Dublin prospered, while many older centres
were abandoned (eg Dorestad, see Willemsen Chapter 9), relocated (eg Hedeby, see
Radtke 2006) and/or temporarily much reduced (eg London and Ribe, see Feveile
2010; Hodges 2012, 114). Around the end of the 1st millennium, the new medieval
towns, like their less numerous Viking Age antecedents, became major nodes of inter-
connectedness, thus creating both widespread centrifugal homogenizing forces and
centripetal foci for the coalescence of local and regional identities (cf Sindbæk 2012a;
Williams 2013). Nevertheless, the emergence of this new way of living was not always
embraced immediately, particularly in traditionally rural regions. For example,
Bergen’s transition from 11th-century royal aspiration to 12th-century metropolis
was a protracted one (Hansen Chapter 14). Moreover, some material expressions of
identity — such as soapstone cooking pots (Sindbæk Chapter 15) and hair combs of
reindeer antler with copper-alloy rivets (Ashby Chapter 18) — circulated mainly
within cultural provinces, despite frequent traffic between them.

organized violence
There have been diverse and shifting perspectives on the degree to which raiding
and conquest were or were not dominant forms of mobility (vis-à-vis gift exchange,
trade and/or rural settlement, for example) in particular contexts during and after the
Viking Age (eg Barrett 2004). Nevertheless, the observation that organized violence
was a defining feature of the age among all the peoples of Europe is both a cliché and
an evidence-based statement (eg Reuter 1985; Winroth 2014, 41). To understand it
from a 21st-century perspective, with a non-violent world among our greatest aspira-
tions, it is helpful to attempt to separate archaeology as heritage from archaeology as
study of the distant past (cf Carver 1995, 188–189). Although a critical distance
cannot provide objectivity (McGuire 2002, 213–218), it may allow us a glimpse of
how contemporaries viewed their actions.
Organized violence was clearly a successful socio-economic strategy for many in
the Viking Age and after. Young people (often men) from disparate rural settlements
(some newly founded in the east and west) made their fortunes and reputations if they
did not lose their lives (Barrett 2010; Glørstad 2014). Military leaders could climb the
social ladder to kingship (Kershaw 2000). Existing kings could aspire to an empire
MAM 37: 01-barrett – Press

6 maritime societies of the viking and medieval world

(Bolton 2009). Fencing plunder and slaves made townspeople wealthy (Holm 1986),
and the overlords who taxed them even wealthier (Loveluck 2013). Thus the worlds
of raiding, trading and rural settlement were directly interrelated.
The misery of those on the opposite end of this success does not go unrecorded
— and was very real despite exaggeration by ecclesiastical annalists with ideological
axes to grind (Dumville 1997). Nor was it restricted to indigenous victims of Viking
raiding. The reinforcement of local identity engendered by the tensions of external
threat could be equally brutal. It is witnessed by examples as diverse as the St Brice’s
day massacre of 1002, when all Danes in England were to be slain by royal decree
(Pollard et al 2012; Whitelock et al 1961, 86), the 10th-/11th-century mass grave of
executed (probably) Scandinavian raiders at Weymouth Ridgeway in Dorset (Loe et al
2014) and the slaughter (probably by a scandalized Frankish aristocracy) of peasants
from the area between the Seine and Loire who took up arms in self-defence against
the ‘Danes’ in 859 (Nelson 1991, 89). Making a living and building community could
be a dangerous business in the Viking Age.
The military- or plunder-economy of the Viking Age was continued into the
Middle Ages by the hybrid creations of the Scandinavian diaspora. These were prin-
cipalities such as the Kingdom of Dublin (Downham Chapter 24), the Kingdom of
Man (McDonald Chapter 22), the Earldom of Orkney (Griffiths Chapter 16; Jesch
Chapter 21; cf Barrett 2007) and the Lordship of the (western Scottish) Isles (Caldwell
Chapter 23). Their strategies of piracy and mercenary service may also have applied
to some Baltic islands (eg Naum 2012), although they were arguably different to those
employed by the Gotlanders (for whom protecting trade was the secret to socio-
economic success) (Carlsson Chapter 5). Organized violence was a ubiquitous feature
of medieval and post-medieval Europe — where ‘war made the state’ and vice versa
(Tilly 1975, 42; cf Munzinger 2006). Nevertheless, the distinctive maritime societies of
the north were situated in liminal zones between larger kingdoms, providing unique
niches for both large-scale mercenary service and piracy. Irish kings hired the fleet of
Dublin (Holm 1986, 338) and, based on analogy with late medieval sources, the thinly
populated Hebridean islands of western Scotland could field a professionally trained
and equipped mercenary force of up to 6000 men (Caldwell Chapter 23). It is impor-
tant to understand these medieval military societies in their own terms, as normalized
outcomes of a long history of analogous activity and creations of a new realpolitik.
They were polities of their age, fully engaged with contemporary ecclesiastical and
secular European culture (eg McDonald Chapter 22) — rather than anachronistic
survivals of a Viking Age past.

transformations
Whatever its complex causes (cf Barrett 2010; Sindbæk 2011), the Viking Age
was characterized by the broadly contemporary emergence of long-range raiding,
sailing ships, kingship and ‘specialized sites of trade and craft’ in Scandinavia (Skre
Chapter 12; cf Nordeide Chapter 13; Westerdahl Chapter 2). The impact of these
developments on neighbouring polities around the Baltic, North and Irish Seas was
MAM 37: 01-barrett – Press

the transformation of the viking age and medieval world 7

not trivial. It is recognized that the Scandinavian diaspora contributed in complex

ways to state formation and/or urbanism in what are now Ireland, Scotland, England,
Russia and the Ukraine (Bradley 1988; Broun 1994; Downham 2007; Duczko 2004;
Hadley and ten Harkel 2013; Woolf 2007), while also playing roles in the emergence of
Normandy as an independent principality (Abrams 2013; Neveux 2002) and the inter-
necine strife of the later Carolingian Empire (Coupland 1995). Moreover, the
Northern and Western Isles of Scotland and parts of England and Ireland received
significant (albeit complex and unquantifiable) immigration (Abrams and Parsons
2004; Barrett 2004; Bowden et al 2007; Brooks 1979; Downham Chapter 24; Harrison
Chapter 20; Kershaw 2013; Macniven 2013; McLeod 2014; Sharples Chapter 17). An
iconic innovation — the adoption of sailing between the construction of two rowing
vessels at Karmøy in western Norway c770–780 and the building of the Oseberg ship
c820 (Bonde and Stylegar 2009; Nordeide Chapter 13; Westerdahl Chapter 2) —
provides a useful way to comprehend the underlying transformation of Scandinavian
society itself. The motive and capacity to mobilize labour and materials on a large
scale had changed fundamentally (Barrett and Anderson 2010, 311; Bill 2010;
Westerdahl 2008).
The subsequent transition from the Viking Age to the Middle Ages in Scandi-
navia and parts of the Baltic Sea region was for many a shift to life in larger and more
powerful kingdoms, administered by a network of new long-lived urban centres and
ideologically underpinned by Christian institutions and royal coinage (eg Bagge 2010;
Risvaag and Christophersen 2004; Thurston 2001). England shared a similar trajec-
tory a little earlier. The rapidly urbanizing 9th–10th century Anglo-Scandinavian
kingdoms of East Anglia and Northumbria were conquered by the kings of Wessex in
the 10th century, creating a temptingly affluent target for renewed raids and ulti-
mately conquest (from Scandinavia in the decades around the end of the 1st millen-
nium, and from Scandinavia and Normandy in 1066) (Astill 1991; Sawyer 2013).
Although long a Christian region, the 10th-century unification of England was not
without its ideological dimension, a key example being royal sponsorship of the
English Benedictine reform (Barrow 2008). In contrast, the Carolingian Empire frag-
mented in the 9th and 10th centuries, with much control in the west being devolved to
autonomous regional magnates such as the Counts of Flanders (Tys Chapter 10). Yet
power in these new smaller polities was also consolidated by encouraging urban-
ization and sponsoring ecclesiastical institutions (Loveluck 2013) — reducing inter-
regional differences.
This new political landscape altered the lives of many maritime societies of the
North, Baltic and Irish Seas. For some on the fringes of consolidating powers, oppor-
tunities emerged to reify what may previously have been opportunistic raiding. As
discussed above, piracy and sea-borne mercenary activity were turned into economic
strategies and cultural norms with considerable longevity in contexts such as the
Earldom of Orkney, the Kingdom of Man and the Lordship of the Isles. Elsewhere
the outcome was different. The economic flexibility previously open to peasants of
coastal Flanders was slowly compromised (Tys Chapter 10). They were economically
swamped as coastal wetlands were replaced with comital estates providing wool for
MAM 37: 01-barrett – Press

8 maritime societies of the viking and medieval world

textile production, which itself became increasingly urban in its organization. Yet the
distinction between ‘pirates’ and ‘sheep farmers’ can be drawn too sharply. It is likely
that both Orcadians and Flemmings died as mercenaries at the Battle of Stamford
Bridge, during Harald Hardrada of Norway’s failed invasion of England in 1066
(Whitelock et al 1961, 138–145).
In many areas, access to long-range trade via small landing places became a
thing of the past as new permanent coastal and riverine settlements (often towns)
attempted, with lordly support, to monopolize trade within their hinterlands
(Deckers Chapter 11; cf Loveluck 2013). This was largely a phenomenon of the 11th
to 12th centuries across the entire geography of present concern, from Riga (where a
trading place and hillfort of 12th-century origin predates the later 13th-century town)
in the eastern Baltic (Mägi Chapter 4) to Bristol in the Irish Sea region (where coins
were first minted between 1017 and 1023) (Loveluck 2013, 352–353). Before this date,
the north could be divided into zones in which coastal settlement was ubiquitous,
albeit seldom urban (eg Norway and Scotland), and those where it was the exception
(eg England and the eastern Baltic) (cf Deckers Chapter 11; Griffiths Chapter 16;
Mägi Chapter 4; Nordeide Chapter 13; Sharples Chapter 17; Skre Chapter 12). The
diversity of contexts in which this shift occurred belies political explanations such
as the Norman Conquest, despite the empirical observation that many of England’s
coastal towns and ports were initiated by Anglo-Norman magnates (Beresford 1967;
Deckers Chapter 11). Moreover, the new towns of the 11th and 12th centuries could
also be internal creations (eg Bergen, see Hansen Chapter 14), complicating their
attribution to a more holistic western European colonial package (cf Bartlett 1994,
167–172).
The new medieval coastal towns are likely to have been partly a response to
major increases in both regional and long-range exchanges of low-value staple goods
— perhaps even more so than conventional wisdom allows (Barrett et al 2004,
630–631 and references therein), and contrary to theories emphasizing large-scale
bulk trade earlier in the 1st millennium ad (eg Hodges 2012; Sawyer 2013). This is
not to suggest that pre- and early Viking Age exchange was unimportant (see Skre
Chapter 12), but the growing volume of goods involved during the Viking Age–
medieval transition was both a contributor to and a symptom of major socio-
economic transformation. The reality of change is shown by an order of magnitude
increase in Scandinavian ship capacities between the 9th and 11th centuries — made
possible in part by the adoption of the sail (Bill 2010; Crumlin-Pedersen 1999;
Westerdahl Chapter 2). It is also supported by evidence ranging from more fish bones
in middens (Barrett 2012; Barrett et al 2004) to shifting political priorities. Viking Age
trade was often a corollary of warfare — a clear example being the growth of
Dublin’s mercantile community around a military encampment turned mercenary
base (Downham Chapter 24; cf Williams 2013). Later medieval military intervention
in northern Europe was increasingly initiated by the need to promote trade (eg
Munzinger 2006; Nedkvitne 2014, 50). These economic shifts must have been com-
bined with a concurrent transformation in social perceptions of the sea, which had
become a road to opportunity in areas where it had previously been a landscape of
risk (Mägi Chapter 4; cf Loveluck 2013). During the Viking Age to medieval
MAM 37: 01-barrett – Press

the transformation of the viking age and medieval world 9

transition there were thus quantitative and qualitative changes in the mosaic of socio-
economic choices made by the peoples of the Baltic, North and Irish Seas. The island
and coastal societies studied in this volume were both a cause and a consequence.1

note
1 I thank the contributors to this volume for their my present research. Aleks Pluskowski drew my
inspiration and the Leverhulme Trust for funding attention to Munzinger (2006).

references
Abrams, L, 2012 ‘Diaspora and identity in the Viking Age’, Early Medieval Europe 20, 17–38
Abrams, L, 2013 ‘Early Normandy’, Anglo-Norman Studies 35, 45–64
Abrams, L and Parsons, D N, 2004 ‘Place-names and the history of Scandinavian settlement in
England’, in J Hines, A Lane and M Redknap (ed), Land, Sea and Home: Proceedings of a
Conference on Viking-period Settlement, Maney/Society for Medieval Archaeology, Leeds,
379–431
Appadurai, A, 1996 Modernity at Large: Cultural Dimensions of Globalization, University of
Minnesota Press, Minneapolis
Arneborg, J, Nyegaard, G and Vésteinsson, O (ed), 2009–10 Norse Greenland: Selected Papers
from the Hvalsey Conference 2008, Journal of the North Atlantic Special Volume 2, Eagle Hill
Institute, Steuben, Maine
Astill, G, 1991 ‘Towns and town hierarchies in Saxon England’, Oxford Journal of Archaeology
10, 95–117
Bagge, S, 2010 From Viking Stronghold to Christian Kingdom: State Formation in Norway, c.900–
1350, Museum Tusculanum Press, Copenhagen
Barrett, J H (ed), 2003 Contact, Continuity and Collapse: The Norse Colonization of the North
Atlantic, Brepols, Turnhout
Barrett, J H, 2004 ‘Beyond war or peace: the study of culture contact in Viking-age Scotland’, in
J Hines, A Lane and M Redknap (ed), Land, Sea and Home: Proceedings of a Conference on
Viking-period Settlement, Maney/Society for Medieval Archaeology, Leeds, 207–218
Barrett, J H, 2007 ‘The pirate fishermen: the political economy of a medieval maritime society’, in
B B Smith, S Taylor and G Williams (ed), West over Sea: Studies in Scandinavian Sea-borne
Expansion and Settlement before 1300, Brill, Leiden, 299–340
Barrett, J H, 2010 ‘Rounding up the usual suspects: causation and the Viking Age diaspora’, in
A Anderson, J H Barrett and K V Boyle (ed), The Global Origins and Development of
Seafaring, McDonald Institute for Archaeological Research, Cambridge, 289–302
Barrett, J H (ed), 2012 Being an Islander: Production and Identity at Quoygrew, Orkney,
ad 900–1600, McDonald Institute for Archaeological Research, Cambridge
Barrett, J H and Anderson, A, 2010 ‘Histories of global seafaring: a discussion’, in A Anderson,
J H Barrett and K V Boyle (ed), The Global Origins and Development of Seafaring,
McDonald Institute for Archaeological Research, Cambridge, 305–314
Barrett, J, Beukens, R, Simpson, I, Ashmore, P, Poaps, S and Huntley, J, 2000 ‘What was the
Viking Age and when did it happen? A view from Orkney’, Norwegian Archaeological
Review 33, 1–39
Barrett, J H, Locker, A M and Roberts, C M, 2004 ‘‘‘Dark Age Economics’’ revisited: the English
fish bone evidence ad 600–1600’, Antiquity 78, 618–636
Barrow, J, 2008 ‘The chronology of the Benedictine ‘‘reform’’’, in D Scragg (ed), Edgar, King of the
English, 959–975: New Interpretations, The Boydell Press, Woodbridge, 211–223
MAM 37: 01-barrett – Press

10 maritime societies of the viking and medieval world

Bartlett, R, 1994 The Making of Europe: Conquest, Colonization and Culture Change 950–1350,
Penguin, London
Beresford, M W, 1967 New Towns of the Middle Ages: Town Plantation in England, Wales, and
Gascony, Praeger, New York
Bill, J, 2010 ‘Viking Age ships and seafaring in the west’, in I Skibsted Klæsøe (ed), Viking Trade
and Settlement in Continental Western Europe, Museum Tusculanum Press, Copenhagen,
19–42
Blackburn, M, 2008 ‘Currency under the Vikings part 4: the Dublin coinage c.995–c.1050’, British
Numismatic Journal 78, 111–137
Blackburn, M, 2011 ‘The Viking winter camp at Torksey, 872–3’, in M Blackburn (ed), Viking
Coinage and Currency in the British Isles, Spink, London, 221–264
Bolton, T, 2009 The Empire of Cnut the Great: Conquest and the Consolidation of Power in
Northern Europe in the Early Eleventh Century, Brill, Leiden
Bonde, N and Stylegar, F-A, 2009 ‘Fra Avaldsnes til Oseberg: dendrokronologiske undersøkelser
av skipsgravene fra Storhaug og Grønhaug på Karmøy’, Viking 72, 149–168
Bornholdt-Collins, K A, 1999 ‘Myth or mint? The evidence for a Viking-age coinage in the Isle of
Man’, in P Davey (ed), Recent Archaeological Research on the Isle of Man, British Archaeo-
logical Reports British Series 278, Oxford, 199–220
Bowden, G R, Balaresque, P, King, T E, Hansen, Z, Lee, A C, Pergl-Wilson, G, Hurley, E, Roberts,
S J, Waite, P, Jesch, J, Jones, A L, Thomas, M G, Harding, S E and Jobling, M A, 2007
‘Excavating past population structures by surname-based sampling: the genetic legacy of the
Vikings in northwest England’, Molecular Biology and Evolution 25, 301–309
Bradley, J, 1988 ‘The interpretation of Scandinavian settlement in Ireland’, in J Bradley (ed), Settle-
ment and Society in Medieval Ireland: Studies Presented to F.X. Martin, O.S.A., Boethius,
Kilkenny, 49–78
Brooks, N P, 1979 ‘England in the ninth century: the crucible of defeat’, Transactions of the Royal
Historical Society 29, 1–20
Broun, D, 1994 ‘The origin of Scottish identity in its European context’, in B E Crawford (ed),
Scotland in Dark Age Europe, The Committee for Dark Age Studies, University of St.
Andrews, St. Andrews, 21–32
Campbell, A (ed), 1962 The Chronicle of Æthelweard, Thomas Nelson and Sons Ltd, London
Carver, M, 1995 ‘The North Sea highway: cultural arena or political barrier’, in J M Fladmark
(ed), Sharing the Earth: Local Identity in Global Culture, Donhead Publishing, London,
177–190
Coupland, S, 1998 ‘From poachers to gamekeepers: Scandinavian warlords and Carolingian kings’,
Early Medieval Europe 7, 85–114
Crawford, B E, 2004 ‘Thorfinn (II) Sigurdson (c.1009–c.1065)’, in Oxford Dictionary of National
Biography, Oxford University Press, Oxford, http://www.oxforddnb.com/view/article/49271
Critch, A, forthcoming ‘How are princely gifts repaid by your powerful friends?’: ring-money and
the appropriation of tradition in Insular Viking politics, ad 900–1065, Unpublished PhD
thesis, University of Cambridge, Cambridge
Crumlin-Pedersen, O, 1999 ‘Ships as indicators of trade in Northern Europe 600–1200’, in J Bill
and B L Clausen (ed), Maritime Topography and the Medieval Town, The National Museum
of Denmark, Copenhagen, 11–20
Downham, C, 2007 Viking Kings of Britain and Ireland: The Dynasty of Ívarr to a.d . 1014,
Dunedin Academic Press, Edinburgh
Duczko, W, 2004 Viking Rus: Studies on the Presence of Scandinavians in Eastern Europe, Brill,
Leiden
MAM 37: 01-barrett – Press

the transformation of the viking age and medieval world 11

Dugmore, A J, Church, M J, Buckland, P C, Edwards, K J, Lawson, I, McGovern, T H, Panagio-

takopulu, E, Simpson, I A, Skidmore, P and Sveinbjarnardóttir, G, 2005 ‘The Norse landnám
on the North Atlantic islands: an environmental impact assessment’, Polar Record 41, 21–37
Dumville, D N, 1997 The Churches of North Britain in the First Viking Age, Whithorn Trust,
Whithorn
Ferguson, Y H and Mansbach, R W, 2012 Globalization: The Return of Borders to a Borderless
World?, Routledge, London
Feveile, C, 2010 ‘Ribe: continuity or discontinuity from the eighth to the twelfth century?’, in
J Sheehan and D Ó Corráin (ed), The Viking Age: Ireland and the West, Four Courts Press,
Dublin, 97–106
Glørstad, Z T, 2014 ‘Homeland — strange land — new land: material and theoretical aspects of
defining Norse identity in the Viking Age’, in J V Sigurðsson and T Bolton (ed), Celtic-Norse
Relationships in the Irish Sea in the Middle Ages 800–1200, Brill, Leiden, 151–170
Graham-Campbell, J, 1995 The Viking-Age Gold and Silver of Scotland, National Museums of
Scotland, Edinburgh
Hadley, D and Richards, J D, 2013 ‘Viking Torksey: inside the Great Army’s winter camp’,
Current Archaeology 281, 12–19
Hadley, D M and ten Harkel, L (ed), 2013 Everyday Life in Viking-age Towns: Social Approaches
to Towns in England and Ireland, c. 800–1100, Oxbow Books, Oxford
Hernæs, P, 1997 ‘Storpolitikk og vikingtog på slutten av 700-tallet’, in I Fuglestvedt and B Myhre
(ed), Konflikt i Forhistorien, Arkeologisk Museum i Stavanger, Stavanger, 57–67
Hodges, R, 2012 Dark Age Economics: A New Audit, Bristol Classical Press, London
Holm, P, 1986 ‘The slave trade of Dublin, ninth to twelfth centuries’, Peritia 5, 317–345
Kershaw, J F, 2013 Viking Identities: Scandinavian Jewellery in England, Oxford University Press,
Oxford
Kershaw, P, 2000 ‘The Alfred-Guthrum Treaty: scripting accommodation and interaction in
Viking Age England’, in D M Hadley and J D Richards (ed), Cultures in Contact: Scandi-
navian Settlement in England in the Ninth and Tenth Centuries, Brepols, Turnhout, 43–64
Kilger, C, 2008 ‘Wholeness and holiness: counting, weighing and valuing silver in the early Viking
period’, in D Skre (ed), Means of Exchange: Dealing with Silver in the Viking Age, Aarhus
University Press, Aarhus, 253–325
Lebecq, S, 2007 ‘Communication and exchange in northwest Europe’, in J Bately and A Englert
(ed), Ohthere’s Voyages: A Late 9th-century Account of Voyages Along the Coasts of Norway
and Denmark and its Cultural Context, The Viking Ships Museum, Roskilde, 170–179
Lieberman, V, 2009 Strange Parallels: Southeast Asia in Global Context, c. 800–1830, Volume 2:
Mainland Mirrors: Europe, Japan, China, South Asia, and the Islands, Cambridge University
Press, Cambridge
Loe, L, Boyle, A, Webb, H and Score, D, 2014 ‘Given to the Ground’: A Viking Age Mass Grave on
Ridgeway Hill, Weymouth, Oxford Archaeology, Oxford
Loveluck, C, 2013 Northwest Europe in the Early Middle Ages c. ad 600–1150, Cambridge
University Press, Cambridge
Loveluck, C and Tys, D, 2006 ‘Coastal societies, exchange and identity along the Channel and
southern North Sea shores of Europe, ad 600–1000’, Journal of Maritime Archaeology 1,
140–169
MacAirt, S and MacNiocaill, G (ed), 1983 The Annals of Ulster (to a.d. 1131), Dublin Institute for
Advanced Studies, Dublin
Macniven, A, 2013 ‘Modelling Viking Migration to the Inner Hebrides’, in A Jennings and
A Sanmark (ed), Across the Sólundarhaf: Connections between Scotland and the Nordic
MAM 37: 01-barrett – Press

12 maritime societies of the viking and medieval world

World, Journal of the North Atlantic Special Volume 4, Eagle Hill Institute, Steuben, Maine,
3–18
McLeod, S, 2014 Beginning of Scandinavian Settlement in England: The Viking Great Army and
Early Settlers, c. 865–900, Brepols, Turnhout
McGuire, R H, 1992 A Marxist Archaeology, Academic Press, London
Mehler, N, 2013 ‘Globalization, immigration, and transformation: thoughts from a European per-
spective’, Historical Archaeology 47, 38–49
Munzinger, M R, 2006 ‘The profits of the Cross: merchant involvement in the Baltic Crusade
(c.1180–1230)’, Journal of Medieval History 32, 163–185
Montgomery, J, Grimes, V, Buckberry, J, Evans, J A, Richards, M P and Barrett, J H, 2014 ‘Find-
ing Vikings with isotope analysis: The view from wet and windy islands’, Journal of the
North Atlantic 7, 54–70
Myrberg, N, 2010 ‘A worth of their own: on Gotland in the Baltic Sea, and its 12th-century
coinage’, Medieval Archaeology 54, 157–181
Naum, M, 2012 ‘Difficult middles, hybridity and ambivalence of a medieval frontier: the cultural
landscape of Lolland and Falster (Denmark)’, Journal of Medieval History 38, 56–75
Nedkvitne, A, 2014 The German Hansa and Bergen 1100–1600, Böhlau Verlag, Köln
Nelson, J L, 1991 The Annals of St-Bertin, Manchester University Press, Manchester
Neveux, F, 2002 ‘L’héritage des Vikings dans la Normandie ducale’, in É Ridel (ed), L’héritage
maritime des Vikings en Europe de l’Ouest, Presses universitaires de Caen, Caen, 101–118
Palla, G, Barabási, A-L and Vicsek, T, 2007 ‘Quantifying social group evolution’, Nature 446,
664–667
Pollard, A M, Ditchfield, P, Piva, E, Wallis, S, Falys, C and Ford, S, 2012 ‘‘‘Sprouting like cockle
amongst the wheat’’: the St Brice’s Day massacre and the isotopic analysis of human bones
from St John’s College, Oxford’, Oxford Journal of Archaeology 31, 83–102
Price, N, 2002 The Viking Way: Religion and War in Late Iron Age Scandinavia, Uppsala
University Press, Uppsala
Radtke, C, 2006 ‘Money, port and ships from a Schleswig point of view’, in L Blue, F Hocker and
A Englert (ed), Connected by the Sea: Proceedings of the Tenth International Symposium on
Boat and Ship Archaeology Roskilde 2003, Oxbow Books, Oxford, 147–151
Risvaag, J A and Christophersen, A, 2004 ‘Early medieval coinage and urban development: a
Norwegian experience’, in J Hines, A Lane and M Redknap (ed), Land, Sea and Home: Pro-
ceedings of a Conference on Viking-period Settlement, Maney/Society for Medieval Archae-
ology, Leeds, 75–91
Reuter, T, 1985 ‘Plunder and tribute in the Carolingian Empire’, Transactions of the Royal
Historical Society 35, 75–94
Sawyer, P H, 1962 The Age of the Vikings, Edward Arnold, London
Sawyer, P, 2013 The Wealth of Anglo-Saxon England, Oxford University Press, Oxford
Sigurdsson, J V, 2008 Det Norrøne Samfunnet: Vikingen, Kongen, Erkebiskopen og Bonden, Pax
Forlag A/S, Oslo
Sindbæk, S M, 2007 ‘The Small World of the Vikings: networks in early medieval communication
and exchange’, Norwegian Archaeological Review 40, 59–74
Sindbæk, S M, 2011 ‘Silver economies and social ties: long-distance interaction, long-term invest-
ments — and why the Viking Age happened’, in J Graham-Campbell, S M Sindbæk and
G Williams (ed), Silver Economies, Monetisation and Society in Scandinavia, ad 800–1100,
Aarhus University Press, Aarhus, 41–66
Sindbæk, S M, 2012a ‘Enter the gripping beast: innovations and actor-networks in Viking Age
towns’, in R Berge, M E Jasinski and K Sognnes (ed), N-TAG Ten: Proceedings of the
MAM 37: 01-barrett – Press

the transformation of the viking age and medieval world 13

10th Nordic TAG Conference at Stiklestad, Norway, 2009, British Archaeological Reports
International Series S2399, Archaeopress, Oxford, 129–135
Sindbæk, S M, 2012b ‘Viking disruptions or growing integration? Contextualising communication
networks in the 10th century North Sea’, in S Kleingärtner and G Zeilinger (ed), Raumbildung
durch Netzwerke? Der Ostseeraum zwischen Wikingerzeit und Spätmittelalter aus archäo-
logischer und geschichtswissenschaftlicher Perspektive, Dr. Rudolf Habelt, Bonn, 19–38
Sindbæk, S M and Trakadas, A (ed), 2014 The World in the Viking Age, The Viking Ship Museum,
Roskilde
Skre, D (ed), 2008 Means of Exchange: Dealing with Silver in the Viking Age, Aarhus University
Press, Aarhus
Steuer, H, 1987 ‘Gewichtsgeldwirtschaften im frühgeschichtlichen Europa — Feinwaagen und
Gewichte als Quellen zur Währungsgeschichte’, in K Düwel, H Jahnkuhn, H Siems and
D Timpe (ed), Untersuchungen zu Handel und Verkehr der vor- und frühgeschichtlichen Zeit
in Mittel- und Nordeuropa: Teil IV: Der Handel der Karolinger- und Wikingerzeit, Vanden-
hoeck & Ruprecht, Göttingen, 405–527
Svanberg, F, 2003 Decolonizing the Viking Age 1, Almqvist & Wiksell International, Stockholm
Theuws, F, 2004 ‘Exchange, religion, identity and central places in the early Middle Ages’,
Archaeological Dialogues 10, 121–138
Thomas, G, 2012 ‘Carolingian culture in the North Sea world: rethinking the cultural dynamics of
personal adornment in Viking age England’, European Journal of Archaeology 15, 486–518
Thurston, T L, 2001 Landscapes of Power, Landscapes of Conflict: State Formation in the South
Scandinavian Iron Age, Springer Scientific, New York
Tilly, C, 1975 The Formation of National States in Western Europe, Princeton University Press,
Princeton
Vésteinsson, O, Lucas, G, Þórsdóttir, K and Gylfadóttir, R (ed), 2011 Upp á Yfirborðið: Nýjar
Rannsóknir ı́ Íslenskri Fornleifafræði, Fornleifastofnun Íslands, Reykjavı́k
Wallace, P F, 1987 ‘The economy and commerce of Viking Age Dublin’, in K Düwel, H Jahnkuhn,
H Siems and D Timpe (ed), Untersuchungen zu Handel und Verkehr der vor- und frühge-
schichtlichen Zeit in Mittel- und Nordeuropa: Teil IV: Der Handel der Karolinger- und
Wikingerzeit, Vandenhoeck & Ruprecht, Göttingen, 200–245
Westerdahl, C, 2008 ‘Boats apart: building and equipping an Iron-Age and early-medieval ship in
Northern Europe’, The International Journal of Nautical Archaeology 37, 17–31
Whitelock, D, Douglas, D C and Tucker, S I (ed), 1961 The Anglo-Saxon Chronicle, Eyre &
Spottiswoode, London
Wiechmann, R, 2007 ‘Hedeby and its hinterland: a local numismatic region’, in J Graham-
Campbell and G Williams (ed), Silver Economy in the Viking Age, Left Coast Press, Walnut
Creek, California, 29–48
Williams, G, 2013 ‘Towns and identities in Viking England’, in D M Hadley and L ten Harkel (ed),
Everyday Life in Viking-age Towns: Social Approaches to Towns in England and Ireland,
c. 800–1100, Oxbow Books, Oxford, 14–34
Winroth, A, 2014 The Age of the Vikings, Princeton University Press, Princeton
Woods, A, 2013 ‘The coinage and economy of Hiberno-Scandinavian Dublin’, in S Duffy (ed),
Medieval Dublin XIII, Four Courts Press, Dublin, 43–69
Woolf, A, 2007 From Pictland to Alba: 789–1070, Edinburgh University Press, Edinburgh
Wormald, C P, 1982 ‘Viking Studies: whence and whither?’, in R T Farrell (ed), The Vikings,
Phillimore & Co., London, 128–153
Zori, D and Byock, J (ed), 2014 Viking Archaeology in Iceland: Mosfell Archaeological Project,
Brepols, Turnhout
MAM 37: 02-westerdahl – Press

chapter 2
SAILS AND THE COGNITIVE ROLES OF
VIKING AGE SHIPS

By Christer Westerdahl

introduction
The Viking Age is part of the European Middle Ages, but in its context it repre-
sents the final centuries of the Iron Age. It is interesting as a transitional period, in
certain important dimensions, most connected with power and all interconnected
with each other, such as Christianization and the all-pervading appearance of kings.
Two decades ago I argued that ships were more important in the Viking Age than
before or after (Westerdahl 1993). However, I have changed my mind to some extent
— not in the sense of the ship shaping the post-mortem picture of the Viking Age, but
regarding the significance of the boat to ordinary people. Judith Jesch (2001, 275) put
it in so many words:
Although the words ‘viking’ and ‘ship’ so often seem to go together, ships were not
necessarily more important to the Scandinavians in the Viking Age than in any
other time in their history. The Viking Age may just have been when other nations
became more keenly aware of Scandinavian nautical prowess.
The Viking Age was not the only period when ships had a particular social and
symbolic significance in Northern Europe. Nor was only that small section of society
that we call the Vikings (those with the resources to equip expeditions) the only part
of a repressive and highly hierarchical world that was dependant on ships and boats.
Waterways were the main networks of communication and transport, in addition to
providing fish, sea mammals, seabirds and other resources. Moreover, one might
question the historiographical exaltation of the Viking Age as nationalistic, roman-
ticist and ethnocentric (eg Svanberg 2003; Westerdahl 2004). As more or less a con-
temporary of the Swedish historian Erik Lönnroth, I empathize with the sentiment of
his 1947 review of the iconic novel Röde Orm by F G Bengtsson:
It is a magnificent testimony that the scalds offer on the spirit of the Viking Age.
But it is monotonous and its meaning is terrible. Behind the gorgeous imagery is a
sea and a world as desolate as the empty eyes of the dragon heads, where the long
ships rested as little as wind, waves and the rapaciousness of men. (Lönnroth 1961;
my translation)
MAM 37: 02-westerdahl – Press

sails and the cognitive roles of viking age ships 15

That said, it must be admitted that ships are a special case, partly because the
culture of the north in general was maritime and partly because there was something
especially significant about ships within Nordic society, particularly if not exclusively
in the Viking Age. The ship-formed stone settings (multi-period, but common during
the Viking Age), the numerous bog finds of vessels or parts of vessels, the ships carved
on picture and rune stones and the inclusion of ships in graves (burnt or unburnt) all
point to this conclusion. There is plenty of information from history and archaeology
regarding the importance of sea-going vessels in the Viking Age. A handy introduc-
tion by one of the principal actors is offered by Crumlin-Pedersen (2010). Moreover,
one must ultimately explain why the sail was adopted between approximately ad 750
and 820. It is this last issue that I wish to focus on in the present chapter, revisiting a
topic I first considered many years ago (cf Westerdahl 1995).

the sail in the north

The present state of archaeological research tells us that the sail was adopted
first during the Viking Age in Scandinavia. Particularly interesting is the question why
the sail was adopted so late and seemingly hesitantly by all the peoples of the north
(not only Germanic groups). The technical advantages of sail over oar appear so
obvious to our time and our context, and northerners knew well the existence of the
sail, even its technicalities, from close contact with the Roman world (Figure 2.1). In
1995 I suggested three explanations. Two were functional and made mainly military
sense. The first maintained that pre-Viking Age society was very much a martial one,
but with an emphasis on surprise raids where you did not want to be seen in advance.
Sails would spoil stealth. The second functional explanation was the need for
coordination in such raids, which relied on multiple vessels. One could not rely on a
fleet driven by fickle winds (Figure 2.2). Rowing time could nearly always be com-
puted, especially with a high degree of technical sophistication in the process of
rowing, something which can be assumed with confidence for this period. My third
explanation was a strong social and cognitive conservatism which involved being

Figure 2.1 The Roman ship carved on a

cattle bone thrown in the River Weser,
Germany, dated to the 5th century ad. Legible
runes of the 24-type variety tell us (probably)
that ‘we are coaxing them (the Romans?) here’
(after Pieper 1989)
MAM 37: 02-westerdahl – Press

16 maritime societies of the viking and medieval world

Figure 2.2 Contrasting rowing (Nydam) and sailing (Gokstad) ships (Sune Villum-Nielsen,
after Westerdahl 1995)

part of a particular rowing crew, a comitatus-type segment of a fundamentally row-

ing society owing allegiance to a chieftain: one man, one oar, one rowlock. In the
Nydam ships of ad c400 it appears that all rowlocks (and perhaps even the oars)
were individually made (Rieck 1995). Perhaps the depiction of the first sails on some
Gotlandic picture stones of a hanfot system of braces (Figure 2.3) in the hands of
almost all the members of the crew is a nostalgic remembrance of rowing as a social
act? I argue further, then and now, that during the Viking Age the sought-after legiti-
macy of the new royal rulers paved the way for a new ideology where leaders wanted
to be seen. The display of large fleets was a prerequisite for intimidation and enforced
domination of a totally different kind from that achieved by earlier hit-and-run
tactics. In a sailing ship the crew is inactive. Sails propel the vessel. The winds are
governed by powers superior to men. Only kings would thrive in such a system.
And in fact they do, according to the imagery of royal court poetry. Only they would
depend on chance and a direct lead from the Gods. Or on Grace from the Lord
himself.
The metaphor of rowing must, however, have been strong even in the days of
sail. In much later medieval provincial laws (which include incipient attempts to
impose efficient taxation) a metaphorical rowing society is conjured up, very probably
MAM 37: 02-westerdahl – Press

sails and the cognitive roles of viking age ships 17

(a)

Figure 2.3 (a) The hanfot bracing

system of the ship depicted on a
picture stone from Smiss I in
Stenkyrka parish, Gotland. The
deceased person to whom the stone
is raised is probably sitting at the
stern. It seems that all crew members
are holding the ends of the braces
(after Nylén and Lamm);
(b) Gotlandic picture stone
Ardre VIII with horse and ship
(b) (C Westerdahl)
MAM 37: 02-westerdahl – Press

18 maritime societies of the viking and medieval world

petrified and archaized, but still functional. We know that basically rowing ships
were still used as ledung/leidang (obligatory ship-levy) vessels into the 14th and 15th
centuries in some cases. Arable lands in the Nordic kingdoms of the Middle Ages
were divided into units corresponding to the archaic principle of one man, one oar,
one bench. Hå/hamna (and equivalents), which literally meant rowlock and fastening
for the oar, were used in the ship levies of the medieval provincial laws to denote the
smallest unit of taxation, a couple of farmsteads or sometimes a hamlet. However,
this metaphor need not hark back entirely to the period before the Viking Age.
Crumlin-Pedersen (1997, 189) has pointed out that the drastic widening of ship beams
that provided stability in the first period of the sail was followed by a return to pre-
Viking long and slender warships (in combination with sails) precisely ‘to maximize
the effect of rowing’ in the last period of the Viking Age (the 10th and 11th centuries).
The last major ship finds without any arrangements for a mast include the
sacrificial Kvalsund boats of west Norway, which date to the 7th or 8th century ad
(see Chapter 13), and the Storhaug and Grønhaug burials from Karmøy, dated to
approximately ad 770 and ad 780 respectively (Bonde and Stylegar 2009; Opedal
1998). The first find with a mast-step, although rather a weak one, is the famous
burial ship from Oseberg in Vestfold, dated by dendrochronology to approximately
ad 820, but deposited in ad 834 (Bonde and Christensen 1993). The time for sails may
have been ripe. The Kvalsund finds had developed a Nordic T-formed keel, con-
sidered an important step towards reducing leeway (Shetelig and Johannessen 1929;
Figure 2.4).
All agree that the oldest depictions of sailing ships in the north are those of the
Gotlandic picture stones (a late group of them: see Figure 2.3). Less known inter-
nationally is the redating of Lindquist’s (1941–42) chronological scheme by Varenius
(1992), which moves the relevant images forward in time. Moreover, this revision has
been confirmed and made even younger by the research of Imer (2004). It is now
thought that the Gotlandic picture stones with sail all belong to the Viking Age,
although whether it started around ad 750 or ad 800 is still an open question. Thus it
is no longer appropriate to argue that these depictions of sailing ships and boats bring
us back to the 6th and 7th centuries ad as was once believed (eg Thier 2003, 184).
Thus an early Viking Age date remains the generally accepted opinion despite
efforts to put the innovation back in time among the north Germanic peoples. My
explanations discussed above therefore remain plausible. However, other of my pre-
vious ideas (Westerdahl 1995, 47) have faded. I am no longer convinced that sails
were important as media for symbols and heraldic figures, although this does still
apply in the case of the cross on the sail of the Sparlösa stone ad c800 (Figure 2.5;
Westerdahl 1996; 2011). Nevertheless, it is obvious that the Gotlandic depictions of
sails (Figure 2.3) contain information of a symbolic character from the very begin-
ning. If they connote the divine ship — parallel in this context to the divine horse
Sleipnir — that ship may be thought of as Skiðblaðnir, always provided with a fair
wind (Westerdahl 1995, 46).
Critics have approached the dating of the first Nordic sail by archaeology in
different ways. Sailing enthusiasts of modern times cannot believe in it. The Sutton
Hoo ship in the 7th century, they claim, could have been sailed (Gifford and Gifford
MAM 37: 02-westerdahl – Press

sails and the cognitive roles of viking age ships 19

Figure 2.4 Cross-sections of two sacrificial boats from Kvalsund, Nerlandsøy, western
Norway, which have the oldest known T-formed keels, normally considered a prerequisite for
sailing vessels (after Shetelig and Johannessen 1929)

1996). Timm Weski (1998) thinks that the journeys of the Saxon invaders of Britain
could not have been made only by rowing, despite the testimony of Procopius
(writing in ad c550; albeit spatially removed from the events described). Weski points
to a very early find: a holed rib in the stem part of the 2nd-century ad Lecker Au log
boat from Dithmarschen in northern Germany, which is of long (c13.5m) and slender
construction (Weski 1998, 68). However, the overall character of the boat makes
interpretation of this rib as a mast-step rather improbable. A mounting point for a
hauling pole seems more likely.
Others point to alleged Saxon sailing mentioned during the 5th and 6th centuries
ad (Haywood 1999; cf Thier 2003). However, these details are found only in a few
(three) texts, and only one seems at all convincing. The others are ambiguous in
meaning, possibly using sailing as a general term for travelling at sea or using a boat.
A seemingly early rune stone, the Eggja monument, has been dated by Ottar
Grønvik (1985) to the end of the 7th century ad. He believes that a mast-detail is
MAM 37: 02-westerdahl – Press

20 maritime societies of the viking and medieval world

Figure 2.5 The Sparlösa

picture stone, which dates
to ad c800, bears a ship and
rider scene with a house on
top that resembles the
figural arrangement of
several Gotlandic stones.
Note the cross on the sail
(C Westerdahl)

mentioned and that the inscription may refer to a shipwreck. Both dating and inter-
pretation are difficult and may reasonably be problematic.
Another approach is to date the appearance of mast stones in the middle of stone
settings. Such cases are known, but appear to be at least Late Iron Age or rather
Viking Age in date (Capelle 1986). Moreover, the objection weighs heavily that a
symbolic ship in the ground might have had cosmological connotations where the
centre of the vessel space would be marked for other reasons. An informed philo-
logical discussion on the introduction of the Germanic word sail has been provided
MAM 37: 02-westerdahl – Press

sails and the cognitive roles of viking age ships 21

by Katrin Thier (2003) where she points to a transfer from Celtic along the Rhine.
However, nothing new on the dating of the sail in the north has come out of this.
Early Irish boats, perhaps curraghs of animal hide, may in fact have been the first to
use sail in the north based, for example, on the poorly dated Broighter model (see
Marcus 1980).
In sum, the Nordic conception of ships must have been heavily influenced by the
introduction of the sail in the period ad 750 to 820. It may be informative to look at
the reaction of by-standers to this rapid development. In the far north, recently dis-
covered rock art in the inland mountains of Arctic Sweden record Sami perceptions of
Nordic ships (Mulk and Bayliss Smith 2006). These motifs (Figure 2.6) are so far
unique in their setting, and may belong to an early part of the Viking Age (cf the ship
on the Sparlösa rune stone of Västergötland dated ad c800 (Figure 2.5; Westerdahl
1996; 2011). Perhaps the first sailing ships were thought remarkable by the Sami,
although they were experienced in boat culture long before that. The magic use of
ship depictions may have had a background in the Sami cultural world, but provided
a means of expression in the context of change.
Most or all Viking Age sails were made of wool (Andersson 2007; Bender
Jørgensen 2012; Möller-Wiering 2007). It is obvious that a prerequisite for sailing was
the large-scale surplus production of this raw material. Moreover, the technology
for the production and refinement of sailcloth was not created overnight. It is prob-
able that the original coastal heather landscapes of western Scandinavia and other
parts of the Atlantic coasts of Europe – sheep-grazing lands – are an effect of this.
Some dates of the creation of coastal heaths in west Norway point to the middle of
the 8th century ad (Bender Jørgensen 2005; 2012 and references therein). The adop-
tion of the sail had non-trivial implications for the costs of producing a vessel,
creating a demand on resources and labour commensurate with that of a hull

Figure 2.6 Sailing ships illustrated on a rock in Sami mountain territory in Padjelanta,
northern Sweden. Compare with Figure 2.5 (after Mulk and Bayliss-Smith 2006)
MAM 37: 02-westerdahl – Press

22 maritime societies of the viking and medieval world

(Andersen 1995, 250). It is in this context too, that sailing is broadly correlated with
the growth of kingship in Scandinavia (cf Westerdahl 2008; Barrett and Anderson
2010).
The immediate candidates for introducing the sail as an innovation from the
outside are the Frisian sailing merchants (Lebecq 1983). Their appearance on the
historical stage coincides with the rise of the first proto-urban sites in the north —
initially Ribe around ad 710–720, followed rapidly by Birka, Hedeby and Kaupang.
On the other hand, the sailing arrangements (keelson with mast-step) and termi-
nology of Scandinavia were adopted elsewhere in western Europe. Thus the idea may
have been received from Frisians, but the actual shaping of it was at least partly a
native one. It is less likely that sails of the river boats of the east were adopted by
Scandinavians, although the possibility exists of multiple influences. Larsson (2000;
2007, 97), for example, has noted possible Byzantine mushroom-shape sail forms on
early Gotlandic picture stones and discussed Russian parallels.
In practice the square sail on one mast was adopted, and would reign supreme
into the late Middle Ages. To give a comparative perspective, in the (eastern)
Mediterranean the square sail was in use from at least the 3rd millennium bc until the
lateen sail was adopted in Egypt during the 7th century ad (Basch 1997). It then took
at least another 600 years before the Mediterranean maritime cultures reintroduced
square sails in earnest, this time together with innovations from the northern cog,
such as the stern rudder. Looking back further, there was a millennium or two
between the introduction of sea-going sailing vessels in the eastern Mediterranean
in the 3rd millennium ad and their local adoption in the western Mediterranean,
despite documented interregional contact (Broodbank 2010). The explanation of this
curious disjunction and the later example of northern Europe may have meaningful
resonances, both being in part the result of differences in political organization
between regions, with concomitant implications for ideology, warfare and access to
resources (cf Barrett and Anderson 2010).

conclusions
In concluding, it appears that the Viking Age (and its posthumous reputation)
would be inconceivable without ships with sails. In fact the introduction of the sail
seems contemporary with the Viking Age in Scandinavia, at least for ship types cross-
ing the North Sea and the Baltic. Concomitantly, the cognitive role of the one-masted
longship and its symbolic connotations are inextricable parts of the social changes of
the age: the growth of larger realms subordinated to warrior kings instead of petty
chieftains. The fleet and its ships, or a pars pro toto like a towering stempost and sail,
could stand as a metaphor for social life, including pagan religion, on-board fellow-
ship in a rowing crew structure (that may even have had implications for inland social
structure and territorial divisions) and a spirit of mercantile adventurism among the
magnates following their sea kings. There were certainly other symbolic factors
pertaining to the Viking Age coast, its routes and their landscapes, but they were also
bound up with the vessels and their constituent parts.
MAM 37: 02-westerdahl – Press

sails and the cognitive roles of viking age ships 23

references
Andersen, E, 1995 ‘Square sails of wool’, in O Olsen, J Skamby Madsen and F Riek (ed), Ship-
shape: Essays for Ole Crumlin-Pedersen, The Viking Ship Museum, Roskilde, 249–270
Andersson, E, 2007 ‘Textile tools and production during the Viking Age’, in C Gillis and
M-L B Nosch (ed), Ancient Textiles: Production, Craft and Society, Oxbow, Oxford, 17–25
Barrett, J H and Anderson, A, 2010 ‘Histories of global seafaring: A discussion’, in A Anderson,
J H Barrett and K V Boyle (ed), The Global Origins and Development of Seafaring,
McDonald Institute for Archaeological Research, Cambridge, 305–314
Basch, L, 1997 ‘L’apparition de la voile latine en Méditerranée’, in D Meeks and D Garcia (ed),
Techniques et économie antiques et médiévales : le temps de l’innovation, Éditions Errance,
Paris, 214–223
Bender Jørgensen, L, 2005 ‘Textiles of seafaring: An introduction to an interdisciplinary research
project’, in F Pritchard and J P Wild (ed), Northern Archaeological Textiles, Oxbow, Oxford,
65-69
Bender Jørgensen, L, 2012 ‘The introduction of sails to Scandinavia: Raw materials, labour and
land’, in R Berge, M E Jasinski and K Sognnes (ed), N-TAG Ten: Proceedings of the
10th Nordic TAG Conference at Stiklestad, Norway 2009, Archaeopress, Oxford, 173–181
Bonde, N and Christensen, A E, 1993 ‘Dendrochronological dating of the Viking Age ship burials
at Oseberg, Gokstad and Tune, Norway’, Antiquity 67, 575–583
Bonde, N and Stylegar, F-A, 2009 ‘Fra Avalsnes til Oseberg: Dendrokronologiske undersøkelser av
skipsgravene fra Storhaug og Grønhaug på Karmøy’, Viking 72, 149–168
Broodbank, C, 2010 ‘‘‘Ships a-sail from over the rim of the sea’’: Voyaging, sailing, and the making
of Mediterranean societies c. 3500–800 bc’, in A Anderson, J H Barrett and K V Boyle (ed),
The Global Origins and Development of Seafaring, McDonald Institute for Archaeological
Research, Cambridge, 249–264
Capelle, T, 1986 ‘Schiffsetzungen’, Praehistorische Zeitschrift 61, 1–63
Crumlin-Pedersen, O, 1997 ‘Large and small warships of the north’, in A Nørgård Jørgensen and
B Clausen (ed), Military Aspects of Scandinavian Society in a European Perspective,
ad 1–1300, National Museum of Denmark, Copenhagen, 184–194
Crumlin-Pedersen, O, 2010 Archaeology and the Sea in Scandinavia and Britain, Viking Ship
Museum, Roskilde
Gifford, E and Gifford, J, 1996 ‘The sailing performance of Anglo-Saxon ships as derived from the
building and trials of half-scale models of the Sutton Hoo and Graveney ship finds’, The
Mariner’s Mirror 82, 131–153
Grønvik, O, 1985 Runene på Eggjasteinen: en hedensk gravinnskrift fra slutten av 600-talet,
Universitetsforlaget, Oslo
Haywood, J, 1999 Dark Age Naval Power: A Reassessment of Frankish and Anglo-Saxon Sea-
faring Activity, Anglo-Saxon Books, Pinner, Middlesex
Imer, L, 2004 ‘Gotlandske billedsten: dateringen af Lindqvists gruppe C og D’, Aarbøger for
nordisk Oldkyndighed og Historie 2001, 47–111
Jesch, J, 2001 Ships and Men in the Late Viking Age, Boydell, Woodbridge
Larsson, G, 2000 ‘Contacts between central Sweden and Russia indicated by ship remains’, in
Study of the Marine Archaeology, Vol 4, Russian Academy of Sciences, Institute for the
History of Material Culture, St Petersburg, 102–114
Larsson, G, 2007 Ship and Society: Maritime Ideology in Late Iron Age Sweden, Uppsala Univer-
sitet: Department of Archaeology and Ancient History, Uppsala
Lebecq, S, 1983 Marchands et navigateurs frisons du haut Moyen Age, Presses universitaires de
Lille, Lille
MAM 37: 02-westerdahl – Press

24 maritime societies of the viking and medieval world

Lindquist, S, 1941–42 Gotlands Bildsteine I–II, Wahlström & Widstrand, Stockholm

Lönnroth, E, 1961 Från svensk medeltid, Aldus/Bonniers, Stockholm
Marcus, G W, 1980 The Conquest of the North Atlantic, The Boydell Press, Woodbridge
Möller-Wiering, S, 2007 ‘Under canvas’, in C Gillis and M-L B Nosch (ed), Ancient Textiles:
Production, Craft and Society, Oxbow, Oxford, 122–126
Mulk, I-M and Bayliss-Smith, T, 2006 Rock Art and Sami Sacred Geography in Badjelánnda,
Laponia, Sweden: Sailing Boats, Anthropomorphs and Reindeer, Department of Archaeology
and Sami Studies/Kungl. Skytteanska Samfundet, Umeå
Opedal, A, 1998 De glemte skipsgravene: makt og myter på Avaldsnes, Arkeologisk museum i
Stavanger, Stavanger
Pieper, P, 1989 Die Weser-Runenknochen. Neue Untersuchungen zur Problematik: Original oder
Fälschung, Isensee, Oldenburg
Rieck, F, 1995 ‘Ships and boats in the bog finds of Scandinavia’, in O Crumlin-Pederson and
B M Thye (ed), The Ship as Symbol in Prehistoric and Medieval Scandinavia, National
Museum of Denmark, Copenhagen, 125–129
Shetelig, H and Johannessen, F, 1929 Kvalsundsfunnet og andre norske myrfund av fartøier,
Bergen Museum, Bergen
Svanberg, F, 2003 Decolonizing the Viking Age 1, Almqvist & Wiksell International, Stockholm
Thier, K, 2003 ‘Sails in the north: New perspectives on an old problem’, International Journal of
Nautical Archaeology 32, 182–190
Varenius, B, 1992 Det nordiska skeppet: Teknologi och samhällsstrategi i vikingatid och medeltid,
Stockholms universitetet, Stockholm
Weski, T, 1998 ‘Anmerkungen zum Wasserverkehr während der römischen Kaiserzeit im Oder-
gebiet und auf der Ostsee’, in Beiträge zum Oderprojekt 5, Deutsches Archäologisches
Institut, Berlin, 59–74
Westerdahl, C, 1993 ‘Skepp och Farleder’, in N Lund (ed), Norden og Europa i vikingetid og tidlig
middelalder, Museum Tusculanums Forlag, København, 69–90
Westerdahl, C, 1995 ‘Society and sail: On symbols as specific social values and ships as catalysts of
social units’, in O Crumlin-Pederson and B M Thye (ed), The Ship as Symbol in Prehistoric
and Medieval Scandinavia, National Museum of Denmark, Copenhagen, 41–50
Westerdahl, C, 1996 ‘Sparlösa-stenen: Symboler och ‘‘politiska’’ attityder i tidig vikingatid’,
Västgötabygden 51(5), 16–21
Westerdahl, C, 2004 ‘Scando-Baltic contacts during the Viking Age’, in J Litwin (ed), Baltic Sea
Identity: Common Sea-Common Culture?, Centralne Muzeum Morskie w Gdańsku, Gdańsk,
27–34
Westerdahl, C, 2008 ‘Boats apart. Building and equipping an Iron-Age and early-medieval ship in
Northern Europe’, The International Journal of Nautical Archaeology 37, 17–31
Westerdahl, C, 2011 ‘Sparlösa, Kälvesten och Rök: Symboler och samhälle’, Västgötalitteratur
2011, 13–50
MAM 37: 03-gustin – Press

chapter 3
TRADE AND TRUST IN THE BALTIC SEA
AREA DURING THE VIKING AGE

By Ingrid Gustin

introduction
When the existence of different socio-political areas and regional identities in the
Baltic Sea region is discussed, one should remember that people did not live their lives
in isolation. People, then as now, had social relations that stretched to other regions
and societies. These could concern social contacts such as blood brotherhood and
marriage alliances, or trade relations and tribute dealings, as well as relations based
on common values (Burström 1991, 39; Callmer 1991).
For the central Baltic Sea region and the Russian river valleys the archaeological
material reveals evidence of the existence of diversified interaction. This is apparent
not only through the raw material, partially manufactured items, and prestige objects
that were mediated between different regions. It can also be seen through the
jewellery and the personal belongings that are found far away from where they were
manufactured or where the prototype existed. In the following study, the presence of
Scandinavian groups in the Baltic Sea region and Russian area provides an example
of the interaction that existed. As will be shown, some Scandinavians were involved
in activities connected to trade. Since trust is vital for trade being conducted, this
article also addresses how trust was created through material culture and how
material culture could facilitate interaction between trading actors.

the scandinavian presence in the central baltic sea region and

along the russian rivers during the viking age
Scandinavian objects have been found within present-day Russia, especially
along the river systems that create the Baltic Sea–Volga and Dnieper routes leading to
the Caspian Sea and the Black Sea. Both routes have their starting point at Staraja
Ladoga. Here the population consisted of many ethnic groups from the very start of
the town. Scandinavian artefacts are present in the oldest layers, which are dated to
the ad 750s. Furthermore, one cemetery in the Ladoga area, Plakun, situated across
from Staraja Ladoga, is regarded as Scandinavian and dates to the 9th and 10th centu-
ries (Callmer 2000, 34ff; Jansson 1987, 784; 1997, 27ff; Nosov 1998a; 1998b, 62ff).
MAM 37: 03-gustin – Press

26 maritime societies of the viking and medieval world

Based on the archaeological material, it seems that Scandinavian families were

present, even living permanently in many places within the territory of present-day
Russia (for further reading, see Callmer 2000 and Jansson 1997). Typical Scandi-
navian objects have, for example, been found in Rjurikovo Gorodišče, Novgorod’s
predecessor from the second half of the 9th century, and Gnëzdovo along the Dnieper
from the 10th century (Figure 3.1). The places can be interpreted as exchange and
craft-production centres with a mixed population, and they were situated along the
important communication paths (Callmer 2000, 37; Jansson 1997, 27ff; Nosov 1998a,
65; 1998b, 73).
Scandinavian merchants seeking Islamic silver started to travel south, probably
along the rivers Volga, Don and Donets, during the end of the 8th century (Noonan
1986, 321ff; 2001, 145ff). To gain access to the dirhams they needed commodities that
were attractive to the Caliphate. One of the most important was fur. There was
plentiful access to fur in eastern Europe, and this product could be acquired en route
through, for example, trade or tribute. Thus, it was not necessary to transport the fur
all the way from Scandinavia to the Khazarian markets and the markets along the

Figure 3.1 Places mentioned in the text (I Gustin and Vicki Herring after Jansson 2000, 110)
MAM 37: 03-gustin – Press

trade and trust in the baltic sea area 27

Volga. The Scandinavians encountered the long-established Finno-Ugric people in

the east, as well as Slavic populations as they travelled. In order to establish them-
selves in the fur trade the Scandinavians needed to become a part of the existing local
communities, and this process of integration may be witnessed in the burial evidence.
For example, this is demonstrated through graves in the cemetery at Gnëzdovo along
the Dnieper, and in the cemeteries at Timerëvo, Petrovskoje and Mikhailovskoje,
which are all in the area around Jaroslav, ie the upper Volga area (Callmer 2000,
34–35; Nosov 1994, 187; 1998b, 61ff).
The reasons for the Scandinavian expeditions and presence in eastern Europe
were not only to trade, take tribute and plunder. The associated material culture is
diverse and suggests that this population was also involved in activities connected to
military interests and craft production. The Russian archaeologist Evgenij N. Nosov
argues that some Scandinavians came to create, or participate in, the top level of the
social elite, an interpretation that draws on both archaeological material and the later
chronicle evidence (Nosov 1998b, 73, 80).
There are also signs of other reasons for the migration of some Scandinavians to
the Russian areas. The absolute largest group of identifiable Scandinavian artefacts
within present-day Russia is female jewellery, oval brooches in particular (Figure 3.2).
The jewellery indicates a relatively large presence of Scandinavian women. The
archaeologist Ingmar Jansson argues that the large number of Scandinavian women
cannot only be connected to the assumed few groups of warriors and traders that
settled in European Russia. The female presence is taken as evidence that Scandi-
navian families with an agrarian background also emigrated eastwards in order to
farm (Jansson 1987, 790; 1997, 26ff, 55).
While the archaeological material from present-day Russia provides evidence for
extensive and diverse contacts between the Scandinavians and the indigenous popu-
lation, there is less material from Finland and the Baltic States. If this paucity of
evidence does not indicate fewer contacts, then at least it implies considerably fewer
Scandinavian immigrants.
However, similarities in weaponry and male costume in Finland, the Baltic
States and Gotland show that contacts between the areas were vigorous. Concerning
the penannular brooches, belt fittings and belt buckles that have been found within
the three areas it has even been viewed as difficult to determine if the types origi-
nated from Gotland, Finland or the Baltic States (Lehtosalo-Hilander 1983, 291ff;
Thunmark-Nylén 1983, 310–311; Tõnisson 1962). This is regarded as the case for
male dress items, whereas the female costume in the region reveals a number of local
traits.
The role of Gotland in the Viking Age connections in the Baltic Sea area is often
discussed, but hardly explained. The inhabitants of Gotland differ from their neigh-
bours in one important aspect concerning material culture: they rejected foreign
objects. This feature of the island’s special character has been viewed as a conscious
strategy on the part of the Gotlanders to cope with the intensive interaction with
people from Finland and the Baltic States (Callmer 1992, 104). Interestingly, finds of
burials of Gotlandic type outside the island are rare.1 The extreme number of silver
MAM 37: 03-gustin – Press

28 maritime societies of the viking and medieval world

Figure 3.2 Distribution of bronze oval brooches from the mid-Viking Age
(after Jansson 1987, 777)

hoards — c700 — that have been found on Gotland does, however, testify to the fact
that the island had wide-ranging contacts (see Chapter 5). The number of 10th-
century Islamic coins in these hoards clearly exceeds the number found in mainland
Sweden or in the other countries around the Baltic Sea (Noonan 1994, 223, table 1;
Östergren 2008). At the end of the 10th century the number of western European
coins, especially German and English coins, increased and after ad 990 they made up
the majority of the coinage. In fact, more German and English Viking Age coins have
been found on Gotland than in their home countries (Östergren 1983, 34).
What activities have the Gotlandic people been involved in outside their own
island? The Gotlanders, in their capacity as travelling farmers, have often been seen
as mediators of different merchandise in the Baltic Sea region during the Viking Age.
In recent years it has become possible to reconstruct a map demonstrating Gotland’s
role as a hub for trade by charting the large number of harbours from the Late Iron
Age along the coast of Gotland. The harbours vary in size, ranging from small fishing
camps for individual farms to extensive harbours and places for exchange, some
MAM 37: 03-gustin – Press

trade and trust in the baltic sea area 29

appearing to be almost permanent in character. Examples of the latter include

Paviken, Bandlundeviken, Bogeviken and Fröjel.2

trade and trust

As has been shown, the trade and exchange that occurred in the Baltic Sea area
and along the Russian rivers resulted in the meeting and interaction of people from
widely different socio-political groups. However, as I pointed out in my thesis,
Mellan gåva och marknad, trade was hardly an unproblematic phenomenon in tradi-
tional societies (Gustin 2004). The evidence of the Icelandic sagas and of anthro-
pological examples shows that buying and selling was perceived as a highly delicate
and even hostile undertaking. Buying and selling was something you did with people
who were geographically or socially distant. Ordinary ties of friendship were not
supposed to be subject to the strain that continuous commercial transactions entailed
(Miller 1986, 46–49; Sahlins 1972, 197ff; see Chapter 12 and references therein).
Why was trade viewed as being such a delicate and risky activity? It is part of the
essence of market trade that the parties each try to maximize their gain from the
transaction. Maximization must inevitably occur at the expense of the other party.
Moreover, long-distance trade was often carried on with persons outside one’s own
group — the main reason for engaging in trade was to obtain products which could
not be found within one’s own area — and trade between strangers was liable to
involve bargaining and deception that did not occur in exchanges within one’s own
group. An ethnographic example is seen in the work of the Norwegian anthro-
pologist, Fredrik Barth. He studied peoples’ interactions and meetings, and identified
differences depending on whether a meeting occurred between people of the same
ethnic group or two separate ethnic groups. At the meetings between people belong-
ing to the same group, both parties assume that they share the same fundamental
values. One acts according to the same rules and therefore there exists the potential
for common actions and activities. On the other hand, when individuals identify as
belonging to ethnic groups other than one’s own this implies that there are limitations
in understanding between the partners. According to Barth, the meeting with the
stranger therefore creates insecurity about the other’s view of ‘right’ opposed to
‘wrong’, and ‘honourable’ opposed to ‘immoral’ (Barth 1969, 13ff).
Anthropological and historical examples show that market trade was not only
viewed as a dangerous activity, but even sometimes as an amoral phenomenon. There
are many examples of business morality being far from honest; it could be associated
with gain-oriented individualism and undue profit. Incidentally, merchants in ancient
Greece and medieval Europe had the same patron deity or saint as thieves; Hermes
and Saint Nicholas. The negative view of merchants seems primarily to have applied
to those working locally. For those engaged in long-distance trade, increments to
cover the cost of storage, transport, and the like were not considered to be undue gain
in the same way as in local trade.
Some archaeological evidence also shows the link between trade and undue
profit in the form of cheating. False coins have been found in assemblages from the
MAM 37: 03-gustin – Press

30 maritime societies of the viking and medieval world

Viking Age. In material from both Birka and Hedeby there are counterfeit dirhams in
which the majority of the silver is replaced with other metals (Rispling 2004; Steuer
et al 2002, 158). In Birka a counterfeit western gold coin has also been found (Rispling
2004, 39–42). Other places have examples of counterfeit coins or counterfeit means of
payments from the Viking Age (see Gustin 2004, 173).
In contrast to the image of trade as an uncertain undertaking, steered by self-
interest and fraught with conflict, there stands the need for trust. The significance of
trust and security for the functioning of trade has been emphasized by anthropol-
ogists, sociologists and historians alike (eg Bourdieu 1995, 158; Dahl 1998, 291;
Giddens 1984, xxiii; Sahlins 1972, 201, 298ff). The administered exchange of com-
modities, which has been considered to be the most important exchange system at the
Viking Age ‘port of trade’ by Karl Polanyi and the substantivists, can be viewed as
one of several ways to avoid some of the risks involved in outright market trade. In
the administered exchange of commodities, contracts were drawn up between dif-
ferent elites and state powers, which were also the parties that organized the
exchange. This diplomatic and administrative procedure aimed, among other things,
to avoid competition and haggling. Administered trade, if anything, confirmed the
links between different elites.
Other examples show that the individual could achieve security in trade by
establishing some kind of trade partnership. In this way, some of the risks otherwise
inherent in market trade could be eliminated. Buyer and seller were linked by a social
relationship. The market part of trade was toned down, since it was already deter-
mined who you could exchange goods with. In transactions between trading
partners, moreover, bargaining and haggling were prohibited. The transaction was
instead carried out as a transfer of things of equal value, and the exchange thus also
served to consolidate the peace between the parties (Dilley 1996, 730ff; Plattner 1989,
209ff; Sahlins 1972, 201, 298ff).
Collective and large-scale ways of achieving security involved the establishment
of different types of institutions around trade. Trust could be created, for example,
through the occurrence of physical and legal protection for people pursuing trade. At
a macro level this is obvious in the ports of trade or emporia founded around the
North Sea and the Baltic Sea in the Late Iron Age. These trading sites offered physical
protection for the people there, thanks to the presence of armed guards and through
the occurrence of palisades, defensive ramparts, and the like. However, the places
gave not just physical but also legal protection. Strangers, who were otherwise
normally without any legal protection, were safe here.
Another measure for creating trust was to surround trade with routines. We
know from written sources that a purchase had to be concluded in the presence of
several witnesses or some potentate. Medieval legal texts also show that a deal had to
end with a handshake. The medieval laws of Swedish and Danish towns mention that
the parties then had to take a drink together. The fact that trade was hedged with
routines as early as the Viking Age is shown by the Bjärköarätt, a collection of laws
on trade and its protection, which is mentioned in Norwegian sources at the start of
the 11th century (Müller-Boysen 1990, 118ff).
MAM 37: 03-gustin – Press

trade and trust in the baltic sea area 31

trust and weights

Trust in the equipment used for weighing and measuring was also of great
significance if trade was to be able to take place without friction. The importance
attached to accurate and reliable weights is evident in several early medieval decrees.
In 744 Charlemagne’s predecessor Pippin urged the bishops to ensure that each civitas
would have a legally valid market and legally valid measures (Steuer 1999, 406).
Another decree issued by Charlemagne at the end of the 8th century ruled that each
town and monastery should provide exact and identical weights (Kisch 1965, 4).
In Scandinavia, the Baltic Sea area and along the Russian rivers, standardized
scales and two new standardized weights occurred in the second half of the 9th
century (Figure 3.3). Within the area that today consists of Sweden and the eastern
Baltic Sea region there was no minting of coins during the 9th century and most of the
10th century. Instead, there existed a metal-weight economy, meaning that both
smaller and larger payments could be made in silver that was weighed out. It is from
this that one should see the importance of the tools of payment, specifically
standardized weights and scales, which appeared in the region. This standardized
equipment, used primarily with high-quality silver in the form of dirhams, resulted in
a single ‘economic language’ across a vast geographical area (Figure 3.4).
The standardized scales were used with spherical weights with flattened poles as
well as cubo-octahedral weights. Both weight types are connected with the Islamic
weight system and have been discussed in detail elsewhere (Gustin 1997; 1999; 2004;
Steuer 1987; 1997). What is important in this context is that both types differ from
other contemporary weights in their particular shape and decoration. Were the cubo-
octahedral weights and the flattened spherical weights that occurred in the Baltic Sea
area in the second half of the 9th century perceived as exact, identical and reliable by
the people of the time? There is a great deal to suggest that this was the case. First of
all, the cubo-octahedral weights and the flattened spherical weights were more clearly
and more systematically an integral part of a standard weight system — the Islamic
one — than other contemporary weights. Secondly, the cubo-octahedral weights and
the flattened spherical weights were made of bronze or bronze-coated iron. The
majority of the weights otherwise found in Viking Age contexts are made of lead.
Lead is suitable for weights in that it has a high specific gravity, but when it comes
to security and permanence, lead suffers a serious disadvantage. The material is soft
and it oxidizes easily. Bronze, on the other hand, is a relatively hard material. That is
why bronze is often used for the manufacture of weights, for example, in ancient
Greece where the reference weights stored in the temples were usually made of
bronze.
Thirdly, it has been claimed that not just the use of unchangeable material, but
also the design of the weights, is of great importance for establishing reliability.
People quite simply got used to certain traditional shapes, which they also accepted
and trusted. Special trust was placed in shapes that were difficult to tamper with.
These shapes have been divided into two main groups. The first consists of regular
geometrical forms while the second consists of various special shapes, above all
figurative representations of animals (Kisch 1965, 92ff). The cubo-octahedral weights
MAM 37: 03-gustin – Press

32 maritime societies of the viking and medieval world

Figure 3.3 Cubo-

octahedral weights,
spherical weights with
flattened poles and scales
with weight-like
ornaments (drawing by
C F Lindberg in
Montelius 1872, 178)

must be regarded as a typical example of the first group, and the distinctive shape
must have helped to make counterfeiting difficult, and hence created confidence.
The ornamentation on the cubo-octahedral weights and the flattened spherical
weights must also have made it more difficult to modify the weights, just as the
ornament on Viking Age coins rendered it difficult to make imitations (see Hodges
1988, 102–103 on the role of ornamentation in early minting). The decorations dis-
tinguish the cubo-octahedral and spherical weight types from other contemporary
weights, which mostly had no decoration at all.
MAM 37: 03-gustin – Press

trade and trust in the baltic sea area 33

Figure 3.4 Distribution of cubo-octahedral weights (after Steuer et al 2002, Taf 139)

The adaptation to a supra-regional weight system and the material, shape, and
ornamentation thus indicate that the cubo-octahedral weights were perceived as
exact and reliable. We have seen that commodity exchange, especially commercial
transactions, must have been regarded as a risky undertaking in bygone societies.
Security and trust are fundamental for the possibility of engaging in transactions and
for the development of market trade. The cubo-octahedral weights, associated with
reliability, weighing, payment, and by extension also with activities such as trade, are
therefore the probable explanation why weight-like ornamentation began to appear
on weighing and measuring equipment such as scales, steelyards and measuring rods
from the second half of the 9th century. During the Viking Age, there developed not
only regulations and rules aimed at bringing the double-edged trade into a more
secure and trustworthy form; the articles of material culture used in commodity
exchanges and commercial transactions were also designed to evoke connotations
of exactitude and reliability.
MAM 37: 03-gustin – Press

34 maritime societies of the viking and medieval world

trust and penannular brooches with facetted end knobs

The trade and exchange that occurred in the Baltic Sea area and along the
Russian rivers resulted in the meeting and interaction of people from widely different
socio-political groups. Social psychologists and sociologists have often discussed
encounters between strangers and how to overcome the general insecurity that is felt
in meetings with strangers. The importance of self-representation and how this
makes interactions with others easier has been pointed out. The sociologist Erving
Goffman argues that people, when they meet, create or show the image of themselves
that they want the other to see. The information that is communicated is the foun-
dation for the definition of the situation and makes it possible for both partners to
anticipate what they can expect of each other and the meeting (Goffman 1959).
The importance of being able to read unknown persons and the character of the
meeting has also been discussed by Richard Jenkins. He has, inspired by Goffman,
identified the importance of predictability when it comes to setting the tone for
encounters between strangers. Through physical cues each person signals their group
belonging, laying the foundation for the desired treatment. One way to achieve pre-
dictability about unknown people is to associate an attribute with them — to place
them in stereotypes and groups, based on the limited information that their verbal
and non-verbal behaviour reveals. Through becoming proficient in recognizing
simple motifs and engaging in this kind of categorization, an understanding of the
unknown is achieved, guided by clues from earlier experiences of a group of people.
In this way both partners can predict what kind of meeting is afoot and can adjust
their behaviour accordingly (Jenkins 1996, 123–124).
During the Late Iron Age it is likely that there was an increase in both positive
and negative interaction within the Baltic Sea. The region had no common language,
and interaction and dealings between peoples must have been highly unpredictable.
Therefore self-representation must have been of utmost importance at meetings
between strangers and members of other socio-political groups.
When Goffman discusses self-representations, the focus is not only on presenta-
tions in verbal communication and communication in the form of facial expressions
and body language. Outward signs and material culture also have meaning and are
bearers of information (Goffman 1959, 222). Also within contextual archaeology it
has been stressed that objects have a communicative dimension and that material
culture conveys information. However, it has been considered difficult to be unam-
biguous in assigning meaning to material culture. Only when it comes to iconological
representations it has been pointed out that it is possible to interpret what is signified
because of the close connection between the signifier and the signified (Hodder 1987,
3; 1989, 72f). Therefore it is interesting to observe that the facetted end knobs on the
penannular brooches seem to reference the cubo-octahedral weights during the latter
part of the 9th century/first part of the 10th century.3 If one takes a close look at the
facetted knobs of the penannular brooches from this period and onwards one can see
that the design corresponds with the weights in that they have the same type of facet-
ing, the same marking of the eye and often the same type of dotted frame (Figure 3.5).
This ornamentation is especially common on Gotland and in Finland.4
MAM 37: 03-gustin – Press

trade and trust in the baltic sea area 35

Figure 3.5 Penannular

brooch with facetted and
weight-like end knobs.
The brooch comes from
Lilla Klintagård,
Väskinde parish,
Gotland. Scale 1:1 (after
Stenberger 1947, Abb
200)

Traces from the manufacture of penannular brooches with facetted end knobs
are known from a number of places. The earliest brooches of this type have been
found within the Finnish area, which indicates that they were made there. Later, it
seems likely that they were manufactured on Gotland and in the present-day Baltic
States (Ģinters 1984, 29–30).
Penannular brooches with facetted end knobs become more common in the
north Baltic Sea area during the latter half of the 9th century, at the same time or
slightly after the cubo-octahedral weights started to appear in this area.5 Is there
anything that supports the idea that facetted end knobs on the penannular brooches
and the cubo-octahedral weights were really associated with each other during the
9th, 10th and 11th centuries? There are a number of cases that indicate this. In burials
from Sweden and the Baltic and Russian areas cut-off facetted end knobs have been
found with different kinds of weights (for example, Ģinters 1984, 30 and references
cited therein; Nosov 1987; Thålin 1984, 15). This shows that the facetted end knobs
on penannular brooches were associated with, and actually used as, weights. As dis-
cussed above, the cubo-octahedral weights probably inspired confidence in com-
mercial transactions. It is therefore reasonable to assume that the facetted end knobs
had similar connotations.
When individuals from different ethnic groups met in the Baltic Sea region, the
Sami lands further north and along the Russian rivers, it must have been of great
importance to present an image of oneself that would give the counterpart informa-
tion about what kind of meeting to expect and what kind of situation he/she would
have to respond to. According to social physiologists these kinds of presentations
generally occur at all meetings. To be able to identify strangers based on the material
culture that adorns them would have been of great importance at meetings between
people from different ethnic and language groups in a time when a common language
did not exist. For example, Færeyinga saga describes how merchants were actually
identified in this way. In the saga it is told how, from a long distance, one could
MAM 37: 03-gustin – Press

36 maritime societies of the viking and medieval world

identify a group of strangers on the island as merchants based on the weapons and
clothing they wore (Müller-Boysen 1990, 45).
It must have been particularly important to point out that one wished for peace-
ful exchange during a time when this form of transaction may have been far from the
norm. In the written evidence from the period there are a number of examples show-
ing how violence could transpire if one of the trading partners viewed the situation in
the wrong light. For example, Æthelweard describes how some Scandinavian ships
visited Wessex at the end of the 8th century. Expecting merchants, the king’s official
and his men quickly gathered, only to be killed as the seafarers did not have peaceful
intentions (Sawyer 1962, 17–18). In laws concerning trade from the early Middle
Ages there are also examples of how important it was that both partners signalled
peace at an early stage. Within the Scandinavian area the intent for peaceful com-
merce was generally signalled by the approaching partner raising a white shield. If a
person was arriving the shield was raised above the head. In the case of a ship, the
shield would be raised on the mast (Müller-Boysen 1990, 87ff).
The Nestor or Primary Chronicle recounts events attributed to the years
between the 9th century and the beginning of the 12th century, providing examples of
how material culture was used as a symbol in trade. Of the year ad 945 it is told that
the Rus’ leader provided messengers with a gold seal, and merchants with a silver
seal, to show that they came with peaceful errands when they went to Greece. It
further explains that these seals, with time, came to be replaced by letters. Merchants
without a letter were imprisoned, or killed if they resisted (Oxenstierna 1998, 47). It is
stretching the evidence to say that the penannular brooches with facetted end knobs
had the same official status as a Rus’ leader’s seal when it came to displaying peaceful
intentions. Nevertheless, it is conceivable that by wearing jewellery incorporating the
form of the cubo-octahedral weights in an area where the weights were well known,
the wearer could signal trustworthiness and knowledge of payment transactions
using weighed silver. Thus one could anticipate accepted behaviour from the other
when it came to the exchange of goods and payment. In this way stress and risk
would have been reduced, and transactions facilitated, thanks to a recognized
symbol.

concluding remarks
Judging by the archaeological and historical sources, contacts between different
areas around the Baltic became increasingly vigorous in the course of the Iron Age.
The fact that trade was a part of the general interaction in the Baltic Sea area and
along the Russian rivers during the Viking Age is obvious not least from the tools of
trade, in the form of standardized weights and scales that quickly began to spread in
the region from the second half of the 9th century. The nature of trade was such that
it could involve great risks for the partners involved; transactions could be experi-
enced as hostile and dangerous, with much potential for conflict. Insecurity was also
a part of all market trade, from the risk of drawing the shortest straw in bargaining
to being tricked or robbed and suffering physical violence. Therefore the spherical
weights with flattened poles as well as the cubo-octahedral weights were designed to
MAM 37: 03-gustin – Press

trade and trust in the baltic sea area 37

inspire confidence and thereby to facilitate commercial transactions and trade. This
might also have been the reason why cubo-octahedral and weight-like ornamentation
began to appear on weighing and measuring equipment such as scales, steelyards and
measuring rods from the second half of the 9th century. Other objects that are likely
to have had the same connotations as the cubo-octahedral weights are penannular
brooches with facetted end knobs. Finds of cut off facetted end knobs in graves show
that they could be used as weights. It is therefore reasonable to assume that the
penannular brooches could have inspired confidence and facilitated trade in the same
way as the weights.

notes
1 According to Ingmar Jansson, only one burial 4 In my PhD dissertation I compiled a corpus of
with typical Gotlandic jewellery has been found penannular brooches with facetted end knobs
in the Baltic States. This grave contains two which have been found in different countries
animal-shaped brooches and one box-shaped and regions (Gustin 2004, 215). The corpus
brooch. The burial is from the cemetery Lauks- shows that penannular brooches with facetted
kola at the lower Dvina (Jansson 1994, 16 and end knobs have been found primarily on Gotland.
references therein). Individual burials at the Nevertheless, a large number of brooches have
cemeteries at Grobiņa, Elbla˛g and Višnevo con- also been found within Finnish territory and in
tain Gotlandic jewellery, but these are typically the Baltic States. In Estonia this type of pen-
combined with local jewellery or show a mix- annular brooch was generally in use during the
ture of objects from the Baltic States, mainland first part of the 11th century (Tallgren 1925, 68).
Sweden and Gotland. Judging from finds of A large number of brooches of this type have been
imported dress items, immigration to Gotland found in Lithuania (Žulkus 1997), and in Latvia
does not seem to have occurred from any other they are viewed as one of the most common types
area either. Jewellery originating in the Baltic of jewellery from the Late Iron Age (Balodis
States has been found on Gotland, but it has only 1940, 220). Penannular brooches with facetted
occasionally been possible to associate the mate- end knobs have also been found in present-day
rial with specific graves (Thunmark-Nylén 1983, Russia, from Ladoga and Gnëzdovo in the west to
306ff). the Jaroslav and Vladimir cemeteries in the east
2 Whether or not Visby should be counted as a (ie the area in which Scandinavians lived). In
Viking Age harbour is discussed elsewhere (eg Sweden, penannular brooches with facetted end
Westholm 1989; Roslund 2001). Excavation has knobs occur relatively frequently at Birka. Only
occurred at some of the sites. Recent excavations a few examples are known from other parts
at Fröjel show that it has been used continually of mainland Sweden. This type of brooch has
from the 6th or 7th century to the early Middle seldom been found in other Scandinavian
Ages. The harbour went out of use around countries except in the Sami area in northern
the middle of the 12th century (Carlsson 1999). Norway (Serning 1956; Sjøvold 1974).
The investigations at Paviken and Bandlunde 5 Penannular brooches with facetted end knobs
demonstrate that these harbours came into use from the first half of the 9th century are not
during the Vendel Period and early Viking Age re- common in the Scandinavian material. On Got-
spectively. Both places are thought to have ceased land all types of penannular brooches are rare
earlier than Fröjel: Paviken around the year during the 9th century (Carlsson 1988, 69). The
ad 1000 and Bandlunde around the end of the 9th same pattern holds for the penannular brooches
century (Brandt 1986, 52; Lundström 1983, 117). that have been found in Birka. With one possible
3 I have in my PhD dissertation given a detailed exception they all belong to the late Birka period,
description of the earliest penannular brooches ie from the mid-9th century to ad 970 (Thålin
with facetted end knobs (Gustin 2004, 300ff and 1984, 19). Interestingly, the number of penannu-
references therein). They are few in number and lar brooches increases dramatically towards the
come from Finland during the transition between end of the Viking period. There are 110 com-
the Vendel period/Viking Age. All the earliest plete penannular brooches from Birka, of these
Finnish penannular brooches that are described 24, ie 22%, have facetted end knobs (Thålin
or depicted, either unornamented or ornamented, 1984, 19ff). The Gotlandic material presents a
lack the traits that are known from the cubo- similar picture. Anders Carlsson has discussed
octahedral weights. the ones with facetted end knobs in his study of
MAM 37: 03-gustin – Press

38 maritime societies of the viking and medieval world

penannular brooches (Carlsson 1988, 20, 70ff). type appears to decrease. From ad 1100–1150 the
Based on Carlsson’s numbers one can see that penannular brooch with facetted end knobs was
during the period ad 800–1000, 40% of all pen- the only type that existed (Carlsson 1988, 73ff).
annular brooches had facetted end knobs. Simul- Gotland is not the only place that shows this
taneously the number of terminal types increased trend of an increasing number of penannular
in relation to the prevailing period. From the brooches with facetted end knobs. Concerning
period ad 1000–1100 a large number of penan- Finland, Salmo (1956, 35) points out that a large
nular brooches, over 500, have been found on number of penannular brooches with facetted end
Gotland. Carlsson has been able to separate older knobs can be dated to the end of the 11th century.
material within this last phase, and in this phase 6 I am grateful to Mats Roslund for comments
there are 184 brooches with facetted knobs. These on the manuscript and for the support of Birgit
comprise 54% of all the penannular brooches in och Gad Rausings Stiftelse för Humanistisk
this phase, despite the numbers of knob types still Forskning and Berit Wallenbergs Stiftelse.
increasing. After this, however, the facetted knob

references
Balodis, F, 1940 Det äldsta Lettland, Almqvist & Wiksell, Uppsala
Barth, F, 1969 ‘Introduction’, in F Barth (ed), Ethnic Groups and Boundaries, Scandinavian
University Books, Oslo, 9–38
Bourdieu, P, 1995 Praktiskt förnuft: Bidrag till en handlingsteori, Daidalos, Göteborg
Brandt, B, 1986 ‘Bandlundeviken — en vikingatida handelsplats på Gotland: grävningsrapport och
utvärdering’, Uppsats för påbyggnadskurs i arkeologi vt 1986 (unpublished), Institutionen för
arkeologi, Stockholms Universitet, Stockholm
Burström, M, 1991 Arkeologisk samhällsavgränsning: en studie av vikingatida samhällsterritorier i
Smålands inland, Stockholms Universitet, Stockholm
Callmer, J, 1991 ‘Territory and dominion in the Later Iron Age in southern Scandinavia’, in
K Jennbert and M Strömberg (ed), Regions and Reflections: In Honour of Märta Strömberg,
Almqvist & Wiksell International, Lund, 257–273
Callmer, J, 1992 ‘Interaction between ethnical groups in the Baltic region in the Late Iron Age’, in B
Hårdh and B Werbart (ed), Contacts Across the Baltic Sea, Lunds Universitet, Lund, 99–107
Callmer, J, 2000 ‘The archaeology of the early Rus’ c. ad 500–900’, Medieval Scandinavia 13, 7–63
Carlsson, A, 1988 Vikingatida ringspännen från Gotland: text och katalog, Stockholms Univer-
sitet, Stockholm
Carlsson, D, 1999 ‘Ridanäs’: Vikingahamnen i Fröjel, ArkeoDok Skrifter nr 2, Arkeodok, Visby
Dahl, G, 1998 Trade, Trust, and Networks: Commercial Culture in Late Medieval Italy, Nordic
Academic Press, Lund
Dilley, R, 1996 ‘Market’, in D Levison and M Ember (ed), Encyclopedia of Cultural Anthropology,
Volume 3, Henry Holt and Co, New York, 728–732
Giddens, A, 1984 The Constitution of Society: Outline of the Theory of Structuration, Polity Press,
Cambridge
Ģinters, V, 1984 ‘Der Ursprung der Ringspannen von östlichen Typ’, in G Arwidsson (ed), Birka
II:1: Systematische Analysen der Gräberfunde, Kungliga Vitterhets, Historie och Anti-
kvitetsakademien, Almqvist & Wiksell International, Stockholm, 23–30
Goffman, E, 1959 Presentation of Self in Everyday Life, Doubleday, New York
Gustin, I, 1997 ‘Islam, merchants or king? Who was behind the manufacture of Viking Age
weights?’, in H Andersson, P Carelli and L Ersgård (ed), Visions of the Past: Trends and
Traditions in Swedish Medieval Archaeology, Almqvist & Wiksell International, Lund/
Stockholm, 163–177
Gustin, I, 1999 ‘Vikter och varuutbyte i Uppåkra’, in B Hårdh (ed), Fynden i centrum: keramik,
glas och metall från Uppåkra, Almqvist & Wiksell International, Stockholm, 243–269
MAM 37: 03-gustin – Press

trade and trust in the baltic sea area 39

Gustin, I, 2004 Mellan gåva och marknad: Handel, tillit och materiell kultur under vikingatid,
Almqvist & Wiksell International, Stockholm
Hodder, I, 1987 ‘The contextual analysis of symbolic meanings’, in I Hodder (ed) The Archaeology
of Contextual Meanings, Cambridge University Press, Cambridge, 1–10
Hodder, I, 1989 ‘Post-modernism, post-structuralism and post-processual archaeology’, in
I Hodder (ed), The Meaning of Things: Material Culture and Symbolic Expression, Unwin
Hyman, London, 64–78
Hodges, R, 1988 Primitive and Peasant Markets, Basil Blackwell, Oxford
Jansson, I, 1987 ‘Communications between Scandinavia and eastern Europe in the Viking Age: the
archaeological evidence’, in K Düwel, H Jahnkuhn, H Siems and D Timpe (ed), Unter-
suchungen zu Handel und Verkehr der vor- und frühgeschichtlichen Zeit in Mittel- und Nord-
europa: Teil IV: Der Handel der Karolinger- und Wikingerzeit, Vandenhoeck & Ruprecht,
Göttingen, 773–807
Jansson, I, 1994 ‘Skandinavien, Baltikum under Rus’ under vikingatiden’, in K Tønnesson (ed),
Norden og Baltikum: Rapport I, Olso universitet, IKS, avdelningen for historie, Oslo, 5–25
Jansson, I, 1997 ‘Warfare, trade or colonization? Some general remarks on the eastern expansion
of the Scandinavians in the Viking period’, in P Hansson (ed), The Rural Viking in Russia and
Sweden, Örebro kommuns bildningsförvaltning, Örebro, 9–64
Jansson, I, 2000 ‘Östersjöländerna och vikingatiden’, in K Dahlbäck (ed), Att förstå det mänskliga:
Humanistisk forskning vid Stockholms universitet, Natur och kultur, Stockholm, 109–137
Jenkins, R, 1996 Social Identity, Routledge, London
Kisch, B, 1965 Scales and Weights: A Historical Outline, Yale University Press, New Haven
Lehtosalo-Hilander, P L, 1983 ‘Gutarnas förbindelser med Finland’, in I Jansson (ed), Gutar och
vikingar, Statens Historiska Museum, Stockholm, 288–305
Lundström, P, 1983 ‘Paviken: handelshamn, hantverksplats och skeppsvarv’, in I Jansson (ed),
Gutar och vikingar, Statens Historiska Museum, Stockholm, 117–119
Miller, I, 1986 ‘Exchange in Medieval Iceland’, Speculum 61, 19–50
Montelius, O, 1872 Svenska fornsaker ordnade och förklarade, Norstedt, Stockholm
Müller-Boysen, C, 1990 Kaufmannsschutz und Handelsrecht in frühmittelalterlichen Europa,
Wachholtz, Neumünster
Noonan, T S, 1986 ‘Why the Vikings First Came to Russia’, Jahrbücher für Geschichte Osteuropas
34, 321–348
Noonan, T S, 1994 ‘The Vikings in the east: coins and commerce’, in B Ambrosiani and H Clarke
(ed), The Twelfth Viking Congress: Developments around the Baltic and the North Sea in the
Viking Age, Birka Project, Riksantikvarieämbetet och Statens Historiska Museer, Stockholm,
215–236
Noonan, T S, 2001 ‘Volga Bulghāria’s Tenth-Century Trade with Sāmānid Central Asia’,
Archivum Eurasie medii aevi 11, 140–218
Nosov, E N, 1987 ‘New Data on the Ryurik Goroditshche near Novgorod’ Fennoscandia
Archaeologica iv, 73–85
Nosov, E N, 1994 ‘The emergence and development of Russian towns: some outline ideas’,
Archeologia Polona 32, 185–196
Nosov, E N, 1998a ‘The Varangian problem: Russian historical tradition vs. new archaeological
evidence’, in A Wesse (ed), Studien zur Archäologie des Ostseeraumes: von der Eisenzeit zum
Mittelalter. Festschrift für Michael Müller-Wille, Wachholtz, Neumünster, 61–66
Nosov, E N, 1998b ‘The first Scandinavians in northern Rus’’, in A Hedman and A Kirpichnikov
(ed), Vikingi i Slavjane, Dmitrij Bulanin, St Petersburg, 56–82
Östergren, M, 1983 ‘Silverskatternas fyndplatser — farmännens gårdar’, in I Jansson (ed), Gutar
och vikingar, Statens Historiska Museum, Stockholm, 34–48
MAM 37: 03-gustin – Press

40 maritime societies of the viking and medieval world

Östergren, M, 2008 ‘Spillings: världens största vikingatida silverskatt’, in A M Pettersson (ed),

Spillingsskatten: Gotland i vikingatidens världshandel, Länsmuseet på Gotland, Visby, 11–40
Oxenstierna, G, 1998 Nestorskrönikan, B. Östlings bokförlag Symposion, Eslöv
Plattner, S, 1989 ‘Economic Behaviour in Markets’, in S Plattner (ed), Economic Anthropology,
Stanford University Press, Stanford, 209–221
Rispling, G, 2004 ‘Catalogue and Comments on the Islamic Coins from the 1990–1995 Excava-
tions’, in B Ambrosiani (ed), Eastern Connections II, Birka Project/ Riksantikvarieämbetet,
Stockholm, 26–60
Roslund, M, 2001 ‘Gutar, främlingar och den förblindande vikingatiden: om staden Visbys tidi-
gaste datering’, in A Andrén, L Ersgård and J Wienberg (ed), Från stad till land: en medeltids-
arkeologisk resa tillägnad Hans Andersson, Almqvist & Wiksell International, Stockholm,
241–251
Sahlins, M, 1972 Stone Age Economics, Aldine-Atherton, Chicago
Salmo, H, 1956 ‘Finnische Hufeisenfibeln’, SMYA – FFT 56, 7–106
Sawyer, P H, 1962 The Age of the Vikings, Edward Arnold, London
Serning, I, 1956 Lapska offerplatsfynd från järnålder och medeltid, Stockholms högskola,
Stockholm
Sjøvold, T, 1974 The Iron Age Settlement of Arctic Norway, Volume II: Late Iron Age, Norwegian
Universities Press, Tromsö/Oslo
Stenberger, M, 1947 Die Schatzfunde Gotlands der Wikingerzeit: II: Fundbeschreibungen und
Tafeln, Kungliga Vitterhets, Historie och Antikvitetsakademien, Almquist & Wiksell Inter-
national, Stockholm
Steuer, H, 1987 ‘Gewichtsgeldwirtschaften im frühgeschichtlichen Europa. Feinwagen und
Gewichte als Quellen zur Währungsgeschichte’, in K Düwel, H Jahnkuhn, H Siems and
D Timpe (ed), Untersuchungen zu Handel und Verkehr der vor- und frühgeschichtlichen Zeit
in Mittel- und Nordeuropa. Teil IV: Der Handel der Karolinger- und Wikingerzeit, Vanden-
hoeck & Ruprecht, Göttingen, 405–527
Steuer, H, 1997 Waagen und Gewichte aus dem mittelalterlichen Schleswig: Funde des 11. bis 13.
Jahrhunderts aus Europa als Quellen zur Handels- und Währungsgeschichte, Rheinland-
Verlag, Köln/Bonn
Steuer, H, 1999 ‘Handel und Wirtschaft der Karolingerzeit’, in C Stiegemann and M Wemhoff (ed),
799: Kunst und Kultur der Karolingerzeit: Karl der Grosse und Papst Leo III. in Paderborn,
von Zabern, Mainz
Steuer, H, Stern, W and Goldenberg, G 2002 ‘Der Wechsel von der Münzgeld — zur Gewichts-
geldwirtschaft in Haithabu um 900 und die Herkunft des Münzsilbers im 9. und 10.
Jahrhundert’, in K Brandt, M Müller-Wille and C Radtke (ed), Haithabu und die frühe
Stadtentwicklung im nördlichen Europa, Wachholtz, Neumünster, 133–167
Tallgren, A M, 1925 Zur Archäologie Eestis: 2: Von 500 vis etwa 1250 n. Chr., Mattiesen, Dorpat
Thålin, H, 1984 ‘Ringspangen’, in G Arwidsson (ed), Birka II:1: Systematische Analysen der
Gräberfunde, Kungliga vitterhets, historie och antikvitetsakademien, Stockholm, 15–22
Thunmark-Nylén, L, 1983 ‘Gotland och Ostbaltikum’, in I Jansson (ed), Gutar och vikingar,
Statens Historiska Museum, Stockholm, 306–322
Tõnisson, E, 1962 ‘Eesti aardeleiud 9.–13. sajandist’, in H Moora (ed), Muistsed kalmed ja aarded.
Arheoloogiline kogumik 2, Eesti Riiklik Kirjastus, Tallinn, 182–249
Westholm, G, 1989 ‘Visby — Bönders hamn och handelsplats: Visbysamhällets uppkomst och
utbredning under förhistorisk tid och äldre medeltid’, in Medeltidsstaden 72. Visby. Staden
och omlandet, Riksantikvarieämbetet och Statens historiska museer, Stockholm, 49–114
Žulkus, V, 1997 ‘Baltische Funde an den westlichen Ostseeküsten’, in V Kazakevièius and V Žulkus
(ed), The Balts and their Neighbours in the Viking Age, Zara, Vilnius, 165–188
MAM 37: 04-magi – Press

chapter 4
BOUND FOR THE EASTERN BALTIC:
TRADE AND CENTRES ad 800 – 1200

By Marika Mä gi

The vast majority of writings dealing with Viking maritime activity concentrate on
movements in the western or southern direction, paying much less attention to the
east — which can be seen as a reflection of recent political realities. The present Baltic
States in particular have often been left out in discussions of Viking Age communi-
cations and sea routes. The Cold-War-era border running along the Baltic Sea created
a situation where Baltic archaeologists faced several constraints induced by the Soviet
system, primarily the lack of exact maps and restricted access to the sea. Society
gradually accepted the subconscious concept that sea-faring is a complicated venture,
with the sea hindering rather than favouring communication with neighbouring
areas. Scandinavia, on the other side of the Baltic Sea, and even Finland, which is in
places less than 100km from the northern coast of Estonia, seemed to be at an un-
reachable distance. This political situation of the recent past was erroneously pro-
jected back in history, creating an over-romantic vision of one-time brave seafarers
and pirates.
These public attitudes were echoed in archaeological research. The Iron Curtain
in a broader sense separated not only Baltic archaeologists from Western academic
thought and publications, but also resulted in very selective information about
eastern Baltic prehistoric societies entering academic discussion in the West. If the
eastern Baltic area was embraced at all, in most cases only Grobiņa was shown, as the
only Viking Age trading place on maps hitherto published in archaeological writings
— clearly as a result of the proper publication of excavations there, printed in
German.
Another important aspect when considering earlier research into maritime land-
scapes and ancient sea-faring in the present Baltic States was the scarcity of carto-
graphic material. Maps that Soviet-era archaeologists could use offered only very
general overviews and were often deliberately misleading. Exact maps were con-
cealed. This lacuna hindered the proper understanding of prehistoric landscapes,
particularly in Estonia where isostatic uplift has been up to 3m in a thousand years.
This phenomenon decreases considerably further south. Even in present Latvia the
differences of coastal outlines, when compared with those of late prehistory, are
not so dramatic. For Estonian archaeologists, however, it remained impossible to
MAM 37: 04-magi – Press

42 maritime societies of the viking and medieval world

evaluate the impact of relative sea-level change on maritime landscapes as recently as

the early 1990s — maps with contour lines belonged to special files. When topo-
graphic maps at the scale of 1:10,000 with contour lines at metre intervals (originally
meant for the use of Soviet military forces) became available for a larger audience,
research into landscape archaeology, especially of maritime landscapes, changed
remarkably.

maritime cultural landscapes in different societies

The structure of cultural landscapes is inseparable from the ruling social system,
and that is likewise true for maritime landscapes. Taking into consideration Viking
Age trade routes and nodal points in the present eastern Baltic, researchers thus face
the question of the structure of the local society in this period. It would also be
incorrect to overlook the ethnic diversity of the eastern Baltic, which was and still is
conspicuous in comparison with the rest of northern Europe (Figure 4.1). The same
variability presumably characterized pre-state political systems in this area — or at
least our interpretations of them tend to be diverse.
The development of maritime cultural landscapes in the north is characterized
by particular regularities that were most likely influenced by social changes. The
upper classes in their efforts to ensure their political influence gained gradually more
and more control over both the inter-regional and international trade routes and
trading harbours. Military activity at sea, including piracy, can also be directly con-
nected with the social system, while more ordinary pursuits like fishing or seal-
hunting are less influenced by it. The scarcity of archaeologically located harbour
sites, which characterizes the areas around the Baltic Sea before the 6th century ad,
was presumably also the result of social conditions: single chieftains simply did not
possess enough political power to control harbours with a larger hinterland. This was
made possible only by gradual power consolidation before and during the Viking Age.
The 6th–7th centuries are frequently considered revolutionary in areas around
the Baltic Sea. Changes in maritime activities were also demonstrated by the dis-
appearance of coastal settlements, a phenomenon which has usually been explained
by intensified piracy that made it too dangerous to live near the sea-shore (eg
Christoffersen and Porsmose 1996; Crumlin-Pedersen 1991; Näsman 1991; Welinder
2009, 395). Fishing villages reappeared on the shoreline only in the medieval period,
the 13th–14th centuries (eg Crumlin-Pedersen 1996), when the consolidated political
power could guarantee some safety against plundering raids from the sea.
Clearly more harbour sites around the Baltic have been located and archaeo-
logically investigated starting from the 8th–9th centuries (eg Filipowiak 1999;
Rebkowski 1999; Ulriksen 1998, 194–195, 217–223). Scandinavian archaeologists
˙ pointed out that the artefact material from harbour sites of this period alters.
have
Imported prestige ware was now complemented by imported but quite common
items like bronze jewellery, glass beads and combs. This phenomenon is believed to
be linked to the development of trade and the intensification of exploitation and dis-
tribution of resources (Bogucki 2004; Christophersen 1991; Ulriksen 1998, 222–223).
Social processes connected to this development resulted in more effective control over
MAM 37: 04-magi – Press

bound for the eastern baltic 43

Figure 4.1 The eastern Baltic in the Viking Age and the 12th century, showing trade routes to
Byzantium and the east crossing the area, and places mentioned in the text (M Mägi)

trade, which resulted in the centralization of trading places. Merchants and other
seamen visiting such places paid taxes to the chieftain(s) controlling the area, and the
latter in turn guaranteed their safety for the time they stayed in the harbour.
The next general change in the development of maritime landscapes around the
Baltic was the move of harbour sites closer to the open sea in the 11th–12th centuries
(for the phenomenon in Scandinavia, see eg Callmer 1991). It has sometimes been
associated with the establishment of towns, and thus the appearance of new power
centres (eg Ulriksen 1998, 222–228). The phenomenon itself is, however, more
MAM 37: 04-magi – Press

44 maritime societies of the viking and medieval world

comprehensive and apparent also in those Baltic regions where no towns developed
in the 11th–12th centuries. It is more likely, therefore, that the shifting of harbour
sites was predominantly caused by the appearance of large cargo-carrying ships,
mainly since the 11th century (Crumlin-Pedersen 1999), which in the northern areas
around the Baltic Sea was also accompanied by the increased speed of land-mass
elevation. The end of the Viking Age and the 12th century were periods when the
Viking Age high-water level in the Baltic Sea sank, which, together with the ordinary
processes of land-mass upheaval, changed the coast-line in the northern areas beyond
recognition in a comparatively short time. At the same time, the bigger new ships
needed harbours with deeper water nearby, which resulted in the abandonment of
several earlier harbour sites.
In Scandinavia, the concentration of trade into single large harbours/trading
centres started centuries earlier than on the eastern coast of the Baltic, and was
caused by the centralization of political power. These were early urban centres,
whose commercial practices and tax revenues were controlled by the kingships being
established (Skre 2007; Ulriksen 1998, 222–228; see Chapter 12). The same develop-
ment can, however, also be seen in areas without the direct jurisdiction of the early
kings. The island of Gotland, for instance, had six main harbours in the Viking Age
(Carlsson 1998; see Chapter 5). By the Middle Ages, all trade was concentrated in one
place — the town of Visby (Westholm 1985), while all other harbour sites had been
abandoned or functioned now only as small insignificant landing places used only by
some single farms.
Societies on the eastern shore of the Baltic Sea differed from the ones on the
western or southern shores, some of them more, some less. In the area of the present
Baltic States, concentration of political power in the Viking Age can be observed in
Lithuania (Kiaupa et al 2000, 38–41; Vaitkunskienė 1995), which is, however, histo-
rically rather more inland, and even nowadays spans only a comparatively short
distance of sea-coast. In Viking Age and 12th- to 13th-century Latvia, chieftains with
far-reaching political power have been assumed primarily in Semigalia, where their
presence has been indicated by strongly fortified hillforts and warrior graves supplied
with abundant weaponry. Princely burials with grave goods providing evidence of a
much higher grade of luxury than average cannot, however, be singled out even in
this area (Vaškevičiūtė 2004; 2007). From the 7th century onward similar burial rites
also characterized the Latgalians, who presumably possessed control over the trade
route along the River Daugava in the inland part of present Latvia. Archaeologists
dealing with the Latgalian area have nearly all identified the 10th–11th centuries as
the time of crucial changes in technology and economy, which resulted in the deeper
stratification of society and the increase of the role that international trade played in
the local strategies (eg Radiņš 1999, 131–153; Šnē 2002; Šnore 1987, 31–33).
Princes (princeps) of both Semigalia and Latgalia have been mentioned in 13th-
century written documents, when the Latgalian princedoms Koknes and Jersika were
politically dependent on the Kiev-Rus’ princedom of Polotsk (Polozk) further east
along the Daugava route. These princes were probably already converted to Eastern
Orthodox Christianity, perhaps being of Kiev-Rus’ origin, and their burial grounds
are neither known nor investigated archaeologically. It is possible, at least according
MAM 37: 04-magi – Press

bound for the eastern baltic 45

to early 13th-century chronicles, that the Livs, an ethnically Baltic Finnish people
inhabiting the lower reaches of the River Daugava, also paid tribute to the princes of
Polotsk. Neither the archaeological evidence nor historical sources suggest that the
society of the Livs was very hierarchical. Archaeologist Andris Šnē has interpreted
the political system of the Livs as comparatively egalitarian, where the majority of
people belonged to the stratum of free peasant-warriors (Šnē 1997). The 13th-century
written sources have described, in contrast to their ethnic Baltic neighbours, the
chieftains of the Livs mainly as ‘the elders and the betters’ (seniores et meliores), and
although their ruler Kaupo was described as ‘like a king and elder’ (quasi rex et
senior), his political influence was clearly very limited (Vassar and Tarvel 1975, 29).
Archaeologically, the 10th- to 12th-century Livs were characterized by burial customs
similar to the ethnic Balts, and although the number of weapon burials in their
cemeteries is high, graves with conspicuously more abundant artefacts cannot be
singled out (Šnore 1996; A Zariņa 2006).
An even more complicated task is to define Viking and later Iron Age society in
Couronia. This area was inhabited by Baltic-Finnish and Baltic mixed populations,
whose settlement areas are the subject of ongoing discussion (see eg Vasks 2008).
During most of the Viking Age and after that up to the 14th century the Couronians
predominantly practised cremation. Most of their graves were individual burials,
quite often abundantly equipped with weapons (eg Balodis 1940; Kulikauskas et al
1961, 387–388; Mägi 2007a; Stankus 1995; Žulkus 2000). The Couronians have been
repeatedly mentioned in written sources in connection with piracy, sometimes
together with the Osilians (the inhabitants of the Estonian island Saaremaa), whom
they also resembled according to their archaeological evidence (Mägi 2005a). Princes
like those of Semigalia or Latgalia have not been mentioned as existing among the
Couronians, nor can they be detected archaeologically.
The character of archaeological evidence changed in the area of present-day
Estonia. Although the Estonian landscape was marked by numerous hillforts like
the rest of the eastern Baltic, burial grounds especially in the mainland part of the
country were predominantly collectivistic, arranged presumably according to family
principle, and consisting of individually indistinguishable cremations. Particularly
during the 11th and 12th centuries, weapons were frequently placed in these ceme-
teries, but they can normally be found mixed with other grave goods and not con-
nected with specific burials. Furthermore, in cases of exceptional inhumations,
weapons have sometimes been found in female graves with abundant jewellery (Mägi
2002, 77–81). In general outline analogous burial rites also predominated in other
areas inhabited by the Baltic Finns during the Viking Age and up to the conversion to
Christianity (eg Kochkurkina 1981, 13ff; Lehtosalo-Hilander 1984, 284; Sedov 1987,
39ff; Shepherd 1999), and suggest social systems different from both the Balts and the
Scandinavians.
The abundance of weapons in cemeteries as well as the hierarchical settlement
pattern, as indicated by a number of hillforts, suggest that the Viking Age and later
Iron Age society in Estonia might be considered hierarchical. Nevertheless, judging
by the burial customs, it was characterized by strongly collectivist traits. One might
suggest that society was divided into clans with a single dominant family in each. The
MAM 37: 04-magi – Press

46 maritime societies of the viking and medieval world

representatives of the clan during periods of peace as well as unrest were chosen
among the clan members; they could vary and possess control only over certain social
aspects. Some members of dominant families probably had a bigger role than others,
but this authority rested on collective ownership and power, ie that of families, and
could not be directly associated with specific individuals. Several researchers have
brought into use the term heterarchy for describing the corporate way of power as it
might have characterized Viking Age and later Iron Age society in Estonia (Mägi 2013
and references therein).
The inhabitants of the Estonian islands, normally called by the general name of
Osilians, were singled out by their maritime activity. There is no doubt that the sea
with all its attractions occupied a prominent place in their lives, and that affected not
only fishing and local barter, but also overseas raids and control over sea routes
passing the Estonian islands to the rivers of the mainland. Recent research has
pointed out a number of late prehistoric or medieval harbour sites along the coasts of
Saaremaa, although none of them could correspond to a centre of international trade
(Mägi 2004).
It has been suggested that the social system on Saaremaa was more hierarchical
than in mainland Estonia, and this seems to be indicated by the more frequent occur-
rence of weapons in cemeteries, by the appearance of individual burials, among them
weapon graves, as early as the 7th century, and by strongly fortified hillforts in the
Viking Age and the 12th century (Mägi 2007a). Still, collective cemeteries also existed
on Saaremaa, alongside the individual graves, until Christianization, and early 13th-
century chronicles leave no doubt that the Estonian islands lacked chieftains with
further-reaching limits of power. Political power in Viking Age and 12th-century
Saaremaa was probably arranged according to collective principles, which among
other aspects could mean that administrative and military power were separated
from each other.
These diverse social structures presumably created confusion among merchants
who arrived from countries with hierarchical warrior-centralized societies or early
kingdoms, beginning with the problem of understanding who was actually holding
power and to whom to turn with requests or proposals. In addition, the language
factor cannot be overlooked. No doubt the Baltic-Finnish languages, belonging to
another language family than Indo-European, were difficult to master for people
from Scandinavia or other European areas. Still, the most essential trade routes
towards the east could not avoid areas inhabited by Baltic Finns during the Viking
Age and Middle Ages (in the greater part of these areas, Baltic-Finnish languages have
now been replaced by Baltic or Slavic ones).
As has been demonstrated by a decade of archaeological study in Estonia (see
eg Mägi 2004; 2007b), several parallel lines of development can be pointed out
between the changes of maritime landscapes in Estonia and other areas around the
Baltic Sea, notwithstanding the differences in late prehistoric social structures. Early
towns with significant importance for international trade did not arise anywhere in
the eastern Baltic before the 13th century, which can be explained by the lack of
centralized power in most of the local societies. Together with some bigger and more
influential trading places, there existed here numerous harbour sites of regional and
MAM 37: 04-magi – Press

bound for the eastern baltic 47

district importance, controlled by local chieftains. Towns were established in Estonia

and Latvia immediately after the German–Danish conquest in the first quarter of the
13th century. Despite the fact that medieval Livonia was not a centralized state either,
but consisted of nearly independent domains of different overlords, the bulk of inter-
national trade was now concentrated in towns. Four medieval towns in Estonia and
eight in Latvia belonged to the Hanseatic League. Four of them were situated on the
coast or in the vicinity of the sea, but all of them were connected with substantial
trade routes between the west and the east.

international trade routes in the baltic sea in the viking age and
early medieval period
The most convenient water routes to Russia, Byzantium and other areas further
east, over the big Russian rivers and the watersheds between them, started on the
coasts of present-day Estonia and Latvia (Figure 4.1). Ships sailing from the west and
along the islands of Åland turned from the southern Finnish coast to the place where
the town of Tallinn (Reval) was founded in the Middle Ages. The water route con-
tinued from there towards the east along the Estonian northern coast that was much
more convenient to navigate than the rocky Finnish coast with its multiple islets and
sparse inhabitation. The Viking Age predecessor of Tallinn and one of the relevant
nodal points on this route was a hillfort and trade centre at Iru, first established in the
Bronze Age (Mägi 2007b). From the north Estonian coast the products could there-
after be transported to Old-Ladoga and Novgorod, and from there along several
rivers to the Volga River and the Caspian Sea. An alternative was to turn southwards
to the Narva River, from there to the large lakes Peipsi and Pihkva, and then to
Pskov, Polotsk and further south.
Another well-known trade route started along the River Daugava, and ran over
several watersheds to the River Dnieper, Kiev, the Black Sea and Byzantium. A trade
centre with a mighty hillfort at Daugmale, about 22km from the sea in the estuary of
the Daugava River, was established as early as the Bronze Age, existed through the
Viking Age and the 12th century and was later replaced by medieval Riga (eg Radiņś
2001; Zemı̄tis 2007). Although the Daugava River has the greatest flow of water in the
eastern Baltic, its navigation was complicated by numerous impassable rapids — in
the early Middle Ages, there would have been more than one hundred rapids merely
in the territory of present Latvia (Radiņś 1998).1
During the Viking Age, but perhaps also later, an alternative was offered by a
route along the present Pärnu, Navesti, Tänassilma and Emajõgi Rivers, through the
large lakes between Estonia and Russia to Pskov, and from there along the Velikaja
River to Polotsk where it was united with the Daugava route.2 Although in central
Estonia this route presently consists of smaller and non-navigable rivers,3 it has been
marked as one big river on nearly all medieval or early modern maps, and the route
has been mapped separately even as late as in the 17th century (Ehrensvärd 2001,
64–65). This route was longer than sailing directly to Daugava, but demanded fewer
re-loadings of goods, thus making the transportation of them cheaper and less
laborious. In late prehistory and the early Middle Ages there were apparently enough
MAM 37: 04-magi – Press

48 maritime societies of the viking and medieval world

merchants choosing the Pärnu–Emajõgi River route instead of the River Daugava to
justify the establishment of three Estonian Hanseatic towns — Pärnu, Viljandi and
Tartu — along this waterway soon after the incorporation of present Estonia in the
western sphere. Viljandi and Tartu got their start as early as the Viking Age (eg Haak
2005; Tvauri 2001, 245–246). No Viking Age trade centre has been found at the site of
later Pärnu. This is to be expected, however, considering the inland location of early
harbours and trade centres, and the infertile soils around later Pärnu. Trade centres
in early Baltic Sea societies without considerable concentrations of political power
were as a rule located not right on the coast, but at a distance from it, and proximity
to fertile soils was essential for their existence (eg Mägi 2004).
The importance of the Pärnu–Emajõgi River route must, however, already have
diminished in the beginning of the medieval period. In addition to changes in terrain,
the character of long-distance trade also shifted in the Middle Ages. International
trade was concentrated in bigger towns and in fewer well-established trade routes,
and the quantity of traded products increased.
Several researchers have pointed to the importance of winter routes in historical
long-distance trade in the north (eg Sindbæk 2003), an aspect that had been virtually
neglected in earlier studies. Frozen waters made northern harbours inaccessible by
sea vessels during several months of the year. It was impossible to move over the sea
in late autumn and early spring when it was partly frozen, but not yet covered with
load-bearing ice. Especially in the Viking Age, when the climate was milder than
during later centuries, the sea might not always have frozen properly at all. As can be
read in the Chronicle of Henry the Livonian, access to the Estonian islands became
impossible for a great part of the year, and remained so for several years in the early
13th century. Even in the 19th century the non-load-bearing ice could sometimes
isolate the islands from the rest of the world for several months.
Inland, the winter routes ran along frozen rivers and over wetlands, which did
not need such long-lasting low temperatures as the sea for proper freezing. Especially
in forested areas, like Estonia, Latvia or north-west Russia, the winter routes were
often more used than summer roads, even in recent centuries. Søren Sindbæk has sug-
gested that rivers in Viking Age north-west Russia were used even predominantly as
winter routes. Experimental trips with prehistoric means of transport have demon-
strated that travellers from Scandinavia heading to Byzantium could not avoid
spending winter in Russia, and the distance from Old Ladoga to Gnezdovo could be
covered more than twice as fast with sledges along frozen rivers as along the same
route with boats (Sindbæk 2003; 2005, 247–254). It seems probable that rivers
through the eastern Baltic region were used as winter routes as well, although it does
not exclude their importance as water routes during navigable seasons.
At least some land-based routes could also have been used predominantly in
winter time. The role of land-based routes seems to have been comparatively impor-
tant for medieval trade in Livonia; the best known example is probably a mainland
route along the River Daugava, which is described in a contract between Germans
and Russians in ad 1229 (Radiņš 1998). The majority of international transport has
nevertheless valued water routes.
MAM 37: 04-magi – Press

bound for the eastern baltic 49

nodal points and trade centres in estonia and latvia

Viking Age and early medieval trade centres in the eastern Baltic are normally
seen in connection with local hillforts; in most cases, medieval towns were later
established in the same places. However, when we make comparisons between
Estonia and Latvia, there are considerable differences in the interpretation of hillforts
and settlements next to them.
It is traditional in Latvian archaeology to consider settlements beside or around
a number of late prehistoric hillforts as early urban centres and trade places. This
view reaches back to the formation of Latvian national archaeology, that is to the
1930s, and seems to be largely based on material collected from graves. Increases
in the abundance of grave goods as well as the occurrence of more burial places in
general led to an opinion that the 10th–11th centuries brought about great changes in
technology and economy which resulted in the formation of professional merchants,
the intensification of trade and the appearance of early urban centres (Radiņš 1999,
131–153).
The most important of these centres was Daugmale in the lower reaches of the
Daugava River (Figure 4.2). It is probably not an exaggeration to claim that Daug-
male is the ‘eastern Baltic’s Birka’. Archaeological excavations started at Daugmale
as early as the 1930s, but the greatest areas of the trading place and harbour site were
uncovered in the 1960s and 1970s, owing to building works for a power station on
the River Daugava. Following these rescue excavations using Soviet-style methods,
the 2ha early urban centre was flooded, making further excavations by later methods
impossible. The large hillfort next to it (with an area about 0.4ha) still stands,

Figure 4.2 The complex of Daugmale, the cemeteries and medieval Riga in the lower reaches
of the Daugava river (M Mägi)
MAM 37: 04-magi – Press

50 maritime societies of the viking and medieval world

although partly damaged. The complex was situated at the meeting point of different
ethnic groups.4 It was probably not a political, but rather a trade and handicraft
centre. One hundred and ninety coins found from there are dated to the 8th–12th
centuries, most of them to the first half of the 11th century (Radiņš 1998; 2006). The
quite sudden abandonment of the site in the 12th century can probably be connected
with the general tendency around the Baltic to move harbours closer to the open sea.
It coincided with the emergence of Riga where another trading place with a hillfort
beside it was established, replaced by the medieval town in the very beginning of the
13th century.
The hasty rescue excavations at Daugmale produced rich finds, but only a
modest number of publications. Most of the latter lack a wide-ranging discussion of
the site, instead concentrating on artefact typologies and issues of ethnicity (eg Radiņš
2001 and references therein). Only a few articles discuss the function of Daugmale as
a nodal point (eg Zemı̄tis 2007).
The find material collected from so-called Livonian cemeteries on the opposite
bank of the River Daugava and on the river island of Dole in the vicinity of the
Daugmale hillfort and trading centre has been published much more extensively.
Latvian archaeologists do not define these cemeteries as directly linked to the Daug-
male trade centre, still considering them to be burial places of near-by Livonian
villages, whose inhabitants took part in the activities on the hillfort and its settlement
(eg Radiņš 2006; A Zariņa 2006, 317–320). The find material of the big Laukskola
cemetery on the northern coast of the River Daugava, looked at in a broader per-
spective, nevertheless suggests that the burial ground was used by different people
who were active at the Daugmale complex and resided there for a considerable time.
Eastern Baltic Late Iron Age female graves of inferred ethnic groups are normally
distinguished by specific jewellery. Male ornaments and weapons, especially the latter,
were in most cases of much more international character. It is, for instance, archae-
ologically nearly impossible to differentiate grave goods of Livonian, Couronian,
Gotlandic, central Swedish, mainland Estonian or Osilian male burials, single arte-
fact types excluded.5 These last, however, do not occur in all graves. More specific
artefacts can be found in Latgalian and Semigalian male burials, and that is true not
only for ornaments, but also for weaponry (eg Griciuvienė et al 2005; Radiņš 1999;
Vaškevičiūtė 2007).
At Laukskola, however, female burials have been recorded with jewellery
characteristic of not only the Livic districts, but also several surrounding areas — eg
Semigalia, Sweden, Couronia, Saaremaa and mainland Estonia. The proportion of
females with non-local grave assemblages was estimated to be about one-third
(Spingis 2008, 373–375). So great a diversity of female burials has not been found in
any other prehistoric burial ground in the eastern Baltic; still, the only explanations
offered so far are intensive cultural contacts and/or perhaps inter-marriages. It can be
presumed that the same cultural diversity also characterized male burials in these
cemeteries, but cannot be traced so easily by relying only on evidence collected from
graves. It is especially clear from cremation graves with artefacts of more inter-
national character. The percentage of weapon burials among cremations was con-
spicuously high, and twice as many men as women had been buried this way at
MAM 37: 04-magi – Press

bound for the eastern baltic 51

Laukskola (G Zariņa 2006). Cremation was the prevailing burial custom on Viking
Age and 12th-century Saaremaa, mainland Estonia, and Couronia (Mägi 2002, 125–
137; Selirand 1974, 178–190; Stankus 1995; Žulkus 1991, 11; 2000). Livs practised pre-
dominantly inhumations, although cremations were not completely unknown among
them either (Šnore 1996; Tõnisson 1974; A Zariņa 2006, 408). It can be suggested,
therefore, that graves of more international character at so-called Livonian cemeteries
next to the Daugmale complex were burial places not only of natives, but also of
people from other areas who had died during their stay in the trade centre.
At the end of prehistory, strongholds were built on the banks of the Daugava
River at distances of 20–40km. They were normally located next to rapids, and
probably functioned as tax-collecting points when vessels and wares had to be landed
and launched again. Besides Daugmale, the hillforts of Jersika and Koknes especially
stand out as influential sites. They functioned predominantly as political centres and,
at least in the 12th–13th centuries, as the residences of local princes. International
trade is believed to have played a much smaller role in them, indicated by the rather
modest number of coins found at these sites (Radiņš 2001).
The trading centre at Grobiņa, with an estimated size of 3ha, has not been
excavated at all, although it is marked as the only Viking Age nodal point in the
eastern Baltic on many maps in archaeological publications. Small-size excavations
were conducted on the hillfort of Grobiņa (with a yard area of 0.36ha) before World
War Two. The culture layer, which included finds of local ceramics and some metal
artefacts, dated the site to the 9th–13th centuries (Nerman 1958). As is indicated by
excavated cemeteries in the vicinity, a Scandinavian colony was situated at Grobiņa
ad 600–800/850, thus mainly in the period predating the Viking Age (Nerman 1958).
Excavations conducted at different burial grounds around the hillfort in the 1960s
and 1980s have at the same time pointed to local inhabitation throughout prehistory
(eg Petrenko and Ozere 1988; Petrenko and Urtāns 1995).
The importance of Grobiņa as a nodal point of international trade is brought
into question by its location on the shore of Couronia, on the bank of the small
Ālande River. The location was presumably determined by a suitable harbour site on
a small lake near the sea coast.6 It is difficult, however, to define Grobiņa as a crossing
point of essential trade routes, although some land-based route running towards the
east probably started there (see also Bogucki 2006). It can be assumed that Grobiņa
was mainly connected with the Amber Way, from south-eastern Baltic coastal areas
along the River Vistula down to southern Europe and Byzantium (Mägi 2011).
Latvian archaeologists also interpret settlements around several Semigalian and
Couronian hillforts as early urban centres (eg Mežotne, Tērvete, Kuldiga and Talsi).
Although they cover, according to general interpretation, several hectares, some-
times even more than ten hectares, only a few of them have been archaeologically
excavated (Urtāns and Asaris 1998; Urtāns et al 1998). One of the biggest and best
investigated complexes of this sort is probably Mežotne at the Lielupe River in
Semigalia, slightly more than 50km south of Riga. The complex consists of two
hillforts, a 13ha settlement between them, and two cemeteries in the neighbourhood.
The settlement is dated to ad 500–1200, the main hillfort to the 9th–14th centuries.
Romas Jarockis (2001) has suggested that since the 9th century at the latest Mežotne
MAM 37: 04-magi – Press

52 maritime societies of the viking and medieval world

can be interpreted as an early urban centre that participated in international trade.

The question of what could have been the internationally important trade route
passing through Mežotne was, however, left open. As with Mežotne, the location
of several other early Latvian centres at a distance from the more important trade
routes suggests that their main role was in exchange between neighbouring regions.
Moreover, a number of them might rather have functioned as predominantly political
centres.
Large hillforts presumably functioning primarily as trade centres on a regional
level also existed in Estonia. Most of them occurred or were rebuilt from some earlier
fortifications in the late 11th and 12th centuries, and some of them remained in use
until the second quarter or the middle of the 14th century. Several such hillforts were
characterized by locations remote from densely populated areas in the middle of
broad wetlands (eg Varbola, Soontagana), and were therefore probably connected
with winter routes. Still, some of the large hillforts were also situated near settlement
centres, at a distance of 5–7km from the coast, and had access to the sea by way of a
smaller river (eg Pöide, Pada). Harbour sites of such hillforts are known and even
excavated. Nearly all of the hillforts were very strongly fortified, several of them
being surrounded by massive drystone walls.
In contrast to Latvia, there were no early urban centres other than the big Late
Iron Age Estonian hillforts, and even where settlements have been detected in their
vicinity, these have not been interpreted as early urban centres in Estonian archaeol-
ogy. Such hillforts probably became heavily populated only seasonally, and for most
of the year there could have been just a small guard group present. These fortifica-
tions were first of all centres of local barter or trade between neighbouring regions,
where few or no artefacts witnessing far-reaching contacts have come into light.
The best example is Varbola, about 50km south of Tallinn. Its 2ha enclosure is
surrounded by a drystone wall, surviving today to a height of 10m. Inside the hillfort,
remains of approximately a hundred buildings with ovens can be counted. Most of
them were probably small horizontal-log buildings of approximately the same size
(eg Tõnisson 1999). It can be assumed that people stayed in such little huts only
temporarily, spending most of the time of year in their ordinary living quarters close
to arable lands. In contrast to most of the hillforts of this sort, Varbola has been
repeatedly mentioned both in German and Russian sources, although it cannot be
connected with any definite district, and presumably did not function as a political
centre. The latter is unlikely because of its location in the middle of forests and
wetlands, with only limited areas of arable soil in the vicinity. Varbola was probably
the crossing point of substantial winter roads, but admittance to the hillfort during
warmer seasons might have been more complicated.
On the island of Saaremaa, a large Late Iron Age hillfort is situated and has been
partly excavated at Pöide. It was originally a Viking Age stronghold, abandoned early
in the 11th century, and rebuilt as a strongly fortified large hillfort in the 12th century
(Lõugas and Mägi-Lõugas 1994). At a distance of 7km from Pöide, on the eastern
coast of Saaremaa, there is a harbour site that was probably controlled by the hillfort.
Excavations have been conducted only at the Viking Age part of the one-time
MAM 37: 04-magi – Press

bound for the eastern baltic 53

harbour, yielding remains of buildings, abundant ceramics, animal bones and boat
rivets, but only a few artefacts linked to neighbouring regions (Mägi 2005b). A
cemetery close to a large hillfort of this sort has been excavated only in Pada, north
Estonia; the 171 inhumation burials there, dated to the 12th–13th centuries, were
provided with very homogeneous grave goods (unpublished excavation reports of
Toomas Tamla; see also eg Limbo 2004). Neither the cemetery at Pada nor the
harbour site near Pöide suggest that very far-reaching international trade was carried
out at these hillforts.
It seems that the most important trade centres in Estonia were functioning as
political centres as well, at least in the later part of prehistory. They did not first
appear in the 11th–12th centuries, like the large hillforts described earlier, but in the
5th–6th centuries, and some of them were later replaced by medieval towns. In this
connection, Tartu and Tallinn should first be mentioned.
The hillfort of Iru has frequently been referred to as the predecessor of Tallinn.
It is situated 8.5km as the crow flies from the medieval town, on the bank of the Pirita
River (Figure 4.3). In the Viking Age it remained 4–5km inland from the river mouth
opening into the Bay of Tallinn. The hillfort was well protected, being located on a
steep and high bank in a ‘U’-shaped river bend, and fortified additionally with walls.
The time of its prosperity was the Viking Age, when about 0.5ha of the hilltop was
used, and in addition there was a settlement outside the walls (Lang 1996, 34–104).
A narrow terrace between the stronghold walls and the river, where the culture

Figure 4.3 Medieval Tallinn and the hillfort of Iru (Tallinn and its surroundings in ad 1689,
Estonian Historical Archives 1-2-III-2)
MAM 37: 04-magi – Press

54 maritime societies of the viking and medieval world

layer was particularly intensive, can be interpreted as a possible river harbour (Mägi
2007b).
Valter Lang has assumed that the emergence (or actually re-emergence, since the
hilltop had already been in use in the Bronze Age) of the Iru hillfort in the 5th–6th
centuries, and its prosperity in the Viking Age, was caused by agrarian settlement
moving there from surrounding areas. He claimed that the number of inhabitants in
nearby areas dropped in that period, arguing from the estimation that the number
of stone graves decreased (Lang 1996, 476–477; 2004). The number of cemeteries of
that period in the vicinity of the hillfort is, however, also very modest. The move
of settlement was, according to Lang, caused by some unspecified social or political
pressure in society, but trade had no role to play. He argued that the unimportance of
trade was proved by the general tendency for north Estonian coastal hillforts (and the
settlements next to them) to be situated not immediately on the seashore, but on river
banks some kilometres inland (Lang 1996, 372–373).
An alternative explanation is, however, that the Iru complex, like other similar
hillforts with or without settlements in north Estonia, was not a fully permanent
living place during the 7th to 10th centuries, but a central-place market as occurred
elsewhere in northern Europe at the same time (Skre 2007). More people probably
gathered there only seasonally, and the community consisted normally of a magnate’s
farm, or perhaps only a group of guards. It is difficult to imagine how the small area
of arable land at Iru could provide a permanent living for an abruptly multiplied
population, or what profit a social elite would gain from the semi-compulsory move
of people to the edge of arable lands, as Lang has suggested. As discussed above,
moreover, the location on a river bank a few kilometres inland from the actual coast
is quite characteristic of Viking Age trade centres.
The find material at Iru, however, does not differ much from that in other pre-
Viking and Viking Age settlement sites in Estonia, simply being more intensive than
average, and bearing traces of handicraft. The number of artefacts indicating far-
reaching contacts is not very impressive, and most of the coins found in various parts
of the complex date to the 11th century. Find material similar to agrarian settlements
does, however, also characterize most of the Viking Age harbours/trading places in
Scandinavia, excluding only the most essential nodal points of international trade (eg
Carlsson 1991; Ulriksen 1998, 113–142).
Archaeological excavations during the last two decades have demonstrated the
possibility that a trade centre, perhaps of more international character, also existed
next to the later medieval town of Tallinn. Most find material known so far belongs
to the Viking Age and the Late Iron Age, but some radiocarbon dates point to even
earlier activity. The area of this settlement was, unfortunately, built over in modern
times, which makes closer study of the site very complicated (Mägi 2015). It is clear,
however, that this centre continued even after the 11th century, when the Iru complex
was abandoned.
One of the earliest Estonian hillforts mentioned in written sources is Tartu on
the River Emajõgi, called grad (hillfort or fortified settlement) Jurjev in Russian
chronicles referring to ad 1030 or ad 1036. It was presumably an important nodal
point on the Pärnu River–Emajõgi River route, with a large settlement nearby (the
MAM 37: 04-magi – Press

bound for the eastern baltic 55

area of the latter is estimated to have been more than 20ha in the 11th century: see
Tvauri 2001, 246). Intensive archaeological excavations at the later medieval town
have proved that the hillfort and settlement complex was first established as early as
the 8th century. Although so far nearly all researchers have considered Tartu as the
centre of trade and handicraft in eastern Estonia, this has been challenged recently by
archaeologist Andres Tvauri.
In his monograph about prehistoric Tartu, Tvauri has taken the view that,
contrary to widespread opinion, trade played only a marginal role in the economy of
late prehistoric Estonia. No arguments for this opinion, however, were presented,
except for the fact that no scales or weights, and only one prehistoric coin, were
found in archaeological excavations at the later medieval town of Tartu. Accord-
ingly, Tvauri suggested that prehistoric Tartu was not a trade centre, but just an
ordinary settlement, perhaps a political centre at the local level, and the water route
through the rivers of central Estonia never existed. He also argued that the upper
classes of Novgorod were probably not interested in one more trade route through
the area of present Estonia (Tvauri 2001, 191–193, 208–212). Nevertheless, the emer-
gence of trade routes cannot be defined only through the interests of the towns in
Kiev-Rus’. Neither have cultural geographical aspects of the location of Tartu been
considered by Tvauri in his argumentation.
In the last decade several excavations have been carried out at smaller, regional
or even district-level harbour sites in Estonia, especially on the island of Saaremaa
(see above). Such harbours were normally located close to district centres. They were
used primarily by the inhabitants of surrounding villages and farms, and their safety
was presumably guaranteed by the most prominent families in the neighbour-
hood, who benefited from taxing vessels landing at the place. Such harbour sites in
Estonia were marked on coastal landscapes with often conspicuous stone graves,
where most likely the members of the family controlling the site were buried (Mägi
2004, 140–142; for a similar phenomenon, eg on Gotland, see Carlsson 1991). The use
of stone graves to mark harbours in the landscape in pre-state society can be com-
pared to the custom of building early (Romanesque) churches or chapels at similar
sites in the Christian world (eg Carlsson 1999; Ulriksen 1998). Some harbour sites
excavated on Saaremaa or on the northern coast of Estonia have been surrounded by
abundant graves and cemeteries, suggesting that these sites functioned not always as
trade or fishing centres, but also as ritual centres, places of regular meetings for rites
and negotiations (Mägi 2009).

conclusions
As demonstrated above, the study of trade routes and nodal points in the eastern
Baltic is to a certain extent still influenced by a lack of theoretical considerations.
First of all, discussions regarding topography and cultural landscapes are often not
sufficient. The frequently different interpretations of similar archaeological sites in
Estonia and Latvia is also obvious. Latvian archaeologists consider hillfort and settle-
ment complexes to be early urban centres, whereas the attitude of Estonian archae-
ologists is rather sceptical, resulting in a complete divergence over the one-time
MAM 37: 04-magi – Press

56 maritime societies of the viking and medieval world

importance of trade routes. It is also possible, however, that the character of central
places, including possible trade centres, really differed according to various social
systems in different areas of the eastern Baltic.
The large settlements beside several Latvian hillforts can possibly be linked with
a more hierarchical social order in these areas, especially when the location of such
centres does not support the interpretation of them as important nodal points of
international trade. It is more convincing to consider them as political and adminis-
trative centres, an idea which is supported by the written sources of the 13th century
describing them as residences of local princes.
The character of trade centres was different in areas inhabited by Baltic Finns.
Although there were often also settlements beside Viking Age hillforts in Estonia,
these were never as large as for instance in Latgalia or Semigalia. It is scarcely even
possible to define settlements at the sites of the later Tartu and Tallinn (Iru) as proper
urban centres. Estonia in the 12th century witnessed a tendency to centralize trade in
single big hillforts, without open settlements next to them. The network of harbours
along the Estonian coast was comparatively egalitarian, consisting of multiple district-
level harbour sites together with probably an even larger number of small landing
places, meant only for local use. The appearance in the 12th century of single trade
harbours of broader international importance (eg Tallinn, Riga), normally located at
the junction of several districts but in the vicinity of some Viking Age and earlier
hillfort of mercantile importance (Iru and Daugmale respectively), seems possible.
The maritime cultural landscape in the eastern Baltic thus corresponds to social
systems in these areas. Collectively organized power in the areas inhabited by Baltic
Finns probably determined the administration of trade centres by several dominant
families. One of the consequences of a more egalitarian social structure was pre-
sumably a smaller percentage of luxury ware among imports, and thus also the infre-
quent occurrence of them as archaeological evidence. Conversely, the rise of early
urban centres and a higher percentage of luxury artefacts among trade wares charac-
terized strongly hierarchical societies, particularly in developing or already estab-
lished kingdoms. In the eastern Baltic early urban centres tended to develop more
frequently in areas where the social system is believed to have been more hierarchical.
In the Viking Age and Middle Ages international trade presumably still followed
routes that best suited the means of transport, and thus did not avoid areas with dif-
ferently organized social orders, at least not without urgent reason. However, archae-
ologically, nodal points in areas characterized with heterarchic power structures can
be complicated to recognize. It is difficult to imagine year-round habitation at trade
harbours in a society without sufficient centralization, however relevant these places
might have been from the point of view of international trade. Even catering for
people there, especially in the winter season when the sea was not navigable, would
have been quite complicated. The nodal points of winter routes were situated some-
where else, often in districts with complicated access in summertime. All these aspects
are reflected in the archaeological evidence of Estonia and, partly, Latvia, but do not
diminish the importance of international trade routes to the lives of local com-
munities through which they passed. Neither do they reduce the relevance of the
region from the viewpoint of traders coming from distant lands.7
MAM 37: 04-magi – Press

bound for the eastern baltic 57

notes
1 Most of the rapids in the Daugava River, have of archaeological material in the eastern Baltic
disappeared now due to the building of several region, as elsewhere, requires great caution.
hydropower plants during the 20th century, 5 As we can read in the early 13th-century chron-
which dammed water in the river. icle, it could also be impossible to distinguish
2 Still, not all Estonian archaeologists agree that warriors from at least some of these areas in a
this trade route existed (see later in this article). battle (Brundage 2003).
3 Possibly caused by changes in the west Estonian 6 The majority of the Couronian coast down to
terrain due to isostatic uplift and reversion to Klaipeda is straight and open to winds, where it
swamps. must have been problematic to find suitable land-
4 Some Latvian scholars believe that Daugmale ing places, especially outside the entrances to
used to be a ‘Semigallian’ site in the first half of local rivers.
7 Writing this article was conducted in the frame-
the Viking Age, ‘Livonian’ in the 11th century and
‘Semigallian’ again after that, when it might have work of the research project ‘Passages of Water-
been Portus Semigallia once mentioned in written borne Interaction and their Impact on Estonian
sources (eg Radinš 2001). Nevertheless, the num- Late Prehistoric and Early Medieval Cultural
ber of other Semigallian sites in the area is quite Landscapes’ (Grant No 9027 of the Estonian
modest. Not all researchers agree with the ethnic Science Foundation), and supported by institu-
definitions of Daugmale, and the Portus Semi- tional research funding IUT 18–8 of the Estonian
gallia is often believed to have been situated at Ministry of Education and Research. The author
the River Lielupe instead, which runs through is grateful to Gordon Snow for checking the
the Semigallian area. The ethnic interpretation language of this paper.

references
Balodis, A, 1940, Det äldsta Lettland, Almqvist & Wiksell, Uppsala
Bogucki, M, 2004 ‘Viking Age ports of trade in Poland’, Estonian Journal of Archaeology 8,
100–127
Bogucki, M, 2006, ‘Grobiņa — a sign of an early future port of trade in the Balt lands’, in
M Bertašius (ed), Transformatio Mundi: The Transition from the Late Migration Period to
the Early Viking Age in the East Baltic, Kaunas University of Technology, Department of
Philosophy and Cultural Science, Kaunas, 93–106
Brundage, J A (ed), 2003 The Chronicle of Henry of Livonia, Columbia University Press, New York
Callmer, J, 1991 ‘Platser med anknytning till handel och hantverk i yngre järnålder: exempel från
södra Sverige’, in P Mortensen and B M Rasmussen (ed), Høvdingesamfund og Kongemagt:
fra Stamme til Stat i Danmark 2, Aarhus universitetsforlag, Aarhus, 29–47
Carlsson, D, 1991 ‘Harbours and trading places on Gotland ad 600–1000’, in O Crumlin-Pedersen
(ed), Aspects of Maritime Scandinavia ad 200–1200: Proceedings of the Nordic Seminar on
Maritime Aspects of Archaeology, Roskilde, 13th–15th March, 1989, The Viking Ship
Museum, Roskilde, 145–158
Carlsson, D, 1998 Vikingahamnar: ett hotat kulturarv, ArkeDok Skrifter 1, ArkeoDok, Visby
Carlsson, D, 1999 Gård, hamn och kyrka: en vikingatida kyrkogård i Fröjel, Gotland University
College, Centre for Baltic Studies, Visby
Christoffersen, J and Porsmose, E, 1996 ‘Den fynske kystzones bebyggelsesmønstre i jernalder,
vikingetid og middelalder’, in O Crumlin-Pedersen, E Porsmore and H Thrane (ed), Atlas over
Fyns kyst i jernalder, vikingetid og middelalder, Odense universitetsvorlag, Odense, 154–160
Christophersen, A, 1991 ‘Ports and trade in Norway during the transition to historical time’, in
O Crumlin-Pedersen (ed), Aspects of Maritime Scandinavia ad 200–1200: Proceedings of the
Nordic Seminar on Maritime Aspects of Archaeology, Roskilde, 13th–15th March, 1989, The
Viking Ship Museum, Roskilde, 159–170
Crumlin-Pedersen, O, 1991 ‘Maritime aspects of the archaeology of Roman and Migration-Period
Denmark’, in O Crumlin-Pedersen (ed), Aspects of Maritime Scandinavia ad 200–1200:
MAM 37: 04-magi – Press

58 maritime societies of the viking and medieval world

Proceedings of the Nordic Seminar on Maritime Aspects of Archaeology, Roskilde, 13th–15th

March, 1989, The Viking Ship Museum, Roskilde, 41–54
Crumlin-Pedersen, O, 1996 ‘Kystforsvaret’, in O Crumlin-Pedersen, E Porsmore and H Thrane
(ed), Atlas over Fyns kyst i jernalder, vikingetid og middelalder, Odense universitetsvorlag,
Odense, 182–193
Crumlin-Pedersen, O, 1999 ‘Ships as indicators of trade in northern Europe 600–1200’, in J Bill and
B L Clausen (ed), Maritime Topography and the Medieval Town: Papers from the 5th Inter-
national Conference on Waterfront Archaeology in Copenhagen, 14–16 May 1998, The
National Museum of Denmark, Copenhagen, 11–20
Ehrensvärd, U, 2001 ‘Topographica Estoniae: Handritade kartor och ritningar över Estland i
svenska offentliga samlingar/Hand gezeichnete Karten und Zeichnungen von Estland in
schwedischen öffentlichen Sammlungen’, Eesti Teadusliku Seltsi Rootsis Aastaraamat/
Societatis Litterarum Estonicae in Svecia XII (1991–1999), 5–289
Filipowiak, W, 1999 ‘Wolin und Szczecin — Hafen und Topographie der mittelalterlichen Stadt’,
in J Bill and B L Clausen (ed), Maritime Topography and the Medieval Town: Papers from
the 5th International Conference on Waterfront Archaeology in Copenhagen, 14–16 May
1998, National Museum of Denmark, Copenhagen, 61–70
Griciuvienė, E, Grižas, G and Buža, Z, 2005 ‘Semigallians, V–XIII cent’, in L Urtė Urbaitė (ed),
Žiemgaliai. Baltu˛ archeologijos paroda katalogas. The Semigallians: Baltic Archaeological
Exhibition Catalogue, Lietuvos nacionalis muziejus, Latvijas Vēstures muzejs, Vilnius, Riga,
23–152
Haak, A, 2005 ‘Archaeological investigations in the Late Iron Age settlement site of Huntaugu and
at the castle of Teutonic Order in Viljandi’, in Ü Tamla (ed), Archaeological Fieldwork in
Estonia 2004, Muinsuskaitseamet, Tallinn, 89–101
Jarockis, R, 2001 ‘Geographical situation, chronological development and early economy: a com-
parative analysis of two hillfort-settlement complexes in Semigallia 900–1400 ad’, in M Auns
(ed), Lübeck Style? Novgorod Style? Baltic Rim Central Places as Arenas for Cultural
Encounters and Urbanisation 1100–1400 ad , Nordik, Riga, 77–87
Kiaupa, Z, Kiaupienė, J and Kuncevičius, A, 2000 The History of Lithuania before 1795, Lithu-
anian Institute of History, Vilnius
Kochkurkina, S I, 1981 Arheologicheskye pamyatniki Karelii V–XV vv, Nauka, Leningrad
Kulikauskas, P, Kulikauskienė, R and Tautavičius, A, 1961, Lietuvos archeologijos bruožai,
Valstybinės ir mokslinės literatūros leidykla, Vilnius
Lang, V, 1996 Muistne Rävala: Muistised, kronoloogia ja maaviljelusliku asustuse kujunemine
Loode-Eestis, eriti Pirita jõe alamjooksu piirkonnas, Eesti Teaduste Akadeemia Ajaloo
Instituut, Tallinn
Lang, V, 2004 ‘Varalinnalised keskused (aolinnad) hilismuinasaegses Eestis’, in T Kala (ed), Kui
vana on Tallinn? 13. mai 2004 konverentsi ettekanded ja diskussioon, Tallinna Linnaarhiiv,
Tallinn, 7–27
Lehtosalo-Hilander, P-L, 1984 ‘Keski- ja myöhäisrautakausi’, in E Laaksonen, E Pärssinen and
K J Sillanpää (ed), Suomen historia, Volume 1, Weilin+Göös Porvoo, Helsinki, 250–405
Limbo, J, 2004 ‘Pada kalmistu indiviididel esinevad hambapatoloogiad (XII–XIII sajandil)’,
Estonian Journal of Archaeology 8, 49–75
Lõugas, V and Mägi-Lõugas, M, 1994 ‘Investigation of ancient monuments at Pöide 1991–1992’,
Teaduste Akadeemia Toimetised, Humanitaar- ja Sotsiaalteadused 43(1), 27–33
Mägi, M, 2002 At the Crossroads of Space and Time: Graves, Changing Society and Ideology on
Saaremaa (Ösel), 9th–13th centuries ad , Gotland University College Centre for Baltic Studies,
Institute of History, Department of Archaeology, Tallinn
MAM 37: 04-magi – Press

bound for the eastern baltic 59

Mägi, M, 2004 ‘‘‘. . . Ships are their main strength’’: harbour sites, arable lands and chieftains on
Saaremaa’, Estonian Journal of Archaeology 8, 128–162
Mägi, M, 2005a ‘On the mutual relationship between late prehistoric Saaremaa and the Livs’, in
S Mäntylä (ed), Rituals and Relations: Studies on the Society and Material Culture of the
Baltic Finns, Annales Academiae Scientiarum Fennicae, Helsinki, 187–206
Mägi, M, 2005b ‘Viking Age harbour site at Tornimäe, eastern Saaremaa’, in Ü Tamla (ed),
Archaeological Field Works in Estonia 2004, Muinsuskaitseamet, Tallinn, 65–75
Mägi, M, 2007a ‘Collectivity versus individuality: the warrior ideology of Iron Age burial rites on
Saaremaa’, Archaeologia Baltica 8, 263–272
Mägi, M, 2007b ‘Iru linnusest Püha Birgitta kloostrini. Merenduslik kultuurmaastik Pirita jõe
alamjooksul’, Studies on Art and Architecture 4, 17–40
Mägi, M, 2009 ‘Harbour sites in ritual landscape — the example of Viltina on southern Saaremaa’,
Estonian Journal of Archaeology 13, 110–131
Mägi, M, 2011 ‘Viking Age and early mediaeval eastern Baltic between the west and the east’, in
S Imsen (ed), Taxes, Tributes and Tributary Lands in the Making of the Scandinavian King-
doms in the Middle Ages, Norgesveldet, Tapir Academic Press, Trondheim, 189–233
Mägi, M, 2013 ‘Corporate power structures as indicated in archaeological evidence: the case of
Estonia in the Middle Iron Age and Viking Age’, Fennoscandia Archaeologica 30, 107–125
Mägi, M, 2015 Rafala: Idateest ja Tallinna algusest, Argo, Tallinn
Näsman, U, 1991 ‘Sea trade during the Scandinavian Iron Age: its character, commodities, and
routes’, in O Crumlin-Pedersen (ed), Aspects of Maritime Scandinavia ad 200–1200: Proceed-
ings of the Nordic Seminar on Maritime Aspects of Archaeology, Roskilde, 13th–15th March,
1989, The Viking Ship Museum, Roskilde, 23–40
Nerman, B, 1958 Grobin-Seeburg. Ausgrabungen und Funde, Almqvist & Wiksell, Stockholm
Petrenko, V P and Ozere, J, 1988 ‘Izrakumi priediena lı̄dzenajā kuršu kapulaukā’, Zinātniskās
atskaites sesijas materiāli par arheologu un etnogrāfu 1986 un 1987 gada petı̄jumu rezult-
tātiem. Arheolog‘ ija, 109–112
Petrenko, V and Urtāns, J, 1995 The Archaeological Monuments of Grobiņa, Statens Historiska
Museum, Riga and Stockholm
Radiņš, A, 1998 ‘Some notes on the Daugava Way: the end of 12th–beginning of 13th century.
Time of changes on the lower Daugava’, in N Blomkvist (ed), Culture Clash or Compromise?
The Europeanisation of the Baltic Sea Area 1100–1400 ad , Gotland Centre for Baltic Studies,
Gotland University College, Visby, 178–191
Radiņš, A, 1999 10.–13. gadsimta senkapi latgaļu apdzı̄votajā teritorijā un Austrumlatvijas
etniskās, sociālās un politiskās vēstures jautājumi, Latvijas Vēstures Muzeja raksti 5 —
Arheolog‘ ija, NIMS, Riga
Radiņš, A, 2001 ‘Daugmale, Jersika, Riga: the development of economic and political centres along
the lower reaches of the Daugava’, in M Auns (ed), Lübeck Style? Novgorod Style? Baltic Rim
Central Places as Arenas for Cultural Encounters and Urbanisation 1100–1400 ad , Nordik,
Riga, 89–94
Radiņš, A, 2006 ‘Lower Daugava area in the 1st to 11th century: ethnic, economic, social and
political change — on the question of activity along Daugava waterway’, in M Bertašius (ed),
Transformatio Mundi: The Transition from the Late Migration Period to the Early Viking
Age in the East Baltic, Kaunas University of Technology, Department of Philosophy and
Cultural Science, Kaunas, 81–92
Rebkowski, M, 1999 ‘The maritime topography of medieval Kołobrzeg’, in J Bill and B L Clausen
˙ (ed), Maritime Topography and the Medieval Town: Papers from the 5th International Con-
ference on Waterfront Archaeology in Copenhagen, 14–16 May 1998, The National Museum
of Denmark, Copenhagen, 55–60
MAM 37: 04-magi – Press

60 maritime societies of the viking and medieval world

Sedov, V V, 1987 ‘Finno-ugry i balty v epohu srednevekovya’, Arheologiya SSSR, Nauka, Moscow
Selirand, J, 1974 Eestlaste matmiskombed varafeodaalsete suhete tärkamise perioodil (11.–13.
sajand), Eesti Raamat, Tallinn
Shepherd, D J, 1999 Funerary Ritual and Symbolism: An Interdisciplinary Interpretation of Burial
Practices in Late Iron Age Finland, British Archaeological Reports International Series 808,
Hadrian Books, Oxford
Sindbæk, S, 2003 ‘Varægiske vinterruter: slædetransport i Rusland og spørgsmålet om den tidlige
vikingetids orientalske import i Nordeuropa’, Fornvännen 3, 179–193
Sindbæk, S, 2005 Ruter og rutinisering: Vikingetidens fjernhandel i Nordeuropa, Multivers
Academic, Copenhagen
Skre, D, 2007 ‘Post-substantivist towns and trade: ad 600–1000’, in D Skre (ed), Means of
Exchange: Dealing with Silver in the Viking Age, Aarhus University Press, Aarhus, 327–341
Šne, A, 1997 ‘Social structures of Livonian society in the Late Iron Age (10th–early 13th century)’,
in O W Jensen, H Karlsson and A Vijups (ed), Arkeologi över gränser: möten mellan lettisk
och svensk arkeologi, Göteborgs Universitet, Institut för arkeologi, Göteborg, 183–207
Šnē, A, 2002 Sabiedrı̄ba un vara: sociālās attiecı̄bas Austrumlatvijā aizvēstures beigās, Intelekts,
Riga
Śnore, E, 1987 Kivtu kapulauks, Zinātne, Riga
Šnore, E, 1996 ‘Daugavas libieši Doles sala’, Arheolog‘ ija un Etnogrāfija xviii, 111–130
Spingis, R 2008, Bruņrupuču saktas ar krūšu važiņrotām un lı̄biešu kultūras attı̄stı̄ba Daugavs
lejtecē 10.–13. gadsimtā, Latvijas Vēstures Institūta Apgāds, Rı̄ga
Stankus, J, 1995 Banduźiu kapinynas, Lietuvos Archeologija 12, Danielius, Vilnius
Tõnisson, E, 1974 Die Gauja-Liven und ihre materielle Kultur (11. Jh–Anfang 13. Jhs.): Ein Beitrag
zur ost-baltischen Frühgeschichte, Eesti Raamat, Tallinn
Tõnisson, E, 1999 ‘Die Wallburg Warbola — ein Zentrum im westlichen Estland: aspekte zur
militärischen, politischen und sozialen Funktion’, in N Blomkvist and S-O Lindquist (ed),
Europeans or Not? Local Level Strategies on the Baltic Rim 1100–1400 ad , Gotland
University College, Centre for Baltic Studies, Oskarshamn, 173–184
Tvauri, A, 2001 Muinas-Tartu, Muinasaja teadus 10, Ajaloo Instituut, Tartu Ülikool, Tartu and
Tallinn
Ulriksen, J, 1998 Anløbspladser: Besejling og bebyggelse i Danmark mellem 200 og 1100 e. Kr. En
studie af søfartens pladser på baggrund af undersøgelser i Roskilde Fjord, Vikingeskibshallen,
Roskilde
Urtāns, J and Asaris, J, 1998 Latvijas rietumu daļas jaunatklātie pilskalni, Fonds ‘Mantojums’, Riga
Urtāns, J, Šnē, A and Asaris, J, 1998 Latvijas pilskalni, Valsts Kultūras Pieminekļu Aizsardzı̄bas
Inspekcija, Riga
Vaitkunskienė, L, 1995 ‘The formation of a warrior elite during the Middle Iron Age in Lithuania’,
Archaeologia Baltica 1, 94–106
Vaškevičiūtė, I, 2004 Žiemgaliai V–XII amžiuje, Leidykla, Vilnius
Vaškevičiūtė, I, 2007 ‘Semigallian warrior weaponry and its reflection in burial rites in the fifth to
the 12th century ad’, Archaeologia Baltica 8, 214–222
Vasks, A, 2008 ‘Kurši Padurē?’, in I Stašulāne, I Kleina and I Pı̄gozne (ed), Pētı̄jumi kuršu senatnē:
Rakstu krājums, Latvijas Vēstures Muzeja raksti 14 — Arheolog‘ ija, Latvijas Nacionālais
Vēstures Muzejs, Riga, 153–165
Vassar, A, 1939 ‘Iru linnapära’, in H Moora (ed), Muistse Eesti linnused. 1936–1938 a. kaevamiste
tulemused, Õpetatud Eesti Selts, Tartu, 53–100
Vassar, A and Tarvel, E, 1975 ‘Die ostbaltischen Stämme im Kampf gegen die deutch-
skandinavische Aggression im 12.–13. Jahrhundert’, Eesti NSV Teaduste Akadeemia
Toimetised, Ühiskonnateadused 1, 26–39
MAM 37: 04-magi – Press

bound for the eastern baltic 61

Welinder, S, 2009, Sveriges historia 13000 f. Kr.–600 e.Kr, Norstedts, Stockholm

Westholm, G, 1985 ‘The settlement at Vi, at the foot of the cliff’, in S-O Lindquist (ed), Society and
Trade in the Baltic During the Viking Age, Acta Visbyensia VII, Gotlands fornsal, Visby,
293–304
Zariņa, A, 2006 Salaspils Laukskolas kapulauks 10.–13. gadsimts, Latvijas Vēstures Institūta
Apgāds, Riga
Zariņa, G, 2006 ‘Salaspils Laukskolas 10.–13. gs. Iedzı̄āvotaju paleodemogrāfija’, in A Zariņa (ed),
Salaspils Laukskolas kapulauks 10.–13. gadsimts, Latvijas Vēstures Institūta Apgāds, Riga,
445–463
Zemı̄tis, G, 2007 ‘10th–12th century Daugmale: the earliest urban settlement along the Lower
Daugava and forerunner of Riga’, in U Fransson, M Svedin, S Bergerbrant and F Androshchuk
(ed), Cultural Interaction Between East and West: Archaeology, Artefacts and Human
Contacts in Northern Europe, Stockholm University, Stockholm, 279–284
Žulkus, V, 1991 ‘Die Prussen und ihre Nachbarn im 1. Jahrtausend nach Christi Geburt’, in V
Žulkus, H-H Timmann and W Timmann (ed), Prussen, Kuren und Masuren: Drei Beiträge
zur Landeskunde ostpreussens, Landsmannschaft Ostpreussen, Weissenburg, 4–14
Žulkus, V, 2000 ‘Die Völkerwanderung und die Westbalten: Die Entstehung der Kuren’,
Archaeologia Baltica 4, 89–108
MAM 37: 05-carlsson – Press

chapter 5
BETWEEN EAST AND WEST:
ECONOMY AND SOCIETY ON THE
ISLAND OF GOTLAND

By Dan Carlsson

In July 1999, more or less by chance, the largest silver hoard known from the Viking
world was found in a cultivated field at the Spillings farm in the northern part of the
island of Gotland, Sweden (Figure 5.1). More precisely, two deposits were discovered
some 3m apart. The two parts (27kg and 40kg respectively) contained some 67kg of
silver objects in total. Both hoards had been deposited in connection with a Viking
Age house. No evidence of containers could be found, but the shape of the deposits in
the ground suggested that the silver objects were contained in bags of leather, textile
or skin. In one of the deposits, there were the remains of a small wooden box, some
17618cm, in which the coin element of the hoard was contained (Östergren 2009).
When excavating the silver a third hoard was discoverer just 2m away. It was a
deposit of bronze objects, including approximately 20kg of mostly pieces of jewellery
from the Baltic area. There were thus three hoards in the same building, and signs
that there had even been a fourth deposit, of unknown content, that must have been
removed at some time in the past. The building itself was not a living house, but
probably a barn or storehouse (Widerström 2009).
In all, the silver deposits contained some 14,300 coins (c17kg), with few excep-
tions, all Arabic. At the time of writing, only 5000 coins have been closely examined,
but they give a good indication of the total contents of the hoards. One complication
during their study has been that most of the coins were highly fragmented. The oldest
is from ad 539 and the youngest from ad 870/871. Besides the Arabic coins, there
were four coins minted in Hedeby, an important trading town in southern Denmark,
today in northern Germany. Among the coins, there was also one from around
ad 830, minted for the Emperor Theophilus at Constantinople (Östergren 2009).
Besides these normal coins, typical of Viking Age hoards on Gotland, there was one
exceptional example with the Arabic inscription ‘Musa rasul Allah’, which translates
as ‘Moses is the envoy of God’. The inscription is probably Jewish, and is a direct
parallel to the words Mohammed is the envoy of God. The coin is a counterfeit,
minted by someone who wanted to draw an advantage from the reputation of official
MAM 37: 05-carlsson – Press

between east and west 63

Figure 5.1 Part of the silver hoard from Spillings (D Carlsson)

Arabic coins. Who this could possibly have been is not entirely clear, but there is
much in favour of it having been the king of the Khazars, living around the mouth of
the River Volga, where it enters into the Caspian Sea (Rispling 2004). In addition to
the coins there were 486 arm rings, 25 finger rings, 34 ingots and a very large number
MAM 37: 05-carlsson – Press

64 maritime societies of the viking and medieval world

of spiral rings and broken pieces of raw silver. Many of the thin spiral rings were
bundled together to a specific weight, and there are clear signs that this material
should be seen as a huge deposit of payment silver for use in the bullion economy of
Viking Age Scandinavia.
The bronze hoard had been deposited in a wooden chest made of spruce with a
heavy padlock of iron. The objects consist of neck rings, arm rings, finger rings and
dress pins, all made of bronze. Many were damaged and melted. Most of the objects
have a clear eastern origin, from modern day Estonia, Latvia and Lithuania, and can
be broadly dated to between ad 700 and ad 1000–1100 (Huttu 2004).
The hoards discovered at Spillings are just a few of many found on Gotland
during the last 100 years (Figure 5.2). Up to the present, approximately 700 examples
from the Viking Age and early Middle Ages are known, dated between ad 800 and
1140. The end of the period of hoarding coincides with the first minting of coins on
Gotland. From then on silver by weight no longer functioned as currency in trading
(Myrberg 2010; Östergren 2004). Every year at least one new silver hoard is dis-
covered. They are found all over the island, except in rocky areas where there have
never been settlements. In some cases there are several hoards from the same farm,
suggesting that there is a hoard for each generation of occupants. The evidence sug-
gests that every farm on the island of Gotland in the Viking Age had at least one silver
hoard, probably more. This wealth of silver raises many questions regarding the
meaning or purpose of the hoards and what they represented in the Viking Age
society of Gotland. The tremendous number is obviously a result of successful com-
mercial activity by the Gotlanders. Fascinatingly, it seems as if every farm on the
island took part in trading in one way or another. One prerequisite for this activity
must have been a stable society. It seems inescapable that the Gotlanders made
treaties with the surrounding world to make trading safe enough to be prosperous.
There are also no overt signs of internal political conflict. Gotlandic society was built
on free farmers, who met every year at the Althing to settle matters in ways similar to
what took place in Iceland (Byock 2002). There are no signs of a king or aristocracy
on Gotland, though some farmers were clearly wealthier than others.
Being the middleman in the case of Gotland was to be the connecting link
between the Arabic area and the western Viking world, meaning Sweden, Denmark
and Norway. Gotland was a natural meeting place and stepping stone in the long-
distance trade between east and west, a position it maintained into the Middle Ages
and the Hanseatic world. Thus, the Gotlanders acquired a remarkable richness that
seems to have been shared by all farms. The basis of this wealth was ultimately the
location of Gotland in the middle of the Baltic Sea, an advantageous position that had
been realized long before the Viking era, with the creation of a well-functioning
trading system based on agreements with surrounding areas and rulers.
Gotland’s position as the middleman between east and west was probably built
up, not only through transit trading, but also by craft working and otherwise adding
value to raw materials and intermediate goods. There is for instance clear evidence of
comprehensive metalworking, not only of Gotlandic artefacts, but also of typical
female brooches found on the mainland of Sweden, as well as in Norway (Gustafsson
MAM 37: 05-carlsson – Press

between east and west 65

Figure 5.2 More than 700 silver and gold hoards have been found on Gotland, and every year
at least one new hoard is discovered. They are evenly distributed over the island, except for the
rocky area crossing the island in the middle from west to east (D Carlsson)

2013). This kind of lucrative trade practice is normally associated with the presence
of a strong monarchy or other central authority. What is remarkable in the case
of Gotland is that there is no trace of a kingdom in the traditional sense, nor of a
nobility. Power on Gotland was based among independent land-holders as far as
MAM 37: 05-carlsson – Press

66 maritime societies of the viking and medieval world

we can judge. That said, some farmers were wealthier than other farmers, and indi-
viduals must have held influential positions within the Gotlandic legal system.
If each Viking Age farm in Gotland took part directly or indirectly in the exten-
sive trade and exchange with the outside world (and thus also had a place at one of
the many trading harbours), how this society was organized remains an intriguing
question. Nevertheless, the administration of the island is partly understood. In the
Viking Age and early Middle Ages Gotland’s farms were divided amongst things
(assemblies) grouped into three part-things and one Althing (Gutnaltinget), at Roma
situated in the middle of the island (Östergren 1992). In total there were 20 things or
districts on the island, led by judges (seniors) presumably designated by the people.
The Althing was the centre of Gotlandic administration, but also probably the central
and most important place of worship during pre-Christian times. The situation is
largely comparable with Iceland, where Thingvellir was the administrative and reli-
gious centre. It is thus hardly surprising that a Cistercian monastery came to be built
in Roma around the years 1152–1164, probably at the invitation of the Althing
(Östergren 1992).
There is no historical or archaeological evidence of a king of Gotland, and
foreign powers such as kings and bishops (in the Christian period) seem to have had
little influence on the island’s daily activities. It is clearly evident from Guta Law,
written down sometime in the early 13th century, that the tithe on the island, which
fell to the Bishop of Linköping on the mainland, was split into three parts and divided
between the parish priest, the church and the poor (Holmbäck and Wessén 1979).
The last part the farmers kept for themselves so that they themselves took care of
poor relief. The freeholder peasants were the dominant class, but in addition the
Guta Law indicates that there was probably a significant element of slavery. For
example, the code stipulated that ‘if someone’s slave . . . will be found working on a
holiday, then the master fined them three Örar, and they will work three years after
the time of slavery is completed’ (Holmbäck and Wessén 1979, 209). The adminis-
trative organization of Gotland was different from the Swedish mainland in many
ways and Gotlandic society also differed from other parts of Scandinavia, probably
as a consequence of its island setting. Despite extremely lively contact with the out-
side world, and patterns of activity that apparently included settlements from the
westernmost limits of Europe to the Black Sea, Gotlandic culture is in many respects
unique, although aspects of society such as religion and art showed some common-
alities with the Viking world in general.
Women’s jewellery provides a clear example. Female Gotlanders consistently
used jewellery sets that do not occur anywhere else. In general, from Iceland in the
west to Finland in the east, women wore oval brooches in pairs. This was not so on
Gotland, where pairs of animal-head brooches (Figure 5.3) were used instead,
between which was worn a large round box brooch. Other forms of jewellery worn
on Gotland also have no counterparts elsewhere in the Viking world, such as tongue-
shaped pendants, fish-shaped pendants and sets of jewellery with lenses of rock
crystal (Carlsson 1983; Thunmark-Nylén 1983a; 1983b). It is important to note that
this difference in jewellery traditions between Gotland and other parts of the Baltic
Sea area is not only limited to the Viking Age. Long before the 9th century the island
MAM 37: 05-carlsson – Press

between east and west 67

Figure 5.3 Animal-head brooches are typical Gotlandic female jewellery, only used on
the island (D Carlsson)

had its own character, with many jewellery forms that are characteristic of Viking
Age Gotland having antecedents earlier in the Iron Age.
It is obvious that Gotlanders, despite their intensive contacts with the outside
world, created and maintained their own identity to a much greater extent than other
areas. It is illuminating that the situation on Gotland stands in sharp contrast to that
on the nearby island of Öland. Both islands have the same bedrock (limestone) and
natural landscape, but exhibit significantly different historical developments in terms
of buildings, burial customs, land use and social structures. Öland, with large regu-
lated villages, an active nobility and a powerful church, had a Viking Age and medi-
eval culture that was very similar to the rest of Sweden, not least in the immediate
area around the medieval town of Kalmar (Fallgren 2006). The physical distance
from Öland to the mainland is short, however, and has never been a great physical
barrier. Conversely, an island like Gotland, isolated from surrounding areas and
spheres of power, could create its own culture, despite extensive and lively contacts
with the outside world.

references
Byock, J, 2002 ‘The Icelandic Althing: dawn of parliamentary democracy’, in J M Fladmark (ed),
Heritage and Identity: Shaping the Nations of the North, The Heyerdahl Institute and Robert
Gordon University, Donhead St Mary, Shaftesbury, 1–18
Carlsson, A, 1983 Djurhuvudformiga spännen och gotländsk vikingatid, Stockholm Universitet,
Stockholm
MAM 37: 05-carlsson – Press

68 maritime societies of the viking and medieval world

Fallgren, J-H, 2006 Kontinuitet och förändring (Continuity and Change): Bebyggelse och samhälle
på Öland 200–1300 e Kr. (Settlement and Society on Öland 200–1300 ad ), Uppsala Univer-
sitet, Uppsala
Gustafsson, N B, 2013 Casting Identities in Central Seclusion: Aspects of Non-ferrous Metal-
working and Society on Gotland in the Early Medieval Period, Stockholm University,
Stockholm
Holmbäck, Å and Wessén, E (ed), 1979 Skånelagen och Gutalagen, Svenska landskapslagar, fjärde
serien, Gebers, Uppsala
Huttu, J, 2004 ‘Bronsdepån i Spillings — skrotförråd eller offer’, Gotländskt Arkiv 2004, 133–141
Rispling, G, 2004 ‘Spännande mynt i Spillingsskatten. I; Gotland Vikingaön’, Gotländskt Arkiv
2004, 123–132
Thunmark-Nylén, L, 1983a ‘Dräkt och dräktsmycken på Gotland under vikingatiden’, in I Jansson
(ed), Gutar och vikingar, Statens Historiska Museum, Stockholm, 153–182
Thunmark-Nylén, L, 1983b Vikingatida dosspännen — teknisk stratigrafi och verkstadsgrup-
pering, Uppsala University, Uppsala
Widerström, P, 2009 ‘Spillings farm: the home of a rich Viking in the northeast of Gotland’, in
A-M Pettersson (ed), The Spillings Hoard: Gotland’s Role in Viking Age World Trade,
Gotlands Museum Fornsalens Förlag, Visby, 41–64
Östergren, M, 1992 ‘Det gotländska alltinget och cistercienserklostret i Roma’, Gotländskt Arkiv
1992, 49–58
Östergren, M, 2004 ‘Guld- och silverskatter’, Gotländskt Arkiv 2004, 107–113
Östergren, M, 2009 ‘Spillings: the largest Viking Age silver hoard in the world’, in A-M Pettersson
(ed), The Spillings Hoard: Gotland’s Role in Viking Age World Trade, Gotlands Museum
Fornsalens Förlag, Visby, 11–40
MAM 37: 06-naum – Press

chapter 6
VIKING AGE BORNHOLM:
AN ISLAND ON THE CROSSWAYS

By Magdalena Naum

Bornholm is a small Baltic Sea island located between the coasts of Scania (southern
Sweden), Denmark, Germany and Poland. In the Viking Age and Middle Ages its
convenient location on waterways connecting Scandinavia with the southern coasts
of the Baltic Sea made the island a well-known place among those traversing the sea
with cargos of merchandise or heading for adventure (Figure 6.1). Despite its strategic
position, however, contemporary writers and travellers had very little to say about
the island. One of the earliest pieces of information comes from Wulfstan, who
around the year ad 880 was sent by the Anglo-Saxon king Alfred the Great to explore
the Baltic Sea. When describing his journey from Hedeby to Truso, Wulfstan notes
that Bornholm, or Burenda-land as he calls the island, has a king of its own (Brink
2007, 69; Lund and Fell 1984, 23). Wulfstan’s remark about the island’s self-rule, as
well as the longevity of Bornholm’s independence, continues to spark debate among
archaeologists and historians (eg Lihammer 2007, 261–262; Nielsen 1998; Randsborg
1980, 163; Wienberg 1986).
Another interesting historical reference regarding Bornholm’s position and
political status comes from the pages of Gesta Hammaburgensis Ecclesiae Pontificum
by Adam of Bremen. In its description of Scandinavia, Adam notes that Bornholm
‘is the most celebrated port of Denmark and a safe anchorage for the ships that are
usually dispatched to the barbarians and to Greece’ (Tschan 1959, 197). He also
describes an act of official conversion of the island. In Adam’s words Egino, the arch-
bishop of Lund, won the islanders to Christ, moving them to tears with his preaching,
and made them destroy their idols and recognize their errors, which they did imme-
diately and without hesitation (Tschan 1959, 192). Although this description is no
doubt a rhetorical overstatement, it suggests that at least from the second half of the
11th century the process of strengthening ties with the Danish kingdom and institu-
tionalized Church was taking place on the island.
Written sources concerned with Viking Age and early medieval Bornholm are
very few and fragmentary. Until 30 years ago archaeology had little to add to this
enigmatic picture, other than to confirm the importance of trade in island life based
on abundant silver hoards. In recent decades, however, the situation changed dramat-
ically. Since the 1980s, when the private use of metal detectors became legalized,
MAM 37: 06-naum – Press

70 maritime societies of the viking and medieval world

Figure 6.1 Viking Age geography of the western basin of the Baltic Sea with some of the
trading routes (after Hermann 1997, fig 1)

archaeology on the island experienced unprecedented development. The number of

mapped farms and sites interpreted as possible settlements dated to the Viking Age
and early Middle Ages increased to over 300 (Figure 6.2) (about 100 of them are
securely dated to the period ad 800–1000: Aarsleff 2005, 34; Nielsen 1994). The
number of excavated silver hoards, which reached over 50 deposits, and over 300
finds of spherical and octahedral balance weights, confirmed the involvement of the
islanders in Baltic Sea trade (Figure 6.3).
Some of the most exciting discoveries were those made in the southern part of
Bornholm. Between 1982 and 1999 three cemeteries dated to the end of the Viking
Age and the early Middle Ages were excavated there (Wagnkilde 1999; 2000; 2001;
Wagnkilde and Pind 1996; Watt 1985). The findings made on a small burial ground,
which was most likely used by a single extended family living at the farm of Rune-
gård, and the excavation of much bigger communal cemeteries at Ndr. Grødbygård
and Munkegård, provided clues about Christianization of the island and possible
MAM 37: 06-naum – Press

viking age bornholm 71

Figure 6.2 Map of Bornholm showing settlement sites of the Viking Age and early Middle
Ages, silver hoards and strongholds in Lilleborg and Gamleborg (after F O Nielsen)

tensions caused by it. Furthermore, they revealed the distinctive character of 11th-
century burial customs on the island.
The results of fieldwork conducted at these cemeteries and settlement sites
provide insights into the changing political, economic and social situation on Born-
holm. Above all, they reveal the inclusive character of the local culture and the
MAM 37: 06-naum – Press

72 maritime societies of the viking and medieval world

Figure 6.3 Distribution of Viking Age and early medieval weights used in trade
(after Aarsleff 2005, fig 32)

human mobility within the island and to the island from abroad (Price et al 2012).
Some of the objects deposited in both the graves and in the remains of Viking Age
houses originated from places located far from the island. Owned or made by indi-
viduals or families that moved to Bornholm from abroad, or brought to the island by
frequent travellers and merchants, these artefacts were incorporated into the local
culture either as fashionable trinkets or as new elements of local tradition. Biog-
raphies of these objects can serve as metaphors for the connections of the islanders
with the outer world. Furthermore, their stories can serve as an original way to
MAM 37: 06-naum – Press

viking age bornholm 73

enlighten the particular history of Viking Age Bornholm, an island positioned on the
crossways of the Baltic Sea.

becoming a part of the danish kingdom

At the end of the Viking Age Bornholm was most likely incorporated, at least
nominally, into the Danish kingdom. But the road to bringing the island into the
structure of the realm was neither short nor easy. The composition of the late Viking
Age hoards, runic inscriptions and the character of early Christian burial practised by
the islanders all suggest a rather complicated and uneasy relationship between the
island and the Danish mainland.
Until the beginning of the Viking Age the political landscape of Bornholm
was dominated by Sorte Muld. This centre, situated in the north-eastern corner of
the island, consisted of as many as 30 farms, covering an area of approximately 60–
70,000 square metres. The special character of this place, which had to be inhabited
by an exceptionally large cluster of people, is even more evident when one considers
that the preferred form of settlement on the island was single farms or small hamlets
(Hamerow 2004, 157–169). Sorte Muld resembles other Iron Age central places, like
Gudme, Uppåkra, Lejre and Tissø in terms of its chronology, settlement and material
culture (Hårdh 2003; Jørgensen 1998; Watt 1991). All of these sites were located near
the sea and all were extremely rich in metal objects and foreign artefacts. They were
most likely strategic places for trade, redistribution and long-distance contacts. How-
ever, unlike other places, which continued to hold their central position into the
Viking Age, the dominating character of Sorte Muld ceased sometime towards the
end of the 8th century (although the surrounding landscape continued to be settled in
the Viking Age and Middle Ages). A similar process of ‘decentralization’ of power
and wealth, and abandonment of villages and concentrated settlements in favour of
smaller settlement units, is observed in cases of other important Iron Age centres on
the island, such as Rødbjerg and Møllegård located in the western part of the island
and Smørrenge in the south.
Radical changes are also visible in the Late Iron Age/early Viking Age burial
ritual on the island. Starting in the 8th century the earlier custom of depositing rich
finds of weaponry and horse-riding gear in male graves disappears and the repertoire
of grave goods is more or less standardized. A similar shift in funerary ritual takes
place in cases of female graves, where the choice of funeral garments and grave goods
is made from limited and commonly used objects, such as oval brooches, beads and
knives (Jørgensen 1987, 43–44; Svanberg 2003, 126–130). Lars Jørgensen attributes
this considerable change to developments in the social sphere: in the 8th century the
rules determining social status, inheritance and the ‘relative ranks of individual
families/farms seem to have become clarified’ (Jørgensen 1987, 44).
The dissolution of large Iron Age cemeteries and the abandonment of larger
settlement units might be a reflection of an increased ‘democratization’ of island
society, and a firm establishment of individual ownership and property rights. This
process led to the establishment of a network of single farms or units of up to three
MAM 37: 06-naum – Press

74 maritime societies of the viking and medieval world

farms, both in the vicinity of the old Iron Age settlements and away from them, in
previously unused areas (Aarsleff 2005; Jørgensen 1987, 51).
The archaeological material recovered on the 9th- and 10th-century farms is
more consistent and standardized. Surprisingly, many of the recognized Viking Age
settlements contain finds of weights and treasure hoards with Islamic and Western
coins, which may suggest that involvement in trade and other maritime activities
became a common practice for Bornholm families (Figures 6.2 and 6.3). Similar ten-
dencies mark contemporary developments on other Baltic Sea islands, like Gotland
and Öland (see Chapters 3 and 5).
Although the island seems to be evenly settled, the areas along the Grødby, Læs,
Præste and Bagge rivers may have had more concentrated settlement. These farms,
particularly those located along the southern coast, produced a large number of silver
hoards (Figure 6.3).
Additional and possibly even more dramatic changes in the settlement structure
took place in the 11th century, involving the establishment of new farms (partly on
heavier clay soils) and the restructuring of older ones (Nielsen 1994). These new
farms, called by Bornholm archaeologists ‘Østersø-bopladser’ (Baltic Sea settlements)
distinguish themselves by new elements of material culture, such as pottery of Baltic
ware and the common occurrence of imported quern stones and whetstones. Some of
these farms have been excavated and the elements of material culture discovered
imply changes towards greater specialization in economic activities and an increasing
importance of agricultural production. Moreover, excavated material from numer-
ous refuse pits and the remains of a building interpreted as a smokehouse discovered
at the 11th-century settlement of Møllebjerg, Rutsker parish (north-western Born-
holm) indicates that the inhabitants were most likely involved in fish processing that
exceeded the needs of just one farm. Common finds of grindstones and the later
introduction of mills in the 12th and 13th centuries are evidence of processing grain.
These changes might have been a response not only to overall shifts in the way
resources were acquired, but also to the increasing weight of tax duties required by
the Church and the King (Thurston 1999; 2001; Wienberg 1986, 58).
The 11th century was a time of economic and cultural change in southern
Scandinavia, when land ownership was of increasingly crucial importance and the
collection of taxes and profits from agriculture, trade and non-agricultural produc-
tion became a major income for the Crown and the Church. This could have been
one of the important reasons for more active engagement in local development on
Bornholm on the part of both Danish kings and the bishops of Lund, which is docu-
mented from the second half of the 11th century onwards. According to Knytlinga
saga, at least during the reign of Cnut the Holy (ad 1080–86), the whole island was a
royal possession administered with the help of allied landlords who controlled the
landscape through a network of royal demesnes (Guðnason 1982, 151–154; Nielsen
1998, 13). In the ad 1060s the island was visited by the bishop of Lund, Egino, who,
according to the contemporary chronicler Adam of Bremen, officially baptized the
inhabitants (see above). Thus the Church organization of Bornholm was probably
incorporated into the structure of the episcopal see of Lund. Soon there was also
major royal engagement in the rebuilding of the stronghold in Gamleborg in the
MAM 37: 06-naum – Press

viking age bornholm 75

forest of Almindingen in the heart of the island. A few decades later, around ad 1150,
a smaller stronghold of Lilleborg was built only 700m away from Gamleborg.
Not all islanders, however, seemed to approve of the expanding power claims of
the Danish kings. The attempts to impose novel administrative divisions and eco-
nomic requirements, following the introduction of changes in the social structure and
attempts to alter the balance of power, were met with resistance and ignited conflicts
within Bornholm society. Archaeological sources can provide interesting insights into
these social and political processes.
The settlement sites dated to the Viking Age and early Middle Ages are often
mapped due to the silver hoards deposited underneath house corners or walls or at
the boundaries of individual properties. These hoards are an interesting source
material revealing more than the whereabouts of the islanders and their mercantile
and/or piratical activities in the Baltic Sea. They are also silent clues regarding a
struggle against integration with the Danish kingdom and about the continuation of
independent economic relations well into the Middle Ages.
The structure of the 10th- and 11th-century hoards is very interesting. Bornholm
has the highest concentration of these finds in southern Scandinavia, although the
average weight of a deposit is lower than elsewhere. Furthermore, the composition of
the hoards differs considerably from those found in other southern Scandinavian
territories (von Heijne 2004). The 10th-century hoards from Bornholm lack
Carolingian coins and have very few early Scandinavian pennies. Instead, Islamic
and German coins and hack silver are the dominant components of the hoards with
some deposits containing specific issues of English coins (see below). The distinctive
character of Bornholm deposits continued in the 11th century. German coins domi-
nate strongly and besides Otto-Adelheid coins, which are commonly found every-
where in southern Scandinavia, coins minted in Franconia, Bavaria, Lotharingia and
Frisia made their way to the island. Unlike in other territories, which by the begin-
ning of the 11th century were regarded as parts of the Danish kingdom, coins from
Danish mints are almost always in the minority. Their number is only significant in
the deposits buried around the year ad 1050, falling to low numbers again towards
the end of the 11th century. Throughout the Viking Age and early Middle Ages
German coins continued to dominate in the contents of the deposits (von Heijne
2004).
The major differences in the structure and composition of Bornholm’s deposits
vis-à-vis hoards buried in Scania, Zealand, Fyn or Jutland might indicate rather
loose economic relationships between the island and the central Danish areas. This
stubborn retention of an independent network of economic relationships might be
a sign of the island’s poor integration into the Danish kingdom. It is interesting to
note, however, that the number of Danish coins rises briefly in the middle of the
11th century, concurrent with Egino’s visit and possibly also with reorganization of
the rural landscape by establishing royal demesnes. The influx of Danish coins, fol-
lowed by their decrease, might be a very tangible indication of the active involvement
of the kings and the Church in island affairs in the middle of the 11th century and, in
the long run, an unsuccessful attempt to change traditional economic contact, which
gravitated towards the German and Slavic coasts of the Baltic Sea.
MAM 37: 06-naum – Press

76 maritime societies of the viking and medieval world

The 11th century was also a period when rune stones were carved and raised
throughout the island. Careful studies of the inscriptions and the patterns of com-
memoration are yet another way to enlighten political and social transformation in
the dusk of the Viking Age. Birgit Sawyer noted that the custom of erecting these
stones was a crisis symptom, a response to a social and economic change caused by
religious and political developments in 10th- and 11th-century Scandinavia (Sawyer
2000, 77). She argues that change was particularly felt in places where the number
and concentration of stones was highest. This would make Bornholm, with its
41 monuments, a place with elevated social and political tensions.
Recently, Anna Lihammer has pointed out yet another interesting feature of the
rune stones. The pattern of commemoration inscribed on these monuments is not
uniform. Some were raised by the family and commemorate a member of one’s kin.
Others were made in memory of individuals unrelated by blood. She hypothesizes
that the first type of relation (vertical or kin) was a statement of inheritance and
family history, and a way of reacting against new political and social developments,
while the horizontal relations (not related by blood) were indications of networks
linked to power coming from outside (Lihammer 2007, 176–177). Interestingly,
among the Bornholm rune stones with readable inscriptions, at least four were raised
to men commemorated by other men without referring to their kin relations, and on
another three stones the relationship between the sponsor and the person commemo-
rated is uncertain (Sawyer 2000, 205). One stone with such a horizontal relationship
commemorates a thegn, ie a person bearing a title linking him to royal power. Con-
sidering the political background of the middle of the 11th century, these divergent
patterns of inscriptions, and use of the titles thegn and dreng on four different stones,
might indicate a split in the Bornholm community.
The titles of thegn and dreng are closely connected with the rule of Sven
Forkbeard and Cnut the Great. They are associated with the rise of a new social class
in Denmark and the formation of strategic alliances between the royal court and
ambitious noble families (see below). These alliances were based on land endow-
ments and privileges granted by the king in exchange for loyalty. They may have been
formed to assure greater integration of various landscapes under Danish rule and to
undermine the traditional local power structure. This undoubtedly led to conflict,
reflected in the custom of raising rune stones and perhaps also in the curious funerary
rituals observed in late Viking Age and early medieval cemeteries.
The network between some Bornholm families and the Crown most likely
resulted in individual conversions, leading to the situation where both traditions —
pre-Christian and Christian —– were simultaneously adhered to by different inhabi-
tants of the island. Christian elements are visible in funerary rituals in graveyards
at Slamrebjerg (Bodilsker parish), Runegård and Ndr. Grødbygård (both in Åker
parish), and in Munkegård (Poulsker parish) dated from the second half of the
ad 900s to the end of the 11th century (Naum 2007; 2008, 179–253; 2009; Wagnkilde
1999; 2000; 2001; Wagnkilde and Pind 1996; Watt 1985). The rituals practised by
the inhabitants of southern Bornholm, where all these excavated burial places are
located, tell an intriguing story of persistence and change in ideological ideas sur-
rounding death.
MAM 37: 06-naum – Press

viking age bornholm 77

Runegård and Slamrebjerg are small burial sites, probably used by single house-
holds. Munkegård and Ndr. Grødbygård, on the other hand, represent large com-
munal cemeteries divided into separate sections for women (to the north) and men (to
the south). All cemeteries clearly exhibit features associated with Christian ceremony,
such as the orientation of burials along an east–west axis, use of coffins and the
emergence of a custom of swaddling the deceased in shrouds. What is striking, how-
ever, is the amount of objects that are found deposited with some of the deceased and
the lingering custom of equipping the deceased for the after-life (Naum 2007, 24–25;
2008, 235–242). Many of the dead are not only buried with jewellery, but also offered
grave goods such as pottery, tools, silver hoards and amulets (Figure 6.4). Further-
more, at the cemetery of Ndr. Grødbygård, in the spaces between the graves as well as
in the top layers of the grave fill, a considerable number of pits containing charcoal,
burnt animal bones and potsherds were found. Some of these pits are hard to date,
but others contained sherds of late Viking Age/early medieval Baltic ware suggest-
ing that they were contemporary with the burials. These pits, and large amounts of

Figure 6.4 One of the burials excavated at Ndr. Grødbygård cemetery. This individual
(presumably a woman) was wearing a beaded necklace consisting of two glass beads, a silver bead
and a silver amulet. She was also given a knife and a pot (courtesy of Bornholms Museum)
MAM 37: 06-naum – Press

78 maritime societies of the viking and medieval world

animal bone also found at the cemetery, could serve as evidence of ceremonial meals
during funerary and commemoration rituals.
The persistence of certain pre-Christian traditions in early medieval funerary
rituals may indicate the relatively late introduction of Christianity to the island, or
lack of its full embracement. Conversely, the particular features of ritual might
indicate that funerals became a stage for silent battles between the old and the new
order. The fact that burials had to be made in communal cemeteries organized in line
with outside ideas, perhaps reinforced by a new elite acting for King and Church, was
contested by the lavishness of multi-stage ceremonies that referred to older traditions.
Another phenomenon that analysis of the Bornholm cemeteries reveals is par-
ticularity in selection of the objects given to some of the deceased. There is a group of
burials where the choice of objects, and the way they were treated and placed inside
the graves, comes somewhat as a surprise. Rather than relating to earlier funerary
rites practised on Bornholm they show parallels with the rituals performed by
Western Slavs in their homelands south of the Baltic Sea (Naum 2007, 23–4, 30–33;
2008, 227–232; 2009, 76–79). Although the isotopic analyses conducted on the bones
from the cemetery neither excluded nor confirmed the burials of the first generation
of immigrants from the southern coast of the Baltic Sea (due to the partial overlap
in strontium values), the ritual and material culture could be indicative of immigra-
tion (Price et al 2012).
The most visible elements of this foreign culture are amulets, female jewellery
and a particular way of depositing pottery (namely the custom of breaking a pot and
placing a single sherd or a small selection of pieces next to the body). While these
practices and objects, together with similar material found in settlement sites (see
below), strongly suggest Slavic immigration to Bornholm, the use of some objects
seems to indicate their simultaneous existence (but perhaps different meaning) in the
cultural worlds of newcomers and the locals. This duality can be illustrated by the use
of silver beads in ritual contexts. In Munkegård and Ndr. Grødbygård the burials
where individuals were given silver beads constitute roughly 13% of all equipped
graves, making it a moderately common part of the deceased’s dress. These beads are
of Western Slavic origins. In the areas they came from they were used as female
ornaments, as necklaces, trinkets attached to headbands or parts of earrings. They
were thus an element of embodied norms related to womanhood, used in everyday
contexts and in the construction of personhood in death. This aspect of display, and
definition of the female body, continued to play an important role after immigration
to Bornholm.
Conversely, silver beads might have had a different symbolic and biographical
meaning for some members of the local community. Fragmented silver (including
beads) served as ‘legal tender’ circulating between tradesmen and those involved in
exchange. As such it is a frequent component of silver hoards deposited on the island.
Some of the individuals buried at Ndr. Grødbygård might have been Bornholmers
who previously participated in trade or had ties of kinship with those who travelled
overseas, and thus had access to silver. These beads would thus be associated with the
biography of the deceased and their families in ways that differed from those buried
with Slavic women.
MAM 37: 06-naum – Press

viking age bornholm 79

facing the south

The traces of ritual and grave goods closely related with the Western Slavic
culture are not the only finds on the island that suggest close connections between its
inhabitants and the southern coast of the Baltic Sea. One of the most common objects
encountered by archaeologists excavating Viking Age and medieval sites in southern
Scandinavia is pottery. Viking Age pottery was handmade, simple, produced at home
for the individual needs of a household. Ceramics on Bornholm were similar to those
used in other Danish islands or in Scania, and the tradition of making these crude but
functional vessels stretched a couple of centuries back in time. However, this long
tradition of making simple undecorated pots was broken in the 11th century. Starting
around the year ad 1000, on Bornholm and in Danish territories oriented towards the
Baltic Sea, a new ceramic style was introduced, called Baltic ware. Having techno-
logical and stylistic roots in Slavic pottery making, these vessels were made with the
use of a turntable, decorated with a plethora of motifs and fired in a more controlled
environment. They very quickly found a place on the tables of local households. On
Bornholm this change is particularly striking in the case of newly established farms,
the so-called ‘Østersø-bopladser’ (Baltic Sea settlements) that emerged in the course
of the 11th century. The earlier tradition of simple handmade vessels might initially
have been sustained in the farms that were continuously used since the Viking Age,
although it seems that they were not made to any considerable extent after the second
half of the 11th century.
The origins of this new pottery tradition, the circumstances of its introduction
and its swift incorporation into local culture has puzzled scholars ever since the 1920s
when the Swedish archaeologist Georg Karlin observed similarities between pottery
found in the early medieval layers of Lund, early 11th-century pottery from the Slavic
areas east of the Elbe and late 11th- to 12th-century vessels from the greater Oder
estuary (Karlin 1923). Archaeologists following his footsteps first assumed that
the pots were imports from the Slavic areas. Later detailed typological studies of the
Baltic ware conducted in Denmark and Sweden, as well as the sheer volume of
the pottery unearthed in both rural and urban contexts, led to the alternative con-
clusion that Slavic potters introduced local production of Baltic ware in Scandinavia.
It was assumed that they emigrated, or were brought to Scandinavia as enslaved cap-
tives, from war-torn areas south-west of the Baltic Sea (Gebers 1980; Liebgott 1979;
Roslund 2007). Furthermore, it was convincingly shown that Baltic ware was sub-
sequently subject to local stylistic development diverging from traditions in the Slavic
areas. The technological and stylistic solutions typical of Slavic pottery were adopted
by the local potters and developed into new forms departing from the original
patterns and from the pottery tradition in the Slavic territory (Roslund 2006, 66–68;
2007, 473, 497–500).
On Bornholm, the Baltic ware closely resembles the pottery used in the western
parts of Western Pomerania and in eastern areas of Mecklenburg-Vorpommern, ie
territory of Liutizi settlement or influence (Figure 6.5) (Naum 2008, 126–131). Hence
the potters that introduced this novel form of ceramics came most likely from this
area. Interestingly, the sherds of Baltic ware that were discovered on the island do not
MAM 37: 06-naum – Press

80 maritime societies of the viking and medieval world

Figure 6.5 Late Slavic pottery from sites located around the Oder estuary and Szczecin
Lagoon (left: after Białe˛cka 1961; Cnotliwy and Rogosz 1982, 110, 124; Herrmann and Donat
1979, 44; Leciejewicz et al 1982, 47, 51) and examples of Baltic ware pottery from Bornholm
(right: M Naum)

show equally clear departure from the original forms and designs. At some farms,
rather the opposite tendency is observed of continuous reuse of forms and decorative
patterns. This could be explained by particularity of local tastes and needs, or seen as
an expression of a discourse surrounding transmission of the potting skill from one
generation to another stressing the necessity to continue along the same stylistic lines
(Naum 2012).
Baltic ware is an interesting object for many reasons. It is an illustrative example
of the process of incorporation of an initially foreign object into an element of a
local culture. It is also an object through which a story of interaction between the
Bornholmers and their southern neighbours can be told.
Before Baltic ware was introduced on the island, single examples of early Slavic
pottery were brought to Bornholm. Some sherds of the 9th- and 10th-century pottery
of Feldberg and Fresendorf type were found in the settlements of Store Gadegård,
Dammegård/Præstegård and Lille Myregård, all located along Bornholm’s southern
coast. The remains of these Viking Age farms hid an important clue to unmasking
the occupation of their inhabitants. All of them contained silver hoards consisting
of, among other things, hack silver of Slavic origin or loose silver pieces of Slavic
jewellery, suggesting that the inhabitants of these settlements participated in trade
with Slavic merchants and possibly visited markets located in the Slavic areas.
This would not be the only indication of mercantile interactions between both
territories. Numismatists and archaeologists observed far-reaching similarities in the
level of fragmentation and the structure of the contents of the hoards deposited on
Bornholm and in the Slavic coastal areas (Hårdh 1996; Skovmand 1942; von Heijne
2004, 24) and assumed that the development of large trading places, such as Wolin,
located in the Oder estuary, had a considerable impact on the routing of the exchange
MAM 37: 06-naum – Press

viking age bornholm 81

(Watt 1988). Maybe it is not a coincidence then that some of the richest and largest
Viking Age farms on the island are grouped along the southern and western shores of
Bornholm, facing major trading routes leading from Hedeby and Slavic ports to Birka
and the east.
Interactions between the territories might not always have been peaceful. From
the pages of Knytlinga saga one can learn about a wealthy Bornholmer named Egil
Ragnarsen (also called Blod-Egil). He was a pirate and gained a bad name partic-
ularly among the Slavic tradesmen who repeatedly fell prey to his attacks (Guðnason
1982, 152–163). Pillaging activities of some islanders might also be reflected in the
contents of selected silver hoards, and possibly in the settlement of enslaved Slavs on
the island.
Networks established between the Bornholmers and their Slavic partners, and a
certain level of familiarity and knowledge of the conditions in each other’s home
territories, were important in the process of Slavic immigration to the island, which
took place in the course of the 11th century (Naum 2008). The causes of migrations
might have been multiple and diverse. The political situation in the territories of
Obodriti, Veleti and the western parts of Pomerania in the 10th and 11th centuries
was marked by episodes of strife, economic pressure and instability. Contempora-
neous chroniclers and observers of these events occasionally noted that people were
fleeing the hopeless and dangerous conditions. Corollaries of war were the taking of
prisoners and the slave trade. The captives, sometimes taken to Scandinavia, were
employed as household slaves or as help in land-clearing and agricultural tasks if not
sold to foreign merchants.
As noted above, in the wake of the Middle Ages, Bornholm, like the rest of
southern Scandinavia, underwent changes in political, social and economic structure.
Land ownership, agricultural production, revenues from the market exchange of craft
objects and tax collection were increasingly sources of profit for those holding power
centrally and locally. On the island the changes meant a certain reorganization of
the landscape and social structure including the establishment of large land holdings,
royal demesnes, the establishment of a Church organization and closer relations with
the Danish Crown. These transformations constitute an important background for
understanding Slavic immigration to the island. In the shifting economic and political
conditions foreigners, either free cottagers or captives brought to the island and per-
sonally tied to different households, might have been an appreciated workforce.
The immigrants brought with them new technology, objects and customs. Not
all of these novelties, however, shared the trajectory of acceptance and assimilation
characteristic of Baltic ware. In fact, it seems that Baltic ware, and possibly a partic-
ular style of knife sheath mounting, were virtually unique in being incorporated into
local tradition. Most likely they both traversed a path of redefinition: from curious
foreign objects to recognizable ones perhaps associated with a particular group of
people — merchants and adventurers who ventured across the sea. A final stage in the
process of adoption was initiated when local production of these objects started and
their availability increased. It is very likely that once they became omnipresent and
easily accessible they started to be regarded as an inherent part of local tradition.
MAM 37: 06-naum – Press

82 maritime societies of the viking and medieval world

Some objects and practices associated with Slavic settlement on the island, such
as female temple rings (rings attached to a scarf or headband and worn by Slavic
women) and rituals accompanying burials, seem never to have crossed the cultural
lines. Others, like the silver beads discussed above, might have been incorporated into
local fashion or ritual.
The transformation of objects which changed in function and meaning as an
intertwinement of human and object biography could also be attested in the case of
other foreign objects brought to the island. Coins turned into pendants are an inter-
esting example. In the Viking Age silver and gold were considered a materialization
of magical ‘luck’ (Gurevich 1995, 159). Thus coins, as objects made of these metals,
were kept as amulets guaranteeing the luck of their possessors. Coins might have
been worn for their biographic reference as objects connected with the events one
participated in, embodying memories of visited places or personal allegiances. On
Bornholm coin-pendants are occasionally found in silver hoards. A few German and
English coins were rolled into beads and placed on the necks of individuals buried at
Ndr. Grødbygård. A pendant made of a Cnut the Great coin struck in ad 1030–35/6
was found worn by an individual buried at Munkegård cemetery.

thegns , drengs and royal farms

English coins, like the one turned into a pendant and worn by an individual
resting at the Munkegård cemetery, are a common find on Bornholm. Coins of Cnut
the Great and Ethelred come from other late Viking Age and early medieval burials.
However, the majority of English pennies come from silver hoards. In 1876 in the
vicinity of a farm at Tyskegård a collection of 34 English coins was found. The coins,
probably kept in a leather purse, were deposited under a large stone. In the 1990s
avocational archaeologists returned to this place and found an additional 48 coins
that most likely were part of the earlier hoard. The content of this hoard is strikingly
homogeneous: all of the coins were of Ethelred’s Long Cross type minted in London,
Lincoln and Winchester around the year ad 997 (von Heijne 2004, 134–135). The
Tyskegård hoard is unique due to such a limited set of coins. It is the only deposit
discovered on the island that contained exclusively English coins, with little wear and
from few mints. There are, however, a few more late 10th- and early 11th-century
hoards in which English coins amount to considerable numbers (eg Store Frigård,
Holsegård and Guldensgård). The majority of English coins found on Bornholm
overall were struck by Ethelred II (types Crux, Long Cross and Last Small Cross) and
Cnut the Great (Pointed Helmet and Quatrefoil). It seems probable that they repre-
sent personal wealth from Danegeld acquired by participants in the raids on England
organized during the reigns of Sven Forkbeard and Cnut the Great.
The earlier expeditions to England, which started in the late 8th century, might
not have involved any participation by Bornholmers. If they did, the material evi-
dence of their involvement is very insignificant. Of more than a thousand Viking Age
metal objects recovered during recent excavations and metal-detector surveys only
three are of English provenience (Aarsleff 2005, 19–20). All of them are fittings in
MAM 37: 06-naum – Press

viking age bornholm 83

Insular style, decorated with incised interlace pattern, one with an inlay of blue glass,
others with round amber beads, dated to the 8th or early 9th century.
The Danish involvement in England took a different turn toward the end of the
10th century. Now the plundering raids were frequent, better organized and encour-
aged by the ruling family. Large sums of silver were not enough for ambitious leaders;
the English throne was at stake. In the organization and successful execution of this
warfare Sven Forkbeard (ad 986–1014) and Cnut the Great (ad 1018–35) relied on a
network of alliances within the south-Scandinavian aristocracy. The finds of English
coins might thus be an indication of networks between the royal house and some of
the Bornholm families. Further and more convincing indication of such alliances is
provided by use of the titles thegn (mentioned once) and dreng (mentioned three
times) on rune stones thought to have been erected in the middle of the 11th century.
The specific meaning of thegn and dreng in a Scandinavian context is unclear.
They have been interpreted as denoting royal retainers (Aakjær 1927–28, 28; Löfving
2001; Sawyer 2000, 103), or an ancestral class of free landowners (thegns being
the owners and drengs their sons) (Lindow 1976, 106; Nielsen 1945, 121). Other
researchers suggest that such tight definitions are impossible, arguing that the terms
denoted slightly different meanings throughout the late Viking Age and early Middle
Ages (Thurston 2001, 114–124; Syrett 2000). It is likely that by the early 11th century,
during the reigns of Sven and Cnut, the term thegn (in addition to the sense of ‘free
landholding man’) described an individual who received at least some land and title
from a king (Syrett 2000, 268). The term dreng, on the other hand, might have been
used to describe a member of a war band, such as a participant in a Viking expedition
who remained under royal supervision and was rewarded by a king.
If these two terms denoted individuals under royal patronage and obligation,
they may indicate the rise of a new social class. This class, invested with land and
possibly some entitlement to surplus production, might have been strategically
created to counter traditional power relationships and to facilitate better control over
territories claimed by the Danish royal power (Thurston 2001, 113–130). Such col-
laboration offered reciprocal gains. For aspiring nobility it brought the prospect of
social and economic elevation; for a king it facilitated an extension of power and the
promise of considerable income.
Participation in the Viking raids of the end of the 10th century and the beginning
of the 11th century is the first clearly visible event that links some families from the
island with royal power. Another reflection of such collaboration is the emergence
of the so-called royal demesnes (kongsgårde), farms with large productive capacity
usually located peripherally in the landscape or adjacent to already existing settle-
ments. Their administrators might have been the thegns and drengs mentioned on the
runic stones. It is not certain when exactly this new form of settlement was intro-
duced on the island. Flateyjarbók from ad c1380 says that there were 20 royal farms
on Bornholm in the time of Cnut the Holy (ad 1080–86). The authors of Knytlinga
saga, written down ad c1250, mention 12 royal farms (Guðnason 1982, 152–154;
Nielsen 1998, 13). The emergence of such farms might have been contemporary with
the overall shift in settlement structure which took place in the 11th century, when
new farms were built in new locations (Nielsen 1994; 1998, 22; see above). However,
MAM 37: 06-naum – Press

84 maritime societies of the viking and medieval world

this locational shift did not lead to major change in the structure of settlement. Single
farms and hamlets, sometimes located very close to each other (less than 200m),
sometimes at much larger distances, still prevailed. The sizes and spatial arrange-
ments of early medieval farms are not always discernible due to a lack of completely
excavated structures, but archaeological survey and late medieval documents suggest
that large farms occupied by extended households and consisting of a number of
buildings existed side by side with smaller ones. What is characteristic of Bornholm’s
early (and later) medieval settlement is the total lack of villages or denser settlements
located along a village green or road.

history from things — a concluding remark

The available historical and archaeological sources dictate the way Bornholm’s
Viking Age and early medieval history is perceived. To a large degree it is a narrative
of the island’s interaction with the Danish kingdom — its gradual incorporation into
the structures of the realm — and a story of economic connections between the island
and other territories around the Baltic Sea. It is hard to escape these dominant frame-
works. Nevertheless, looking at political, social and cultural processes through the
prism of objects and their biographies can offer an alternative way to describe
Bornholm’s past. Biographies of objects are intertwined with the biographies of indi-
viduals, families and groups. Thus from an archaeological perspective the objects can
serve as metaphors for the life stories, personal connections, beliefs or cultural senti-
ments of individuals. In Viking Age and early medieval Bornholm, objects were trans-
formed from foreign to well-known, as illustrated by the example of Baltic ware.
They were used to contest or support changes introduced from above, as indicated by
distinctive funerary practices and by two different patterns of raising rune stones.
They were accumulated during raids in western and central Europe, exchanged with
merchants from the southern coasts of the Baltic Sea and buried in the ground to
protect the household from misfortune and/or to symbolize prosperity and family
networks. Finally they probably served as reminders of the past, evoking familiarity
and helping to re-articulate the identities of immigrants who settled on the island
towards the end of the Viking Age.
Objects, and the way they were made, acquired and handled, indicate cultural
connections and mobility between Bornholm and the territories of the Baltic Sea rim.
They also reveal the particularity of the island’s cultural, political and economic
traditions. Like other Baltic Sea islands, including Gotland and Öland, Viking Age
inhabitants of Bornholm increasingly looked to the sea as a source of promising
revenue and wealth. Their role as middlemen in trade, their piratical activities, the
strategic location of their homes, and their ability to provide safe harbours for those
heading further east, north or south, provided substantial earnings in the form of
silver deposited in hoards. However, while Gotland and Öland seemed to profit
mostly from trade and tribute collected in the east, the Bornholm hoards give the
impression that contacts with the south-western coasts of the Baltic Sea were of
primary importance.
MAM 37: 06-naum – Press

viking age bornholm 85

Bornholm and Gotland are comparable not only because of an exceptional

accumulation of silver. Their settlement structure (encompassing single farms and
lacking villages), distinctive early medieval burial customs, indications of a group of
farmer-merchants, and strong sense of independence are all grounds for assumed
similarities between these two island societies (eg Aarsleff 2005; Staecker 1996, 79).
These resemblances must be due in part to their peripheral locations in relation to the
mainland and their dependence on external contacts. The political history of the
islands follows different trajectories.
Similarities (and differences) between Bornholm and southern Scandinavia are
visible in the activities of everyday life. From the archaeological perspective one of
the most apparent changes is the introduction of Baltic ware. Although this shift
happens more or less simultaneously in all the western areas oriented toward the
Baltic Sea, the potters that were introducing new styles of ceramic on Bornholm came
from different geographic areas than those in the workshops of Scania, Zealand,
Falster and Lolland. Baltic ware found on Bornholm resembles that made and used in
Western Pomerania, the Oder estuary and the Szczecin Lagoon. Conversely, early
medieval pottery in other Danish territories was probably influenced by stylistic
solutions developed by potters from Rügen and inland territories of Mecklenburg-
Vorpommern. The island’s connections with the southern coast of the Baltic Sea are
also visible from Slavic settlement on the island, which has left very tangible archae-
ological traces.
Bornholm’s Viking Age and early medieval history, its connections and partic-
ularity converge very well with the stereotypical image of an island as presented in
saga literature. It comprises a landscape of separation and at the same time exhibits
an essential outward orientation. It is a place inhabited by colourful and strong-
willed figures and a site of extraordinary events (Zilmer nd). It is a landscape on the
crossways not only in the geographic sense, but also politically and culturally.

references
Aakjær, S, 1927–28 ‘Old Danish thegns and drengs’, Acta Philologica Scandinavica 2, 1–30
Aarsleff, E, 2005 Detektorfund — mangler og muligheder. Et metodisk studie af vikingetidens
detektorfund på Bornholm, Unpublished MA thesis, Copenhagen University
Białe˛cka, F, 1961 ‘Ceramika ze stanowiska wykopaliskowego 4 w Wolinie’, Materiały Zachodnio-
pomorskie 7, 293–370
Brink, S, 2007, ‘Geography, toponymy and political organisation in early Scandinavia’, in J M
Bately and A Englert (ed), Ohthere’s Voyages: A Late 9th-century Account of Voyages Along
the Coasts of Norway and Denmark and its Cultural Context, The Viking Ship Museum,
Roskilde, 66–75
Cnotliwy, E and Rogosz, R, 1983 ‘Wykop V’, in E Cnotliwy, L Leciejewicz and W Łosiński (ed),
Szczecin we wczesnym średniowieczu. Wzgórze zamkowe, Ossolineum, Wrocław, 87–168
Gebers, W, 1980 ‘Ostseekeramik auf den dänischen Inseln’, in H Hinz (ed), Kiel Papers ’80 from
Fuglsang/Lolland: Siedlungforschungen auf den dänischen Inseln und im westlichen Ostseer-
aum, Christian-Albrechts-Universität, Kiel, 139–168
Guðnason, B (ed), 1982 Danakonunga sögur. Skjöldunga saga, Knýtlinga saga, Ágrip af sögu
danakonunga, Íslenzka Fornritafélag, Reykjavik
MAM 37: 06-naum – Press

86 maritime societies of the viking and medieval world

Gurevich, A, 1995 ‘The French historical revolution: The Annales School’, in I Hodder, M Shanks,
A Alexandri, V Buchli, J Carman, J Last and G Lucas (ed), Interpreting Archaeology: Finding
Meaning in the Past, Routledge, London, 158–161
Hamerow, H, 2004 Early Medieval Settlements: The Archaeology of Rural Communities in North-
west Europe 400–900, Oxford University Press, Oxford
Hårdh, B, 1996 Silver in the Viking Age: A Regional-economic Study, Almqvist & Wiksell Interna-
tional, Stockholm
Hårdh, B, 2003 ‘The contacts of the central place’, in L Larsson and B Hårdh (ed), Centrality–
Regionality: The Social Structure of Southern Sweden During the Iron Age, Almqvist &
Wiksell International, Stockholm, 27–66
Hermann, J, 1997 Ralswiek auf Rügen: Die slawisch-wikingischen Siedlungen und deren
Hinterland. Teil I — die Hauptsiedlung, Archäologisches Landesmuseum Mecklenburg-
Vorpommern, Lübstorf
Herrmann, J and Donat, P (ed), 1979 Corpus Archäeologischer Quellen zur Frühgeschichte auf
dem Gebiet der DDR (7. bis 12. Jahrhundert.) Lieferung 2, Akademie Verlag, Berlin
Jørgensen, L, 1987 ‘Family burial practices and inheritance systems: the development of an Iron
Age Society from 500 bc to ad 1000 on Bornholm, Denmark’, Acta Archaeologica 58, 17–53
Jørgensen, L, 1998 ‘En storgård fra vikingetid ved Tissø, Sjælland — en foreløbig præsentation’, in
L Larsson and B Hårdh (ed), Centrala platser, centrala frågor: Samhällsstrukturen under
järnåldern, Almqvist & Wiksell International, Stockholm, 233–248
Karlin, G, 1923 ‘Medeltida keramik i Norden’, in G Hallström (ed), Nordiska arkeologmöte i
Stockholm 1922, KVHAA, Stockholm, 147–161
Leciejewicz, L, Łosiński, W, Rulewicz, M and Wesołowski, S, 1982 ‘Wykop I’, in E Cnotliwy,
L Leciejewicz and W Łosiński (ed), Szczecin we wczesnym średniowieczu. Wzgórze zamkowe,
Ossolineum, Wrocław, 15–62
Liebgott, N-K, 1979 ‘Spätslawische Keramik in Dänmark’, in B Chropovsky (ed), Rapports du
IIIe Congrès international d’archéologie slave Bratislava 1975, Slovenska Akademia Ved,
Bratislava, 493–505
Lihammer, A, 2007 Bortom riksbildningen: människor, landskap och makt i sydöstra Skandi-
navien, Lund University, Lund
Lindow, J, 1976 Comitatus, Individual and Honor: Studies in North Germanic Institutional
Vocabulary, University of California Press, Berkeley
Lund, N, and Fell, C (ed), 1984 Two Voyagers at the Court of King Alfred, William Sessions
Limited, York
Löfving, C, 2001 Gothia som danskt/engelskt skatteland: Ett exempel på herarki omkring år 1000,
Gothenberg University, Gothenberg
Naum, M, 2007 ‘Early Christians, immigrants and ritualized practice: a case study of south-eastern
Bornholm’, Lund Archaeological Review 11–12, 17–36
Naum, M, 2008 Homelands Lost and Gained: Slavic Migration and Settlement on Bornholm in the
Early Middle Ages, Lund University, Lund
Naum, M, 2009 ‘Memories, practice and identity: a case of early medieval migration’, in
M Georgiadis and C Gallou (ed), The Past in the Past: The Significance of Memory and
Tradition in the Transmission of Culture, British Archaeological Reports International Series
1925, Archaeopress, Oxford, 71–86
Naum, M, 2012 ‘Ambiguous pots: everyday practice, migration and materiality. The case of medi-
eval Baltic ware on the island of Bornholm (Denmark)’, Journal of Social Archaeology 12,
92–119
Nielsen, F O, 1994 ‘Middelaldergårde på Bornholm’, Hikuin 21, 125–138
Nielsen, F O, 1998 Middelalderens Bornholm, Bornholms Amt, Rønne
MAM 37: 06-naum – Press

viking age bornholm 87

Nielsen, K M, 1945 ‘Var Thegnerne og Drengene kongelige Hirdmænd?’, Aarbøger for nordisk
oldkyndighed og historie, 111–121
Price, T D, Naum, M, Bennike, P, Lynnerup, N, Frei, K M, Wagnkilde, H, Pind, T and Nielsen,
F O, 2012 ‘Isotopic investigation of human provenience at the eleventh century cemetery of
Ndr. Grødbygård, Bornholm, Denmark’, Danish Journal of Archaeology 1, 93–112
Randsborg, K, 1980 The Viking Age in Denmark: The Formation of a State, Duckworth, London
Roslund, M, 2006 ‘Kulturmötets konsekvenser. ‘‘Slaviseringen’’ av den skandinaviska keramik-
tradition’, in M Burström and J Rönneby (ed), Arkeologi och mångkultur: rapport från
svenskt arkeologmöte 2006, Södertörns högskola, Huddinge, 59–76
Roslund, M, 2007 Guests in the House: Cultural Transmission between Slavs and Scandinavians
900 to 1300 ad , Brill, Leiden
Sawyer, B, 2000. The Viking-age Rune-stones: Custom and Commemoration in Early Medieval
Scandinavia, Oxford University Press, Oxford
Skovmand, R, 1942 ‘De danske skattefund fra Vikingetiden og den ældste Middelalder’, Aarbøger
for Nordisk Oldkydighed og Historie 1942, 7–265
Staecker, J, 1996 ‘Searching for the unknown: Gotland’s churchyards from a gender and mission-
ary perspective’, Lund Archaeological Review 1996, 63–86
Svanberg, F, 2003 Death Rituals in South-east Scandinavia ad 800–1000, Decolonizing the Viking
Age 2, Almqvist & Wiksell International, Stockholm
Syrett, M, 2000 ‘Drengs and thegns again’, Saga-Book 25, 243–271
Thurston, T, 1999 ‘The knowable, the doable and the undiscussed: tradition, submission, and the
‘‘becoming’’ of rural landscapes in Denmark’s Iron Age’, Antiquity 73, 661–671
Thurston, T, 2001 Landscapes of Power, Landscapes of Conflict: State Formation in the South
Scandinavian Iron Age, Springer, London
Tschan, F J (trans), 1959 Adam of Bremen: History of the Archbishops of Hamburg-Bremen,
Columbia University Press, New York
von Heijne, C, 2004 Särpräglat: vikingatida och tidigmedeltida myntfynd från Danmark, Skåne,
Blekinge och Halland (ca 800–1130), Stockholm University, Stockholm
Wagnkilde, H, 1999 ‘Slaviske træk i bornholmske grave fra tiden omkring kristendommens indfør-
else: En oversigt over gravpladser og skattefund fra 1000 tallet på Bornholm’, META 2, 3–20
Wagnkilde, H, 2000 ‘Gravudstyr og mønter fra 1000-tallets gravpladser på Bornholm’, Hikuin 27,
91–106
Wagnkilde, H, 2001 ‘Slawische Relikte in borholmer Gräbern aus der Zeit der Einführung des
Christentums im 11. Jahrhundert’, in O Harck and C Lübke (ed), Zwischen Reric und Born-
höved: Die Beziehungen zwischen Dänen und ihren slawischen Nachbarn, vom 9. bis ins 13.
Jahrhundert: Beiträge einer internationalen Konferenz, Leipzig, 4.–6. Dezember 1997, Franz
Steiner, Stuttgart, 57–77
Wagnkilde, H and Pind, T, 1996 ‘Tæt på 1000-tallets indbyggere i Aaker sogn’, Bornholmske
Samlinger iii (17), 167–186
Watt, M, 1985 ‘En gravplads fra sen vikingetid ved Runegård, Åker’, Fra Bornholms Museum
1984–1985, 77–100
Watt, M, 1988 ‘Bornholm mellem vikingetid og middelalder’, in A Andersen (ed), Festskrift til Olaf
Olsen på 60–års dagen, Kongelige nordiske oldskrifselskab, Copenhagen, 105–122
Watt, M, 1991 ‘Sorte Muld: Høvdingesæde og kultcentrum fra Bornholm i yngre jernalder’, in
P Mortensen and B M Rasmussen (ed), Fra stamme til stat i Danmark Volume 2, Jysk
Arkæologisk Selskab, Aarhus, 89–107
Wienberg, J, 1986 ‘Bornholms kirker i den ældre middelalder’, Hikuin 12, 45–66
Zilmer, K, nd, Islands and Interaction in the Icelandic Narrative Memory, http://portal2.
nottingham.ac.uk/shared/shared_viking/documents/Zilmer.doc
MAM 37: 07-brink – Press

chapter 7
TRADING HUBS OR POLITICAL
CENTRES OF POWER? MARITIME
FOCAL SITES IN EARLY SWEDEN

By Stefan Brink

In all early complex societies we find places that have had a special function of some
kind, apart from being agrarian or fishing communities (in some cases with no
agrarian/fishing activity at all). Normally we call them Central Places, and the defi-
nition is that these places had some function or significance exceeding the partic-
ular site or settlement, hence some kind of ‘power’ over a wider area, a hinterland
(cf Brink 1996, 237). In Sweden, especially eastern Sweden, we know of, from the
early medieval period (ad c1100–1300), some royal strongholds or farms, often called
Sw husabyar, being part of a bona regalia (see eg Brink 2000). From the Scandinavian
prehistoric period (hence before ad c1100) we can deduce some other focal sites in the
landscape, the most notable settlements containing the element -tuna (see eg Brink
1996, 263–264; Hellberg 1985). It is notable that the -tuna-settlements very often geo-
graphically coincide with the later husaby-settlements. Whether this is an indication
of succession or just accidental, due to the fact that both are located on sites suitable
for a central place, is unknown. Furthermore, the -tuna-settlements occur in a land-
scape setting, where we can see that they are the focal site within a central-place
complex, where we have other neighbouring settlements indicating some special or
specialist function, such as forging, communal cult, communal assembly or even per-
haps a military force (Brink 1996, 238; 1999). These central-place complexes are still
to a large extent to be identified, analysed and dated. At the moment we only have a
vague knowledge of them.
In some of the identified Late Iron Age (probably ad c600–1000) central-place
complexes we find a settlement with a name containing the element -husa(r) in a
prominent situation in the landscape. Most of these cases are to be found along the
eastern coast of Sweden, and the sites are always, with one or two exceptions, located
strategically along the coast. These so-called -husa-names are very enigmatic, but also
problematic in several ways. As for today several scholars have noted their existence
and commented on them, but no comprehensive analysis have yet been conducted.
Many have seen the obvious resemblance between the husaby-names and the husa-
names, which has led some scholars to assume a genetic relationship and a functional
MAM 37: 07-brink – Press

maritime focal sites in early sweden 89

succession from -husa to husaby, but others have denied this. In this article I will try
to shed some light on this enigmatic name group. I will start by presenting a few
cases, and their toponymic and geographical context (more cases can be found in
Brink 1998), which might be indicative of the origin and background to these enig-
matic place names. I cannot yet present a comprehensive list of -husa names, nor
discuss a precise dating, but can point out that the -husa-names and settlements are to
be found mainly along the Swedish east and south coast and must be an Iron Age
phenomenon (ad c1–1100).

*husar in tjust
In the north-west part of the province of Småland in southern Sweden we find
the old folkland or land Tjust. It is mentioned already in the middle of the 1st mil-
lennium by Jordanes in his Getica as theu(s)tes, but the settlement district has an
obvious root back in the Bronze Age. Tjust is situated by the coast and the core
settlement is located at the inner part of two extensive gulfs (Figure 7.1). During the
Iron Age we find an obvious central-place complex (or perhaps two, see Brink 1999,
426) around the church at Lofta, where we find some remarkable ancient monuments
and place names. The church has obviously been built on land belonging to a today
lost settlement called *Husa(r)1 (Moberg 1962, 123). This settlement has been divided
into three parts: Norrsjö < Norðr-Husar ‘the northern part of Husar’, Södersjö <
Suðr-Husar ‘the southern part of Husar’ and Vivilsjö < Vivils-Husar ‘the vivil’s
Husar’. The latter name is most interesting since it probably contains a word for a
pagan cult leader, *vivil, ON vı́fill (Hellberg 1979, 129), underlining the typical char-
acter of and specialist functions found at a central-place complex.
During the Iron Age (ad c1–1100) it was possible to reach *Husar in Tjust from
the sea, via a long and narrow inlet. This settlement *Husar has a centrally, strate-
gically and well-defended location in the old land of Tjust. It is well protected
although easily accessible from the sea, and it is strategically located along the Swedish
coast. In this case it looks as if this settlement with the name *Husar, is the focal site in
the central-place complex we find here, not -tuna or any other central-place element.

norsa and runsa in uppland

In central Uppland we find some central-place complexes, which are very com-
plex and difficult to analyse (Figure 7.2). One of them — or rather two — are situated
very close to Stockholm airport, Arlanda, which is a modern name constructed from
the name of the medieval hundred (OSw hundare) of Ärlinghundra, and the even
older folkland or land, Arland (see Brink 1999, 428–429). The central-place complex
of interest for us on this occasion is the one we find by Lake Fysingen, where three
hundreds meet. Close to the church at Norrsunda, we find a husaby: Husby or
Åshusby. On this farm we have one of the largest burial mounds to be found in
Sweden, the so-called Nordians hög (Nordian’s Mound) (6069m). Lake Fysingen is
connected to Lake Mälaren and the sea by a river, Verkaån. These waters were of
great importance during the Late Iron Age and have been used for travels to the inner
MAM 37: 07-brink – Press

90 maritime societies of the viking and medieval world

Figure 7.1 The central parts of the hundred (härad) Tjust in the province of Småland,
eastern Sweden, with the parish Lofta. In the centre of this parish we find a lost {Husa(r),
which has been split up in Vivelsjö (< Vivils-Husar), Norrsjö
(< Norðr-Husar) and Södersjö (< Suðr-Husar) (S Brink)
MAM 37: 07-brink – Press

maritime focal sites in early sweden 91

Figure 7.2 The

hundred (hundare) of
Ärlinghundra (older
Arland) in the province
of Uppland, central
Sweden, with the two
husa(r) names Norsa
(< Nor-Husar) and Runsa
(< Run-Husar) (S Brink)

parts of southern Uppland. Verkaån debouches into the main sailing route to Lake
Mälaren and the sea. On each side of the river we find two -husa(r)-names: {Norsa
(< Nor-Husar) and Runsa (in Runusum 1313) (< Run-Husar).
At Runsa we have an impressive and strategically situated hillfort (Olausson
1995, 31–32, 131–142), which may be looked upon as a fortified settlement. The first
element in the name is obscure, but may refer to some rune stone standing in this
*Husar settlement. The latest phase of the hillfort seems to be the Migration Period,
and it is not impossible that the name Runsa is to be connected to this fortified settle-
ment/hillfort. The lost {Norsa, on the opposite side of Runsa, is today incorporated
in the manorial estate of Rosersberg. Some field names and old survey maps indicate
that the settlement of *Norsa must have been located close to the manor. The first
element Nor- is also rather obscure, it may denote the narrow inlet, or some small
basin, suitable as a harbour (see Hellberg 1986).
MAM 37: 07-brink – Press

92 maritime societies of the viking and medieval world

oppusa and {sjöhusa in rönö, södermanland

On a peninsula between Runnviken and lake Sundbysjön we find in a cemetery
the largest burial mound in the province, called Uppsa kulle (Uppsa’s hill) (Figure 7.3).
The first element in the name of the mound contains the name of the nearby hamlet of
Oppusa (i oppusom 1425), which goes back to a *Upp-Husa(r). Oppusa has a most
strategic situation at the old watercourse into the ancient Rönö hundred, which has
been a rich and large Late Iron Age settlement district (Figure 7.4). If we follow the
watercourse downstream to the sea, we find another interesting settlement with
the name Sjösa (i søsæ 1473, Siiøsa 1491), which has been — probably correctly —
interpreted as a *Siø-Husa(r). These two names, it seems, relate to each other, Sjösa
meaning ‘the Husa(r) by the sea’ and Oppsa ‘the Husa(r) upstream or higher up’
(Vikstrand 2000, 218–220).
As can be seen in the map (Figures 7.3 and 7.4) Oppusa (*Upp-Husar) is situated
in a central-place complex. There are central places such as Röntuna and the hundred
thing assembly at Aspa Löt. Moreover, there are settlements indicating pagan cult,
such as Torsberga (Thórr’s Hill) and Ludgo (Liuthguthuwi 1293), containing — in
my opinion — the word vi ‘pagan cult site’ and a much disputed first element (see
Wahlberg 2003, 197–198, cf Elmevik 2008).

Figure 7.3 The hamlet

Oppusa (< *Upp-Husar)
with the large burial
mound Uppsa kulle
(5569m), situated in the
parish of Runtuna in the
province of Södermanland
(S Brink and Vicki
Herring after Vikstrand
2000, 220)
MAM 37: 07-brink – Press

maritime focal sites in early sweden 93

Figure 7.4 The two husa(r)

names Sjösa (< *Siø-Husar) and
Oppusa (< *Upp-Husar) in the
eastern part of the province of
Södermanland, Sweden
(S Brink and Vicki Herring after
Vikstrand 2000, 219)

husa in össeby-garn, uppland

At a strategic site along the Iron Age water route from the sea up to Uppsala and
the central parts of Uppland we find the settlement Husa (Vikstrand 2008, 197).
Opposite Husa we find Vada (Figure 7.5), which probably was the main central place
in the central-place complex we find here. There is a monumental cemetery contain-
ing 114 burials, including three huge burial mounds, Sjökullarna (Raä Vada 10:1),
MAM 37: 07-brink – Press

94 maritime societies of the viking and medieval world

Figure 7.5 Husa in the parish of Vada, in the province of Uppland, central Sweden
(S Brink and Vicki Herring after Vikstrand 2008, 191)
MAM 37: 07-brink – Press

maritime focal sites in early sweden 95

looking very much like the famous three Royal Mounds at Old Uppsala. Vada also
has an early church, and the name means ‘a place in a stream where you can wade’.
Here Vada seems to be the major settlement, according to the archaeological remains.
However, it is not impossible that also here Husa was the main focal site. It is notable
that the large cemetery with the three large burial mounds is situated between Vada
and Husa, and it is not impossible that they should be connected to Husa instead of
Vada.

sunnersta in uppland
Where the River Fyris debouches into Lake Ekoln, which in its turn leads out to
Lake Mälaren and the sea, we find the settlement Sunnersta (in sundusum 1295) —
again — very strategically located, by the watercourse up to Old Uppsala. Imme-
diately north of Sunnersta we have Ultuna, famous for its boat burials. No doubt
Ultuna was one of the major central places in the region (see Hulth 2013). Also the
name is very interesting. It contains in the first element the name of the god Ullr. On
an esker by Sunnersta we find a hillfort, and cemeteries are found scattered here and
there in the former hamlet.

vikhus in västmanland
In the small parish of Lilla Rytterne in southern Västmanland is the castle
Viksjö. It takes its name from the nearby Vikhus by, during the Middle Ages the
largest hamlet in the district, hence the original Vikhus (curie vikhusum 1288) (Brink
2002). During the Iron Age an inlet went all the way up to the settlement, which the
first element of the name alludes to (vik ‘inlet’). This name has a direct counterpart in
Viksjö in Järfälla, Uppland, mentioned already on a runestone as i uikhusum (U92).
This Vikhus in Rytterne does not seem to fit with the rest of the known -husa(r)
settlements, since it is not part of a central-place complex. It is very much a solitaire.
This makes this -husa(r) settlement so much more interesting; is it a reminiscence of a
lost and by now obscure central-place complex, or does it have some other back-
ground (something which is developed below)?

conclusion
The origin of the -husa(r) names is most obscure. Lars Hellberg (1979, 150–152),
who was the first to highlight them as a toponymic group, saw them as royal settle-
ments, run by a steward. Hellberg has a very interesting but speculative explanation
for the element -husa(r), which also is most difficult to prove. He starts in Schleswig,
Germany, where we find Husum, 35km west of the important marketplace Hedeby.
Close by we have the most interesting name Schwesing (in Swezing 1381, de Swesen
1433, Swesum 1491), which has been interpreted as *Svea-Husar ‘the Husar belong-
ing to the Swedes’ (Laur 1992, art Schwesing). Hellberg is inclined to assume an
Iron Age ‘Swedish’ settlement in the area, and he connects this to the idea of a
MAM 37: 07-brink – Press

96 maritime societies of the viking and medieval world

‘svenskevælde’, a reign of Swedes, around ad 900 in Hedeby and the vicinity (a most
disputed idea). We also know that Husum was a Frisian marketplace from ad c700.
In this historical context, Hellberg assumes that Husum (nom. Husar) was trans-
ferred, from the ‘prototype’ Husum, to Scandinavia as a name for a marketplace, a
meaning which he assumes later on changed to ‘central place’ in general, and, in
Hellberg’s theory, to ‘an administrative centre under royal control’. Hellberg’s
ingenious hypothesis has not been discussed to any extent, and this hesitation is
certainly to be seen as a rejection by scholars (cf Brink 2007, 73).
We have seen above that the -husa(r) settlements are nearly always to be found
in coastal locations and in Iron Age central-place complexes — the one exception
being Vikhus in Västmanland. In some cases — especially in Småland — the -husa(r)
settlement seems to have been the actual central place — the focal site in the com-
plex ({Husar in Lofta, {Husar in Hossmo, {Husar on Bolmsö). In Uppland and
Södermanland there is nearly always another settlement in the immediate vicinity,
which seems to have been the focal site of power (Ultuna, Röntuna, Vada, etc).
Maybe we should separate the southern Swedish -husa(r) settlements from the central
Swedish ones? In Småland it looks as if the place-name element -husa(r) denoted the
central place in a central-place complex. This seems to have been the case with
{Husa(r) in Lofta, and also {Husa(r) in Hossmo (the predecessor to the town of
Kalmar) and {Husa(r) on the island of Bolmsö in the large lake of Bolmen.
In central Sweden, hence old Svitjod, we nearly always find Husa(r) beside or in
the vicinity of a major central place. In these cases it is tempting to look for another
explanation. When I discussed the case of Vikhus in Västmanland I noticed that this
settlement had an obvious function turned towards the lake, sea and water ways, and
proposed as a possibility that -husa(r) could be connected to the ancient naval orga-
nization (ledung) or to shipping and trade (Brink 2002, 57). Per Vikstrand (2008, 198)
has picked up this idea in his discussion of Husa by Vada. He highlights two impor-
tant cases, namely Skepphusa in Kungsåra, Västmanland (j skiphusom 1399) and
Skäsa in Mörkö, Södermanland (i skædhusa 1444). The first name contains the word
OSw skip ‘ship’ and the latter OSw skædh ON skeið ‘long ship, war ship’. These two
latter names seem to have the meaning of ‘ship house’, hence a building where you
keep a ship, in the case of Skäsa (< *Skædhusar) obviously a war ship. Actually most
-husa(r) sites in central Sweden can potentially have had this origin, based on their
location. Vikhus was placed in an inlet in Lake Mälaren. Sjösa in Södermanland is
where a watercourse debouches into the sea. Runsa, Norsa and Sunnersta are in the
same position. They are all in an inlet or at the mouth of a river where it debouches
into a larger lake or the sea. Maybe this is the background for the central Swedish -
husa(r) names?
We end up with something not expected, namely a situation where we cannot
see the Swedish -husa(r) names along the eastern coast as a coherent group. Instead
it seems we have to divide them into two groups, one for southern Sweden and one
for central Sweden. This unfortunate result finds support in another way. We also
find several -husa names in the province of Skåne: Abusa (< Abbahusa), Kabusa
(< Kabbahusa), Bjäresjö (< Biærghusa), {Mattwse (< Matthiashusa), Skabersjö
MAM 37: 07-brink – Press

maritime focal sites in early sweden 97

(< Skaparahusa), etc (Hallberg 1990). These names seem to be later than the -husa(r)
names, and most contain a personal name as the qualifier, which is never the case
with the older -husa(r) names. The -husa names in Skåne seem to represent a very
local name fashion. Again we have to look not for a general solution, but a regional
one.

note
1 In the article a * denotes a reconstructed name (eg {Norsa); < a word or name which is derived
or word (eg *Husar); { a name which is out of from an older form (eg Norrsjö < Norðr-Husar).
use or a settlement which has ceased to exist

references
Brink, S, 1996 ‘Political and social structures in early Scandinavia: a settlement-historical pre-study
of the Central Place’, Tor, Journal of Archaeology 28, 235–281
Brink, S, 1998 ‘Land, bygd, distrikt och centralort i Sydsverige: Några bebyggelsehistoriska nedsla’,
in L Larsson and B Hårdh (ed), Centrala platser, centrala frågor: Samhällsstrukturen under
järnåldern. En vänbok till Berta Stjernquist, Almqvist & Wiksell International, Stockholm,
297–326
Brink, S, 1999 ‘Social order in the early Scandinavian landscape’, in C Fabech and J Ringtved (ed),
Settlement and Landscape, Jutland Archaeological Society, Aarhus, 423–439
Brink, S, 2000 ‘Husby’, Reallexikon der germanischen Altertumskunde 15, 274–278
Brink, S, 2002 ‘Namn, landskap och bebygelese i Rytternebygden’, in O Ferm, A Paulson and
K Ström (ed), Nya anteckningar om Rytterne socken: Medeltidsstudier tillägnade Göran
Dahlbäck, Länsmuseet, Västerås, 47–68
Brink, S, 2007 ‘Geography, toponomy and political organisation in early Scandinavia’, in J Bately
and A Englert (ed), Ohthere’s Voyages: A Late 9th-century Account of Voyages Along the
Coasts of Norway and Denmark and its Cultural Context, The Viking Ship Musem, Roskilde,
66–73
Elmevik, L, 2008 ‘Ludgo, Luggavi, Luggude härad — och Locknevi’, in E Brylla and S Strandberg
(ed), Namn från land och stad: Hyllningsskrift till Mats Wahlberg 25 maj 2008, Uppsala
universitet, Uppsala, 37–46
Hallberg, G, 1990 ‘De skånska namnen på –husa’, Namn och bygd 78, 51–61
Hellberg, L, 1979 ‘Forn-Kalmar. Ortnamnen och stadens förhistoria’, in I Hammarström (ed),
Kalmar stads historia, Volume 1, Kalmar stad, Kalmar, 119–166
Hellberg, L, 1985 ‘Aktuell forskning om tuna-namnen 1–5’, Ortnamnssällskapet i Uppsala.
Namnspalten i UNT 3, 20–24
Hellberg, L, 1986 ‘Ordet nor i uppländska ortnamn 1–2’, Ortnamnssällskapet i Uppsala.
Namnspalten i UNT 4, 16–17
Hulth, H, 2013 Ultuna by - i händelsernas centrum. Boplats och rit. Bronsålder, yngre järnålder
och efterreformatorisk tid, Societas archaeologica Upsaliensis, Uppsala
Laur, W, 1992 Historisches Ortsnamenlexikon von Schleswig-Holstein, 2nd edn, Wachholtz,
Neumünster
Moberg, L, 1962 ‘Härads- och sockennamn i norra Kalmars län’, in H Stale (ed), Landstingsbygd.
Kalmar läns norra landstings minnesskrift 1862–1962, Landstinget, Västervik, 115–134
Olausson, M, 1995 Det inneslutna rummet — om kultiska hägnader, fornborgar och befästa
gårdar i Uppland från 1300 f Kr till Kristi födelse, Riksantikvarieämbetet, Stockholm
MAM 37: 07-brink – Press

98 maritime societies of the viking and medieval world

Raä Riksantikvarieämbetet (The Swedish National Heritage Board), Fornlämningsregistret,

http://www.fmis.raa.se/cocoon/fornsok/search.html
Vikstrand, P, 2000 ‘Konungen och helgedomen’, in B Sandnes, J Sandnes, O Stemshaug and
L Stenvik (ed), Oluf Rygh. Rapport fra symposium på Stiklestad 13.–15. mai 1999, Norna,
Uppsala, 213–223
Vikstrand, P, 2008 ‘Äldre ortnamn i Össeby-Garns socken’, in Namn från land och stad.
Hyllningsskrift till Mats Wahlberg 25 maj 2008, Uppsala universitet, Uppsala, 189–201
Wahlberg, M (ed), 2003 Svenskt otnamnslexikon, Språk- och folkminnesinstitutet, Uppsala
MAM 37: 08-rundkvist – Press

chapter 8
ACCESSIBILITY AND VULNERABILITY:
MARITIME DEFENCE AND POLITICAL
ALLEGIANCE ON THE VIKBOLANDET
PENINSULA, ÖSTERGÖTLAND,
SWEDEN

By Martin Rundkvist

The first decisive step in the formation of the medieval kingdom of Sweden appears to
have taken place about ad 1000 when two ethnic groups, the Svear and the Götar,
elected a shared king: Olof Eriksson skotkonungr (Gahrn 1988, 29). The Svear lived
around Lake Mälaren, the Götar on either side of Lake Vättern (Figure 8.1), and their
fertile lands were separated by the rugged forests of Tiveden and Kolmården. At any
one time during the later 1st millennium these two groups most likely had several
petty kings each, warlike characters whose exploits appear to be reflected dimly in the
Scilfingas and Geatas of Beowulf. Written sources for the land of the Götar in that era
are so few that the field of study is just barely proto-historical.
When the area entered the first flickering historical torchlight in the 11th cen-
tury, the lands of the Götar were divided into two halves separated by Lake Vättern:
Västergötland and Östergötland. They somehow belonged together as lands of
the Götar, but the western part was politically and culturally orientated towards the
Danish kingdom to the south-west, and the eastern part showed affinities with the
Swedish kingdom to the north-east. Viking Period settlement in Östergötland was
largely confined to a wide west–east belt of plains through the province (Figure 8.2)
and expanded up two river valleys in the forests to the south. The easternmost
quarter of the fertile plains belt is a peninsula, Vikbolandet, where the sea is never
farther away than 9km. Vikbolandet was densely settled at the time. It is highly acces-
sible from the Baltic, and thus vulnerable, and it was orientated immediately towards
the lands of the Svear. In this paper I review the evidence for Vikbolandet’s relation-
ship with powers from the sea in the Viking Period. We shall look at fortifications,
boat burials, precious-metal finds, runestones and the first royal manors of the united
kingdom of Sweden.
MAM 37: 08-rundkvist – Press

100 maritime societies of the viking and medieval world

Figure 8.1 Map of southern Sweden with its medieval landskap provinces
(M Rundkvist and Vicki Herring)

fortifications
There is no evidence that the Viking Period inhabitants of Vikbolandet erected
any fortifications to protect themselves. There are many hillforts of 3rd- to 5th-century
date on the peninsula, but the limited excavations that have been done so far have
turned up no evidence for Viking Period activity at those sites. On the other hand,
brief episodic use of the hillforts in times of unrest need not have left many traces.
The two long arms of the sea that define the Vikbolandet peninsula clearly posed
a strategic problem not only for people living on the peninsula, but also for areas
further inland from it (Figure 8.3). Fieldwork has documented and dated two impres-
sive structures designed to keep invaders entering these inlets from reaching inland
Östergötland. Note that these defences cannot have been of much use to people living
on the peninsula itself. When enemies struck, the people of Vikbolandet, who as we
shall see were rather affluent, basically had the choice of fending for themselves
or siding with the attackers. Mats G Larsson (2002, 146–148) has suggested that the
peninsula’s inhabitants may in fact have been members of an early version of
Svealand’s royal ledung organization for naval defence.
Some way up the river valley that empties into the inlet of Slätbaken is one
of Sweden’s very few dyke structures, a 3.4km long earthen rampart with a moat,
blocking the passage between two lakes (Nordén 1938, 240–255). According to radio-
carbon dating, it was constructed in the 9th century (Stjerna 1999). Its moat had silted
MAM 37: 08-rundkvist – Press

accessibility and vulnerability 101

Figure 8.2 Map of Östergötland province with -by farmstead names indicating Viking Period
settlement (after Franzén 1937)

up completely by ad 1100, indicating that upkeep ceased long before that date.
Archaeologists have dubbed the structure the Götavirke by analogy to the Danevirke,
the ramparts defending the southern border of Viking Period Denmark against the
Franks. Building the Götavirke was a huge undertaking: the structure is the first clear
evidence from Östergötland for a central power able to command large amounts of
labour. We appear to be dealing with a nameless 9th-century king of much of
Östergötland. Most of the rampart currently functions as a road bank.
There is no archaeological evidence for any Viking Period fortifications in the
Bråviken inlet north of Vikbolandet; all we have here is a place name of uncertain
age indicating a barrage near the inner end of the inlet. At the narrowest point of the
Slätbaken inlet, however, is an underwater barrage that has been dated to the
MAM 37: 08-rundkvist – Press

102 maritime societies of the viking and medieval world

Figure 8.3 Map of the Vikbolandet peninsula with fortifications, precious-metal hoards and
boat inhumation cemeteries (M Rundkvist and Vicki Herring)

10th century (Högmer 1988; 1989; 1999; Westerdahl 1986). It was a first line of
defence, 25km to the east of the Götavirke. These structures show that over time
there were a succession of kings in Östergötland with an area to defend, someone
they needed to defend it against, and the means to defend it. It is also clear that they
did not trust their naval power nor the military might of Vikbolandet’s inhabitants to
keep invaders out.

the identity of the attackers

Who were these potential attackers? Anyone around the Baltic with the means
to equip a ship could of course attempt to raid Östergötland. However, the nearest
historically attested major power was that of the Svear. The distance by sea from
Birka in Lake Mälaren to the mouth of the Bråviken inlet is only 110km. When a
group of ships left Lake Mälaren and set sail southward along the main shipping
route to western Europe, then the first non-Svear area they encountered was Öster-
götland. It would have been easy to turn starboard into the Bråviken inlet. Con-
versely, fleets could easily have come out of that inlet and turned to port with ill
intent.
For as long as the Svear and Götar saw themselves as distinct from each other,
and for as long as they were organized in separate polities, the potential for violence
remained. There must have been constant negotiation to keep aggression to an
acceptable level. Marriages must have been arranged and reciprocal feasts organized.
Looking at 1st-millennium ancient monuments and place names, there are actually
MAM 37: 08-rundkvist – Press

accessibility and vulnerability 103

greater differences between the two halves of Götaland than between Östergötland
and Svealand. If the earliest written sources had not presented evidence for a sepa-
ration between Svear and Götar, then we would have had little reason to believe that
there was an important border zone in the woods north of Östergötland.

boat inhumations
One trait in the material culture of Vikbolandet suggests particularly that the
area’s inhabitants had detailed knowledge of life in Svealand, specifically north of
Lake Mälaren. I am thinking of boat inhumation cemeteries. Such cemeteries are a
famous feature of the archaeological record in Uppland and Västmanland. For simple
reasons of preservation, the boat graves of sites such as Vendel and Valsgärde are
among the most well-known Vendel and Viking Period burials in all of Sweden
(Lidén et al 2001), regardless of the fact that they were extremely rare at the time.
However, there are actually three boat inhumation cemeteries in Östergötland as
well, and two of them are located in Vikbolandet. These sites are hardly ever men-
tioned in the literature, simply because there has been very little to say about them.
The first boat inhumation in Östergötland was excavated only in 2005 by myself and
Howard Williams, at Skamby in Kuddby parish near the centre of the peninsula
(Rundkvist and Williams 2008). It dates from the 9th century.
The Skamby grave is one of ten at the site, identified by boat-shaped depressions
in the surfaces of large oval stone pavements. Preservation conditions were not very
good (Figure 8.4), all bone having been obliterated, but we could establish that the
grave was not exactly like any one of the Svealand boat graves. There was no feast-
ing gear and no weaponry, the horse gear was minimal, most likely symbolic — and
then there was a set of 23 large amber gaming pieces (Figure 8.5). The last time
anything similar was found in a Swedish burial was when Hjalmar Stolpe excavated
the chamber graves of Birka in the 1870s; one of them had 15 amber gaming pieces
(grave Bj 524, Arbman 1940–43, 160–161, Taf 149). As for the Skamby graves’ super-
structures, there is nothing quite like them at the boat inhumation cemeteries of
Svealand. In addition to a covering stone pavement, the excavated grave originally
had a large standing stone at its centre, and possibly further ones at the edge of the
superstructure.
So at Skamby, we see ritual customs clearly inspired by what was going on in
Svealand, but not copying it slavishly. The people of Vikbolandet were cautiously
appropriating customs that might mark them as Svealand affiliates, but at the same
time keeping a distinct symbolic distance.

finds of precious metal

Another source that may indirectly speak of military matters is the precious-
metal finds (cf. Randsborg 1980). Judging from coin dates, precious-metal hoarding
in Östergötland began very late (eg Forsberg 1968; Malmer and Wiséhn 1982). There
are stray single coins from tpq (terminus post quem) ad 780 onward, but the first
hoard is tpq ad 951. Finds of precious metals are always interesting to plot on a map,
MAM 37: 08-rundkvist – Press

104 maritime societies of the viking and medieval world

Figure 8.4 Plan of a

boat inhumation at
Skamby in Kuddby
parish, 9th century
(after Rundkvist and
Williams 2008)

as each one marks a spot where a powerful person has been. However, we cannot
assume that each precious-metal find represents the same kind of event.
These finds were most likely not all buried for the same single reason. Most
scholars at the very least make a distinction between retrievable dry-land finds
(hoards), irretrievable wetland finds (sacrifices), and semi-retrievable finds in graves.
Viking Period precious-metal sacrificial finds from wetlands are rare in Östergötland
(three unambiguous ones and one from an island in Lake Åsunden, outside the plains
belt). None are known from Vikbolandet. Nor are any large amounts of precious
metal known from the province’s graves.
The retrievable finds most likely did not all remain in the ground for the same
single reason. A perennial question is whether a retrievable find was left in the
ground voluntarily or if its owner was somehow prevented from coming back to it. If
there is any one part of Östergötland where we may expect that an unusual number
MAM 37: 08-rundkvist – Press

accessibility and vulnerability 105

Figure 8.5 23 amber gaming pieces found in a boat inhumation at Skamby in Kuddby parish,
9th century (after Rundkvist and Williams 2008)

of hoard owners might have fallen victim to sudden attacks from abroad, then it is
Vikbolandet.
Looking at Viking Period precious-metal hoards from Östergötland’s plains belt
(and disregarding objects found singly, the size of a finger ring or smaller), we are
dealing with 19 sites evenly distributed across the plains, four of which are in or near
Vikbolandet. The following statistics divide the plains belt into four equally wide
slices. Hoards are not more common in Vikbolandet than anywhere else, nor do their
dates there (tpq ad 961 and ad 964) deviate from the norm. The hoards do contain
progressively less silver as we move west from Lake Vättern to the Baltic, decreasing
from 1000g of silver per hoard at Lake Vättern to 189g in Vikbolandet. However, two
of Vikbolandet’s hoards contain gold, which means that their average metal value is
actually the second highest in the province (assuming a 12:1 value relationship by
weight). The hoard evidence thus does not suggest that wealthy people were more
likely to be killed in raids in Vikbolandet than elsewhere in Östergötland. They were,
however, able and willing to commit great sums in precious metals to the earth. This
may be interpreted to mean either that Vikbolandet was an affluent area, that there
was a level of perceived threat there, or both.
People from Vikbolandet clearly travelled overseas themselves, possibly on
lucrative raids. Runestones of the early 11th century in Tåby, Dagsberg and Styrstad
parishes (Ög 30, Ög 145 and Ög 155) commemorate men who died in the east. The
Styrstad man met his fate on the Svear nobleman Ingvar’s ill-fated expedition of
ad c1040, and is thus unlikely to have contributed much to local wealth accumula-
tion. Nevertheless, his participation emphasizes Vikbolandet’s relationship with
Svealand at the time.
MAM 37: 08-rundkvist – Press

106 maritime societies of the viking and medieval world

husby
The people of Östergötland appear to have converted to Christianity around
ad 1000. This was just about the time when, as I mentioned initially, Olof Eriksson
succeeded in getting both the Götar and the Svear to elect him as their king. This was
in turn most likely one important reason why the Götavirke rampart was allowed to
fall into disrepair. Though Olof’s political position may have been precarious, during
his 27-year reign he began attempts to stitch his united kingdom together, which
would have reduced organized hostilities among his subjects. Nevertheless, a man
commemorated on a runestone in Söderköping (Ög 177) died in the early 11th cen-
tury because he was ‘cut down’. We do not know by whom.
In the 11th century, administrative centres for the nascent kingdom appeared in
the form of royal manors named Husaby (Olausson 2000; Rahmqvist 1994, 108; see
Chapter 7). This appellation was usually tacked onto manorial farms that had borne
quite different names before, and which had sometimes served as regional foci
already in preceding centuries. The distribution of Östergötland’s eight Husaby
manors shows quite plainly where the kings of the 11th century had their main points
of support: closest to Svealand. One is on Östergötland’s northern border, five are on
Vikbolandet, and only the remaining two are near the middle of the plains belt.
Suggestively, one of these royal manors was sited immediately outside the defunct
Götavirke rampart. Thus, summing up, it seems that in the 9th and 10th centuries the
kings of Östergötland despaired of defending Vikbolandet. Conversely, in the 11th
century that area became the main bridgehead in Östergötland of Swedish royal-
federal power.1

note
1 The observations discussed in this paper are also aspx?id=2104 — as an inexpensive hardback
included in Rundkvist (2011). It can be ordered book or a free PDF file.
from http://vitterhetsakad.bokorder.se/showTitle.

references
Arbman, H, 1940–43 Birka I: Die Gräber: Untersuchungen und Studien, Kungliga Vitterhets,
Historie och Antikvitetsakademien, Almqvist & Wiksell International, Stockholm
Brate, E, 1911–18 Östergötlands runinskrifter, Sveriges runinskrifter 2, Royal Swedish Academy of
Letters, Stockholm
Forsberg, C, 1968 ‘Östergötlands vikingatida skattfynd’, Tor 1967/68, 12–37
Franzén, G, 1937 Vikbolandets by- och gårdnamn 1, Nomina Germanica 1, University of Uppsala
Gahrn, L, 1988 Sveariket i källor och historieskrivning, Historiska institutionen vid Göteborgs
universitet, Gothenburg
Högmer, A, 1988 ‘Den marinarkeologiska undersökningen av Stegeborgssunden 1985–87’,
S:t Ragnhilds gilles årsbok 1987
Högmer, A, 1989 ‘Skyddade Stegeborg Ansgars Birca?’, Populär Arkeologi 1989(1), 27–29
Högmer, A, 1999 ‘Datering av pålar vid Stegeborg’, Marinarkeologisk tidskrift 1999(3), 10–12
Larsson, M G, 2002 Götarnas riken. Upptäcktsfärder till Sveriges enande, Atlantis, Stockholm
Lidén, K, Isaksson S and Götherström, A, 2001 ‘Regionality in the boat-grave cemeteries in the
lake Mälaren valley’, in B Arrhenius (ed), Kingdoms and Regionality: Transactions from the
MAM 37: 08-rundkvist – Press

accessibility and vulnerability 107

49th Sachsensymposium 1998 in Uppsala, Archaeological Research Laboratory, University of

Stockholm, Stockholm, 27–40
Malmer, B and Wiséhn, I, 1982 Myntfynd från Östergötland, Sveriges mynthistoria, Landskaps-
inventeringen 1, Royal Coin Cabinet, Stockholm
Nordén, A, 1938 Östergötlands järnålder 1:2, Kolmården — V. Husbydalen — Fornborgarna, Self-
published, Stockholm
Ög no, see Brate 1911–18
Olausson, M, 2000 ‘Husabyar, krig och krigare’, in M Olausson (ed), En bok om husbyar, Skrifter
Riksantikvarieämbetet Arkeologiska undersökningar 33, National Heritage Board, Stock-
holm, 125–150
Rahmqvist, S, 1994 ‘Ortnamn påverkade av administration i äldre tid’, in G Ulfsparre (ed),
Ortnamn värda att vårda. Föredrag från Riksantikvarieämbetets symposium Ortnamnskul-
tur, Stockholm 5–7 maj 1993, National Heritage Board, Stockholm
Randsborg, K, 1980 The Viking Age in Denmark: The Formation of a State, Duckworth, London
Rundkvist, M, 2011 Mead-halls of the Eastern Geats: Elite Settlements and Political Geography
ad 375–1000 in Ösergötland, Sweden, Kungl. Vitterhets historie och antikvitets akademien,
Stockholm
Rundkvist, M and Williams, H M R, 2008 ‘A Viking boat grave with amber gaming pieces
excavated at Skamby, Östergötland, Sweden’, Medieval Archaeology 52, 69–102
Stjerna, N, 1999 Götavirke. Arkeologisk undersökning 1999. Östergötland, Söderköpings kom-
mun, Västra Husby socken, Hageby Raä 12 & Hylinge Raä 111, Borgar och befästningsverk i
Mellansverige 400–1100 e.Kr, Archaeological Research Laboratory, University of Stockholm,
Stockholm
Westerdahl, C, 1986 ‘Stegeborgsspärrarna vikingatida!’, Meddelanden från Marinarkeologiska
sällskapet 9, 46
MAM 37: 09-willemsen – Press

chapter 9
DORESTAD AS A FLUVIATILE SOCIETY

By Annemarieke Willemsen

introduction
The inhabitants of Dorestad must have had an ambivalent attitude towards
water. To them, the river was two-faced: on the one hand, it was the very source of
their wealth, but on the other hand it was a problem and even a threat. Dorestad was
a harbour town, placed strategically on the branching of two waterways, the Lek and
the Kromme Rijn. These were uncanalized, meandering rivers with many bends that
kept shifting place. Dealing with this moving water determined the appearance of
Dorestad and defined life in it. The typical look of Dorestad, with all buildings
directed towards the river and a wide ‘promenade’ of platforms alongside its eastern
flank, follows directly from its dealing with wetness. Even the very feel of town and
countryside there must have been determined by the water: the green, yellow and
brown of the settlement and the yards was reflected everywhere in the blue and black
of the river and marshes (Willemsen 2009, 105).
In the Netherlands, not much attention has been given to the maritime aspect of
Dorestad. The Netherlands themselves are a maritime society par excellence, deter-
mined by sea and rivers even now. In the context of this wetland country, the mari-
time character of Dorestad was self-evident on excavation. Although so prominent in
the appearance and features of the settlement, the resulting implications were never
really thought through. In the European context Dorestad was part of a network of
emporia serving as hubs within the Carolingian empire, or (in large measure) as
places for trade with it. The idea of river-based trading centres in the northern
Netherlands was worked out by Frans Theuws (2004). Many emporia were sea ports,
placed at the smallest possible distance from open water. Dorestad is not located on
the sea, but nonetheless it has all the characteristics of a port. There are various ways
of looking at this. With the bird’s-eye view of Dries Tys (eg Chapter 10), the basin of
the Scheldt and Rhine in the early Middle Ages can be seen as one large delta, with
Dorestad as a converging point within it. From a ship captain’s point of view, it must
be emphasized that Dorestad was only 50km (just a few hours of sailing) from the
coast, either to the north or to the west. Also the Vikings could reach it and pull back.
Dorestad was an inland sea port. Its position in the hinterland, but at the same time
at the waterside is what made it unique and determined what it became. Dorestad
was a maritime society, but in a narrow sense. It was oriented on the river, and it
looked out to the sea and the rest of the world via the river.
MAM 37: 09-willemsen – Press

dorestad as a fluviatile society 109

fluviatile town planning

In 2009, for the exhibition on Dorestad in the National Museum of Antiquities
in Leiden, The Netherlands (www.rmo.nl/dorestad), a new reconstruction of the
Carolingian town in its hey-day was produced (Figure 9.1). It shows a bird’s-eye view
over the elongated settlement alongside the River Kromme Rijn, seen from the north,
which means that the harbour and trade quarter of Dorestad are in the foreground.
The illustrator Wim Euverman gave his reconstruction the title ‘Dorestad ad 833’,
that means just before the Vikings came. In ad 834, according to written sources,
Dorestad was attacked for the first time by ‘pagani’ (Lebecq 1983, 335). In the
decades following there was regular plundering of the town every few years, until
ad 863, when in a notably harsh campaign Dorestad was reputedly set on fire, with
part of the population killed and another part fleeing and never returning (Lebecq
1983, 307). This seems to have been the final blow to a town already in decline. It
meant the end of Dorestad, the vicus famosus as it was called by Liudger (Lebecq
1983, 97). It was the northernmost emporium in the Carolingian empire and a hub
for traffic and trade between the Empire and British and Scandinavian territories. It
was won by the Franks over the Frisians in ad 695, and significantly enlarged in the
8th century. It flourished especially under the protection of Charlemagne and his
successors in the first half of the 9th century. After the disappearance of Dorestad, the

Figure 9.1 ‘Dorestad ad 833’ by Wim Euverman (after Willemsen 2009, 22)
MAM 37: 09-willemsen – Press

110 maritime societies of the viking and medieval world

area became inhabited on a small scale again in the 10th century, but it held only the
memory of the earlier town. In a charter by Otto I from ad 948, the area was indi-
cated as a ‘villa, quondam Dorsteti, nunc autem Vvik nominata’: where once was
Dorestad, but that is now called Wijk (Lebecq 1983, 416). The present-day town on
the location of Dorestad has been called Wijk bij Duurstede from at least ad 1300
onwards (Van Es et al 1998, 58).
The bird’s-eye view illustrates how much of Dorestad was actually in the river.
The settlement itself is on the western bank, stretching for almost 3km, closely fol-
lowing the bends of the stream, with a maximum width of about 500m. But the settle-
ment also used the marshy western zone of the river bed to reclaim land. Extensive
wooden constructions excavated in the harbour area (Hoogstraat excavations)
covered at least the whole northern length of the river bank and were at places up to
200m wide. The constructions, initially interpreted as jetties, are now thought to be
platforms. They are oriented like the house plots and have a corresponding width, so
it looks like the already existing land divisions were extended into the river bed. In
their 2009 publication, Van Es and Verwers (2009, 343–350) conclude that these con-
structions must be seen as extensions of the living area and that they may well have
held buildings, if not houses, then at least warehouses and shops. The vast layers of
household waste excavated from underneath the wooden constructions were deliber-
ately put down as filling and reinforcement, a requirement for (rather than a result of)
the presence of platforms. The thousands of finds from the harbour area, initially
interpreted as items thrown off or lost from the jetties, must have been already
present in the dumped soil and may have come from other parts of Dorestad. Possibly
the large amount of pottery placed here was intentional, to strengthen the layer.
Dorestad thus claimed substantial land on the eastern side of the river bank. The
individual parts of the wooden constructions correspond with land plots on the ridge,
implying that the platforms were claimed and/or maintained as projections of land on
the other side of the street. The houses of Dorestad usually had entrances in both
short sides as well as on the long sides. In his excavations of a row of houses on the
river front, Jan van Doesburg found functional differences in the waste pits on these
short sides, with production waste towards the river, where there may well have been
a shop or a window to sell from, and kitchen waste on the ‘back’ end (Willemsen
2009, 31). The orientations of the carrying poles of the platforms dictate that any
large buildings on them must have been placed with one short side towards the river
and the other towards the street. As the platforms were narrow, and room to man-
oeuvre around whatever stood on them must have been limited, it is possible that the
buildings closest to the river were directed more towards the east, while those closer
to the street looked more to the west, to the living quarters. In this way, the pre-
decessor of the present-day Hoogstraat, which initially ran alongside the river, must
have become a central street in the course of the 8th century.
The chronological development of the wooden constructions was reconstructed
from the dendrochronological dating of the wooden poles and the finds from the
sequence of layers of waste in between. The first traces of occupation on the Rhine
bank in this Hoogstraat area can be dated to ad c625–650 (Van Es and Verwers 2009,
47). In this oldest phase of Dorestad there were separate jetties here. From the
MAM 37: 09-willemsen – Press

dorestad as a fluviatile society 111

beginning of the 8th century, the space between these was gradually filled in and
extended in an eastern direction. There was a large, organized campaign around
ad 725, extending most of the platforms to a length of 50m. Afterwards, the plat-
forms were constantly extended and repaired, until the end of the 8th century. The
platforms consist of elements measuring on average 10–12m, often clustered in twos
or threes as if belonging to one owner. The platforms were lined with vertical poles
and in between were shallow ditches for drainage. From the river they will not have
given the impression of a closed harbour front, but rather of groups of platforms with
small bays in between. When the water was high, they may have functioned as piers.
At lower tides it may have been possible to manoeuvre ships until they laid alongside.
The walkways on the platforms will have been of wood as well, but nothing survives
of these. Only the lower parts of the vertical elements have been preserved.
There were remains of over 150,000 wooden poles, showing that there was
continuous work on the constructions. The wood necessary for both the surviving
parts and what one might infer to have been on top has been estimated at around
2000 wooden poles (or a few hectares of forest) a year. According to the dendro-
chronological research, the lion’s share of this wood was imported from the large
forested areas of Germany, transported downstream on the Rhine. Again this shows
the position of the river: it was necessary as a transport route for the material needed
to control it.

excavating dorestad
Dorestad became untenable after the decline of the Carolingian Empire, and it
largely disappears from the sources after ad 864. Although the location of Dorestad
appears to have remained known throughout later ages, the actual Carolingian
settlement was only found in ad 1842 (Willemsen 2004, 71–77). Then, at a time of
great hardship, large quantities of animal bones were dug up outside Wijk bij
Duurstede. The bones were sold to be ground into fertilizer. During these digs, that
produced no less than one million pounds of bones, hundreds of objects like jewel-
lery, silver coins and pottery sherds surfaced. The workers unknowingly had found
— and destroyed — layers of household waste and hundreds of graves of the
inhabitants of the early medieval town. The finds were either sold or given away, and
many of them reached either the Society for the Province of Utrecht (Provinciaal
Utrechts Genootschap) or the National Museum of Antiquities (Museum van
Oudheden) in Leiden. There, curator L J F Janssen recognized the scientific value of
the discoveries. In ad 1842 and in ad 1844/45 he carried out excavations at the site,
well executed and documented considering the time he worked in. He concluded that
the settlement found was ‘the long-famous Dorestad’ and that it had been very exten-
sive. He also located the cemetery of Dorestad that is referred to as ‘De Heul’, from
which hundreds of burials were demolished in the fertilizer diggings (Willemsen
2010). For the remainder of the 19th century, there was little or no academic attention
to Dorestad. Even when, 40 years later, large bone digs were carried out again, there
was no archaeological supervision. The Swedish archaeologist H Stolpe, who
MAM 37: 09-willemsen – Press

112 maritime societies of the viking and medieval world

witnessed the diggings in ad 1880, was shocked by what he called the ‘scandalous
destruction of remains important to the cultural history of the Netherlands’.
Although the antiquities found were scattered, as he wrote in a letter to his director at
the Stockholm Museum in June 1880, he saw enough of them to conclude that they
‘corresponded exactly’ with the objects he had found at Birka. He recognized
Dorestad as the place of origin of all his ‘finer ware, and likely of the rectangular
enamelled brooches’. Birka and Dorestad were indeed two of the focal points in a
trade network that included most of northern Europe and the material culture of
both towns shows significant overlaps.
In the 1920s, the famous Dutch archaeologist J H Holwerda, at that time
director of the National Museum of Antiquities, carried out large excavations at the
site of Dorestad, mainly in the same areas where Janssen had worked. Through his
book Dorestad en onze vroege Middeleeuwen (Dorestad and our early Middle Ages,
Holwerda 1929), his excavations and his theory on the shape and size of Dorestad
became well known and widespread. Looking back, it is amazing that he drew such
major conclusions on such a small number of deliberately placed ditches. It seems
that he knew too well what he wanted to find. Holwerda’s sketch of the size and
location of Dorestad had great impact for a long time. When in the second half of the
20th century the area of Wijk bij Duurstede increasingly became farming and build-
ing land, again there was not much interest, as Dorestad seemed fully excavated, its
borders defined and no information missing. It took the archaeological world a very
long time to realize that the picture Holwerda sketched with authority was largely
wrong, that large areas of Dorestad were still in the ground and that they were
quickly disappearing in recent and planned building activities. To rescue what could
still be rescued, from 1967 to 1978 Dorestad was excavated by the National Archaeo-
logical Service (Rijksdienst voor het Oudheidkundig Bodemonderzoek, ROB), now
part of the Cultural Heritage Agency (Rijksdienst voor het Cultureel Erfgoed, RCE).
This was the largest excavation campaign the Netherlands have ever known. In an
almost full-time dig, about 35ha were examined, including other cemeteries, rows
of house plans and most of the harbour. Because of the find of the famous large
Dorestad brooch at the bottom of a water well (Figure 9.2), all finds of these ROB-
excavations (and the matching documentation) were allocated to the National
Museum of Antiquities in 1978, although it took well into the 21st century before all
the material was actually transferred to Leiden. In the decades following many more
excavations were carried out, and a large project was carried out in 2007–8. All in all,
approximately 80ha of the area of Wijk bij Duurstede have been examined at some
point in history. That is more than any other emporium, and in theory Dorestad
should be one of the most fully understood early medieval settlements in the Western
world. Unfortunately, most of the excavations remained unpublished or inadequately
published. In a large-scale backlog project entitled ‘Dorestad — Vicus famosus’,
financed by the Dutch Organization for Scientific Research (NWO), national services,
universities, museums and governments of the Netherlands worked together to fill in
some of the largest gaps in the period 2009–13 (www.vicusfamosus.eu).
What parts of Dorestad were preserved — and which parts were not — was
determined by river activity as well. In the northern quarter, much of the Carolingian
MAM 37: 09-willemsen – Press

dorestad as a fluviatile society 113

Figure 9.2 Dorestad

Brooch, ad c800 (National
Museum of Antiquities
Leiden no f 1978/1.1, after
Willemsen 2009, 82)

wood was still in place, kept wet for centuries, and the dig had to be drained con-
tinuously to excavate here. The southern part was swallowed by the river and never
found. As the river was meandering, it caused extensive erosion in its outer curves,
whereas in its inner curves sediment was deposited. The northern quarter was situ-
ated in an inner curve where the river moved eastward, leaving an increasingly wide
bank. In the beginning of Dorestad the Kromme Rijn flowed close to the present-day
Hoogstraat, but by the middle of the 9th century it flowed near the present-day
Nieuweweg, about 200m to the east (Van Es and Hessing 1994, 235). Along with the
‘Upkirika’, the church mentioned in ad 777 (Lebecq 1983, 409) and thought to have
been inside the Roman castellum, the ripaticum (harbour toll) for the River Lek and
the land between Rhine and Lek close to this church was donated to the Church of
Utrecht by the king. This means the castellum must have been embraced by the rivers
Rhine and Lek (Van Es and Hessing 1994, 98). Here, the river widened westwards,
eating away the land. In the days of Dorestad, this was not yet serious, but it became
so afterwards when eventually the area was flooded. During Dorestad’s occupation
the River Lek had already become more and more active, while the Kromme Rijn
became silted up and took less and less water. The harbour works reflect investments
to deal with this situation. The changes could be controlled when wood and labour
were in good supply, but were irreversible. Finally the River Lek took over and the
Kromme Rijn was dammed in ad 1122. These developments were part of broader
patterns. In the course of the 1st millennium ad all alluvial ridges and backlands of
MAM 37: 09-willemsen – Press

114 maritime societies of the viking and medieval world

the Kromme Rijn area became wetter and consequently less and less suited for occu-
pation and agriculture. In the Carolingian era already only the highest parts were
suited for settlement and all available land was built on. Agricultural products (eg
grain) were imported in large quantities from elsewhere (Van Es and Hessing 1994,
14–18). There were regular floods and they seem to have become more frequent,
resulting in wet periods. One particularly severe example happened shortly after
ad 870. This at least partly accounts for the absence of late 9th- and 10th-century
occupation at Dorestad (Van Es and Hessing 1994, 227).

a river-based market
Dorestad can best be characterized as in Van Es and Hessing’s (1994) article on
Frisians, Franks and Vikings, where they were inspired by De Ligt’s study on market
types in the Roman Empire (De Ligt 1993) and classified Dorestad as a permanent
interregional annual fair. An interregional annual fair is a meeting of tradesmen,
usually once a year, that lasts some time (three to eight weeks), has a wide range
(participants cover distances well over 300km), renders an enormous turnover of
especially luxury goods (exchange of the most expensive things), and that has an
entrepôt character (most wares are supplied directly and from a large distance). The
products traded at Dorestad were various:
from Frankish side came for instance precious jewellery and weaponry, luxury
glass, high-quality pottery like ‘Tatinger’ jugs decorated with tinfoil, manuscripts,
spices that may have been imported by way of the Mediterranean, gold brocade,
Chinese silk, Frisian cloth and exotic shells. Wines from the Rhineland were among
the most luxurious that the Frankish kingdom had to offer. Contacts between
Dorestad and the Rhineland are demonstrated clearly by archaeology, be it at the
level of daily utensils like millstones from the Eifel or household pottery from the
Vorgebirge near Cologne. Remarkable imports like limestone sarcophages indi-
cate relations with the central Meuse-region. What came in return on the part of the
English and Scandinavians is less clear: mainly perishable goods like slaves, fur,
whalebone and amber. (Van Es and Hessing 1994, 104–105)
Van Es adds that Dorestad had passed the stage of interregional annual fair in
two aspects: the long duration of the trade season (the full navigable season of half
a year between April and October) and the permanent presence of merchants on
the spot. This more permanent character of the market at Dorestad is typical for
maritime fairs, where trade goes on as long as the trading place is reachable, and
merchants stay as long as they are selling. Because the trade season was so long in
Dorestad, merchants settled in the town and local craftsmen specialized. The winter
season was used to build up stock, and evidently also to repair the harbour and the
quays where the fair took place. Was this a real market with stalls? ‘Why not?’
answer Van Es and Hessing (1994, 108). Therefore, in a recent reconstruction draw-
ing by Paul Becx, market stalls have been placed upon the wooden platforms
(Willemsen and Kik 2010, 25).
How did a ship arrive in Dorestad, or leave it? There were three main routes:
from the north, from the west and from the south-east. Ships coming from the north
MAM 37: 09-willemsen – Press

dorestad as a fluviatile society 115

could navigate southbound over the Zuyderzee, past the port of Medemblik and past
the small ecclesiastical centre of Utrecht, following the Kromme Rijn upstream until
reaching the northern quarter of Dorestad. This route connected the town with
Scandinavia. A second route led from the coasts of Holland and Zeeland inland,
upstream the River Lek. There was a harbour at Walacria near present-day Domburg
and via this route Dorestad was connected to the Scheldt basin and thus to the British
Isles and the northern French coast. The third alternative, most important as a supply
route, came from the south-west, through present-day Germany, downstream the
Rhine. Ships navigating this way reached the town on the other, southern side. This
waterway became the main route between the Carolingian Empire and the north.
The geographic position of Dorestad was essential to the type of market and
town it became. A land-based annual fair would be set up for some time at a site
where travellers could reach it easily, but they would come especially for the occa-
sion. Dorestad, conversely, was river-based, making it easy to reach for much of the
year. The fair here may well have started off as one confined in time as well, but
clearly it paid to keep the market in place. Dorestad’s location, embracing the
branching of the Rhine and Lek, made it possible to profit from all this river traffic,
and the connection to the Scheldt via the Lek was an advantage Dorestad had over for
instance Utrecht and Vechten (Van Es and Hessing 1994, 94). On routes between the
north, the coast and the Rhineland it was impossible to avoid Dorestad. For this
reason, Dorestad was given a toll privilege (praeceptum negotiatorum) and toll was
collected on the goods passing by (Lebecq 1983, 436). For the same reason Dorestad
became a staple market, where whole supplies were unloaded, knowing that enough
ships would sail by in the next weeks to sell everything. In this way, it became a
permanent fair where you could trade with a broad naval community.

fluviatile material culture

Some categories of finds recovered from the Dorestad excavations illustrate the
use and importance of the river to this community. These are objects directly related
to fishing, crossing the river and handling ships. What was on these ships has not
been included here, as the lion’s share of the tens of thousands of finds are trade
goods. Most of Dorestad’s material culture either arrived or left by river transport.
Fish hooks (Figure 9.3) and fish-traps are both known from Dorestad. Two funnel-
shaped birch traps had been laid in the area called Hoogstraat IV, in one line parallel
to the river bank, oriented roughly north-west to south-east, in a way that fish would
swim downstream into them (Van Es and Verwers 2009, 242–243, 248). Two creels
(rectangular baskets to contain caught fish under water) were in the river bed as well
(Van Es and Verwers 2009, 35, 242–246) (Figure 9.4). As indicated by the fish bones
from the 1967–78 excavations (Prummel 1983), 25 species of freshwater fish, salt-
water fish and shellfish were eaten in Dorestad (Figure 9.5; Figure 9.6). The sea fish
and shellfish were imported from the coast or the Zuyderzee, indicating that the
transport was profitable. Some species, such as herring, may have been traded over
even greater distances (Enghoff 2000). Remains of whale products were found as
well, seen in connection with whale strandings on the North Sea coast. The many
MAM 37: 09-willemsen – Press

116 maritime societies of the viking and medieval world

Figure 9.3 Fish hook

(National Museum of
Antiquities Leiden
no WD 713, after
Willemsen 2009, 113)

bone skates from Dorestad (pointed on both ends with traces of wear showing that
they were used for sliding) are also indicative of the riverine and swampy context of
the settlement, be they used under shoes or sledges (Van Es and Verwers 1980, 247;
Willemsen 2009, 115).
The many boat hooks (Figure 9.7) excavated point to a constant manoeuvring of
ships. The hooks are c20–30cm long and come in two types. Most common is the
fork-shaped one, used for pushing, but a few are large bent ones used for pulling
ships towards you. The boat hooks come from the harbour area, but the distribution
has not yet been analysed. In summer, ships must have come and gone all the time. In
the beginning, Dorestad seems to have been a beach harbour, where ships were pulled
up on the sand. This may well have remained possible later on, because of the
swampy nature of the settlement area, but the investments in the river area discussed
above allowed ships to navigate close to the platforms. They must then have been
pulled and poled alongside until they could be secured.
A few remains of ships have also been found at Dorestad (Van Es and Verwers
2009, 36–38, 244–247, 250–256; Vlierman 1996, 83–87). The main example consists
of clinker-built planks, each 2–3cm thick with animal hair (probably wool) caulk
between them, found scattered over about 30m (in excavation pits 386 and 387) in a
shallow natural depression in the harbour area named Hoogstraat II, at about 2.70m
below sea level. They form a small part of a 9th-century ship of Scandinavian or
MAM 37: 09-willemsen – Press

dorestad as a fluviatile society 117

Figure 9.4 Creel in situ during excavation (after Willemsen 2009, 112)

Figure 9.5 Bony plates (scutes) of sturgeon (National Museum of Antiquities Leiden
no WD69 7089, after Willemsen 2009, 34)
MAM 37: 09-willemsen – Press

118 maritime societies of the viking and medieval world

Figure 9.6 Oyster shells (National Museum of Antiquities Leiden, Dorestad Collection,
after Willemsen 2009, 35)

Slavonic origin that was for some reason dismantled in Dorestad (Vlierman 1994,
135). The ship is unique in the Netherlands, but its construction is reminiscent of the
Skuldelev ships of the 10th and 11th centuries. In between the shipwreck remains laid
part of a quernstone (find no 387.4.86, at 2.20m below sea level) and a circular basket
with fragments of bone and stone (find no 387.4.87, at 2.10 to 1.70m below sea level)
that may be a remainder of the cargo (Van Es and Verwers 2009, 36–38).
In the course of the one and a half centuries that the harbour of Dorestad was
maintained, thousands of ships must have visited Dorestad. But there is no indication
that there was ever a wharf here, or facilities for repairing ships. Although used ship
rivets have been found, a ship’s chain and a few pieces of used rope and caulk, the
amounts are too small to be more than incidental. Here it should be noted that there
are very few pieces of textile found in Dorestad in general. As so much wood survived
in situ — although almost none has been conserved — it is clear that re-using ship
timbers was not common in Dorestad. Just once or twice planks of ships were used
for lining a well, but almost all of the hundreds of wells were made of re-used wine
barrels. These were plentiful and already had the right shape. The reliable supply of
wood probably played a role. Later, in the 10th and 11th centuries, in the Dutch river
area, wood from British ships of Scandinavian type was intensely re-used in harbour
works, wells and even coffins — in Utrecht, Tiel and Vlaardingen, for example
(Oudhof and Bartels nd). This was after the regular supply of wood by long-distance
MAM 37: 09-willemsen – Press

dorestad as a fluviatile society 119

Figure 9.7 Boat hooks (National Museum of Antiquities Leiden, Dorestad Collection,
after Willemsen 2009, 107)

transport down the Rhine had vanished — a supply that Dorestad obviously could
still depend on.

maritime propaganda
The maritime orientation of Dorestad is reflected by the image chosen by the
town to advertise itself. Ninth-century coins minted in Dorestad during the reign of
Louis the Pious prominently show a ship, depicted detailed enough to see it is a so-
called hulk (Figure 9.8). This is a wide and short trade vessel with lots of cargo space,
the Carolingian counterpart of the Scandinavian knarr. As a vessel suited for both
river navigation and sea-going trade, this may indeed have been the vessel most seen
in Dorestad. Choosing a ship as an emblem, Dorestad advertised itself as water-
oriented; choosing the hulk, it presented itself as a place of trading vessels. In issuing
these coins, the 9th-century emperor emphasized the role of Dorestad as a
Carolingian trade port. This was the image presented to the whole world, as their
rapid circulation made coins the ultimate means of propaganda within the Empire
and outside, in the whole trade network.
MAM 37: 09-willemsen – Press

120 maritime societies of the viking and medieval world

Figure 9.8 Denarius of

Louis the Pious with hulk
and DORESTATVS
(National Museum of
Antiquities Leiden
no WD 792.1.17, after
Willemsen 2009, 120)

More debated, but nonetheless important, is the possible origin of the name
‘Dorestad’. It is generally accepted that the part ‘Dore’, ‘Dure’ or later ‘Duur’ refers
to a fortified structure. The Roman fort at this site must be meant, referred to in
ad 697 as the ‘castrum Dorestadum’ (LeBecq 1983, 328), still standing at this time.
The fortress, identified with the Roman castellum Levefanum, as that was known to
have been positioned at the branching of the rivers Rhine and Lek, seems to have
been swallowed by the river in the 11th century. Although it was never excavated,
phosphate concentrations and suggestive finds of high-status objects and weaponry
indicate that the fortress was used throughout the early medieval settlement period.
The part ‘stado’ or ‘(ge)stade’ seems to refer to a type of harbour, one where you
could pull ships onto land. This would define Dorestad as a fort with a beach harbour
(Willemsen 2009, 7). The duality of the name Dorestad is reflected by its layout, with
the fortress in the south and the harbour quarter in the north, connected by a road.
Its material culture is two-faced as well: Roman stone and elite weaponry and jewel-
lery on the one hand, but harbour works and trade goods on the other. Being badly
equipped for new types of ships of greater draught may be one of several reasons that
Dorestad did not remain a port in the later Middle Ages.

epilogue
That Dorestad was easily accessible by water contributed directly to its wealth,
but also put pirates on its trail. Doom came over the water, in swift ships. Whatever
MAM 37: 09-willemsen – Press

dorestad as a fluviatile society 121

route the Vikings took inland, they always came across Dorestad. Her open character
made the town easy prey, and moreover its fame had preceded it. The Vikings, when
they came, knew where to go, what could be found there and how much space there
was to land. Because the town embraced the rivers, there was effectively no way of
protecting her from attacks. The water in general was one of the reasons for both the
rise and the downfall of the riverside trade port of Dorestad.

references
De Ligt, L, 1993 Fairs and Markets in the Roman Empire: Economic and Social Aspects of Periodic
Trade in a Pre-industrial Society, Brill, Leiden
Enghoff, I B, 2000 ‘Fishing in the southern North Sea region from the 1st to the 16th century ad:
Evidence from fish bones’, Archaeofauna 9, 59–132
Holwerda, J H, 1929 Dorestad en onze vroegste middeleeuwen, Nijhoff, Leiden
Lebecq, S, 1983 Marchands et navigateurs frisons du haut Moyen Age, Vol 2: Corpus des sources
écrites, Presses Universitaires de Lille, Lille
Oudhof, J W M and Bartels, M nd De middeleeuwse haven van Tiel, Archeologie in Tiel 3,
Rijksdienst voor het Oudheidkundig Bodemonderzoek, Amersfoort
Prummel, W, 1983 Excavations at Dorestad, Volume 2: Early Medieval Dorestad, an Archaeozoo-
logical Study, ROB, Amersfoort
Theuws, F, 2004 ‘Exchange, religion, identity and central places in the early Middle Ages’,
Archaeological Dialogues 10, 121–138
Van Es, W A, Van Doesburg, J and Van Koningsbruggen, I B S, 1998 Van Dorestad naar Wijk bij
Duurstede: Het ontstaan van een stad ca. 600–1550 na Chr., Uniepers, Abcoude
Van Es, W A and Hessing W A M (ed), 1994 Romeinen, Friezen en Franken in het hart van Neder-
land, Van Trajectum tot Dorestad 50 v.C.–900 n.C., Matrijs/Rijksdienst voor het Oudheid-
kundig Bodemonderzoek, Utrecht/Amersfoort
Van Es, W A and Verwers, W J H, 1980 Excavations at Dorestad, Volume 1: The Harbour, Hoog-
straat I, ROB, Amersfoort
Van Es, W A and Verwers, W J H, 2009 Excavations at Dorestad, Volume 3: Hoogstraat 0, II, III,
IV, ROB, Amersfoort
Vlierman, K, 1996 ‘Kleine bootjes en middeleeuws scheepshout met constructiedetails’,
Flevobericht 404
Willemsen, A, 2004 ‘Scattered across the waterside: Viking finds from the Netherlands’, in R Simek
and U Engel (ed), Vikings on the Rhine: Recent Research on Early Medieval Relations
between the Rhinelands and Scandinavia, Fassbänder, Vienna, 65–82
Willemsen, A, 2009 Dorestad, een wereldstad in de middeleeuwen, Walburg, Zutphen
Willemsen, A, 2010 ‘Welcome to Dorestad: a history of searching and finding ‘‘the Dutch Troy’’’,
in H Kik (ed), Dorestad in an International Framework: New Research on Centres of Trade
and Coinage in Carolingian Times: Proceedings of the First ‘Dorestad Congress’ held at the
National Museum of Antiquities Leiden, The Netherlands June 24–27, 2009, Brepols,
Turnhout, 7–16
Willemsen, A and Kik, H (ed), 2010 Dorestad in an International Framework: New Research on
Centres of Trade and Coinage in Carolingian Times: Proceedings of the First ‘Dorestad
Congress’ held at the National Museum of Antiquities Leiden, The Netherlands June 24–27,
2009, Brepols, Turnhout
MAM 37: 10-tys – Press

chapter 10
MARITIME ENVIRONMENT AND
SOCIAL IDENTITIES IN MEDIEVAL
COASTAL FLANDERS:
THE MANAGEMENT OF WATER AND
ENVIRONMENT AND ITS
CONSEQUENCES FOR THE LOCAL
COMMUNITY AND THE LANDSCAPE

By Dries Tys

The salt marshes of Flanders were, despite being at the fringe of Neustrian and
Carolingian power, never marginal. They were settled by a maritime society which
practised sheep herding, fishing and exchange, and which lived in beach sites as well
as on terp mounds. The material culture of its settlements seems to indicate a certain
connection with the maritime world of Frisian and Anglo-Saxon ‘North Sea cultures’,
as Anthony Heidinga (1997) called them. However, the coastal settlers also had to
respond to certain social and political powers which existed on the sandy edges of the
lands. It was only when official powers started to interfere with ecological conditions
from the 10th century onwards (thereby becoming ‘maritime’ themselves) that the
North Sea lifestyle and social identity was thoroughly transformed.
From this date, the environmental policy of the counts of Flanders did indeed
have an enormous impact on the ecology and potential of the tidal environment of
coastal Flanders, on its infrastructure and on the spatial layout, landscape and iden-
tity of coastal society. In the lower salt marshes of Flanders they developed large
private coastal estates specializing in the production of wool and meat. In order to
intensify the produce of their domains, the counts focused on innovation and invest-
ment by developing a large-scale programme of water management in the comital
estates. This led to the embankment of coastal Flanders and the construction of
numerous dykes, dams and sluices. The new infrastructure transformed the environ-
ment of coastal Flanders into a hydraulic landscape. A ‘landscape strategy’ was aimed
at managing natural and social resources through the control and manipulation of
the environment as a political and economic tool.
MAM 37: 10-tys – Press

the management of water and environment 123

From the 12th century on, control of the coastal zone shifted towards local/
regional social formations subsequent to comital power strategies. Owing to the com-
plexity of their relationship with the environment, these groups were confronted with
a continuous search for technological solutions to the problem of water management.
The complexity of the watershed, the costs to keep it in balance and social repro-
duction strategies in a commercial rural society caused conflicts with consequences
for the social characteristics of the landscape.

living, worldview and the development of tidal-water

technology in medieval coastal landscapes
Frans Verhaeghe (2006) has raised the question as to whether (and, if so, how)
living in coastal and maritime environments led to common patterns of material
behaviour and social praxis. In other words, was there a medieval North Sea culture?
Did living in the same maritime environment at the edge of the North Sea with access
to the same maritime resources lead to common ideas of place and ecology, and
maybe even common landscapes? This paper explores similar questions and exam-
ines the functioning of coastal societies against the background of the changing per-
ception of the maritime environment as well as the changing role of coastal ‘contact
zones’ in the southern North Sea region, and more specifically coastal Flanders,
between the 7th and the 15th centuries.
Coastal Flanders developed during the 11th and 12th centuries into a highly
commercial, embanked polder area. Traditionally the story of the embankment of the
area is told from the perspective of the struggle of people against natural conditions
and especially the danger of inundations from the sea (Beekman 1932; Blanchard
1906; Verhulst 1959; see also Brown 1997, 315). In this respect, the embanking and
reclamation of the medieval coastal region, resulting in the well-known polder land-
scape, was considered to be a necessary and almost compulsory step — without ques-
tioning its wider social context or its meaning for coastal societies. In these works the
development of dykes and water technology was solely determined by ecological
dynamics and the human reactions they provoked, rather than social dynamics.
In north-west Europe, most coastal plains were embanked and reclaimed
between the 11th and 13th centuries (Allen 1997; Rippon 2001). The ecological con-
sequences of this control of tidal and inland waters were important, as described by
Cook and Williamson (1999, 2):
Within the embanked areas, the natural patterns of creeks and levees and the peat
and the mineral alluvium were manipulated and ultimately transformed by various
forms of agricultural use; new patterns of soil and topography were created, new
landscapes and new habitats formed.
Embankments, however, were not necessary for societies to develop means of
existence and settlement in coastal tidal environments, as is shown by many examples
in, for example, the British Fenlands and early medieval Frisia and Flanders (see Allen
2000, 1211; Darby 1983; Heidinga 1997; de Langen 1992; Tys 2004).1 Unembanked
salt marshes provided useful natural resources such as fowl, fish, reeds, potential for
MAM 37: 10-tys – Press

124 maritime societies of the viking and medieval world

salt production and vegetation suitable for sheep breeding and wool production. In
key areas of early medieval Europe, people developed a tolerant mentality towards
these ecosystems and towards the danger of being inundated (Squatriti 1998).
In c10th-century Flanders the decision was taken to start the reclamation and
embankment of the coastal plain through the application of tidal-water technology.
Was this a deliberate choice reflecting social changes rather than environmental
changes (inundations caused by storm surges due to climate change: Gottschalk
1980), or did the environmental opportunities offer such economic profits that the
protection of the power base through embanking became vital?
At least the development of a particular strategy in the coastal plain seems to
have been based on a certain knowledge of the dynamics of the natural environment,
as well as on certain social and economic dynamics (Brown 1997, 304–315). The
ambitious princely development strategy of the counts of Flanders between the 10th
and 12th centuries seems to be of vital importance here. What was the interaction
between social groups and their changing landscape, and did this lead to changing
attitudes towards the environment (van der Leeuw 1990, 305)? In order to try to
answer questions like this we will look at the environmental dynamics within coastal
Flanders, relating them to changes in the material landscape, the management of the
environment and the development of fisheries.

the coastal landscape and environmental dynamics up to ad c900

Detailed and systematic geomorphological research in Flanders by Cecile
Baeteman and her team has shown that during the period between ad 550 and 750 the
Flemish coastal plain became a stable and not significantly inundated area of salt
marsh interspersed with tidal channels, which were actively silting up and evolving
into channel ridges (Baeteman 1999; Baeteman et al 2002). This period also saw the
end of steady sea-level rises (up to 7cm per century since c3000 bc). Recent inter-
disciplinary research has sought to evaluate the evidence for the development of
settlement in coastal Flanders using a reassessment of the geomorphological evidence,
analyses of medieval and early modern cadastral sources, field and dyke systems, and
archaeological sources (Loveluck and Tys 2006; Tys 2003; 2013). This research has
suggested the existence of a settlement hierarchy of farmsteads on the contemporary
ground surfaces of silted-up tidal channel ridges focused on nucleated settlements of
multiple households, located on terp mounds such as Leffinge and Bredene (Loveluck
and Tys 2006; Tys 2005). Nucleated settlements like Leffinge became the central foci
for their microregions of approximately 15km2 at the administrative, religious and
economic levels from at least the 10th century. Moreover, there are indications of
their central role from at least the Carolingian period. These sites have parallels with
the role of ‘trade terps’ in Frisia, as suggested by Halbertsma (1963) and others.
The populations of the coastal landscape, largely located on tidal channels up
until the 10th to 11th centuries (Verhaeghe 1977; see below), possessed a maritime
focus within a landscape which is unlikely to have fulfilled all their subsistence needs.
The primary activities for the nutritional support of their households would have
been sheep husbandry and fishing (Verhulst 1995; 2002). The population, like others
MAM 37: 10-tys – Press

the management of water and environment 125

involved in specialist production of limited products, must also have been involved in
exchange for cereals and other necessities such as wood. The presence of significant
quantities of imported pottery, such as black burnished wares (so called Hamwic 13),
Rhenish red painted wares, Mayen ceramics and others in almost every single farm-
stead in the unembanked coastal area is evidence that these farmsteads had access to
networks of trade and exchange with other parts of the southern North Sea world
(Loveluck and Tys 2006; Tys 2007). Even though the exact nature of the exchange of
commodities is not known, the maritime inhabitants seem to have profited from a
combination of specialist husbandry in relation to coastal resources and access to
exchange networks. The likely major trade goods would have been wool or finished
wool garments (the so-called pallia fresonica), fish and salt (Ervynck et al 2004;
Tys 2005).
Retrogressive analysis of property relations in the coastal plain shows that the
populations of these terp-focused settlement hierarchies were free proprietors (Tys
2003, 266–273). That is to say, they may have owed some dues to respective regional
lords, whether counts or kings, but with the exception of these possible obligations
there is no evidence that they came under any other significant socio-political control.
A similar situation may be present in the regions of Friesland where Schmid and
others have suggested that ‘free peasant traders’ (Bauerenkaufleute) lived in the
higher coastal salt marshes involving themselves in wool production and trade from
their Langenwurten (Schmid 1988, 134–137).
In short, in the coastal plain of western Flanders there is evidence for unbroken
activity relating to settlement, probable specialist husbandry and commodity pro-
duction, and cross-channel and coastal exchange from the 7th century onwards. The
continental settlement pattern along the coast of the southern North Sea, between
the 7th and late 9th centuries ad, was thus a complex hierarchy of significant coastal
settlements acting as micro-regional central places involved in inter-regional
exchange, focused on a wider network of rural settlements also possessing imported
items and raw materials.

the counts of flanders and the transformation of the landscape,

ad 900 – 1200
In this landscape context the counts of Flanders started to construct their power-
ful princely position at the end of the 9th century. We know that Count Baldwin II
must have held several circular fortresses (diameter c200m) on the coastal plain which
were constructed shortly before ad 890 (the so-called castella recens facta), most prob-
ably initially intended to be refuges for the coastal population in the troubled ad 880s
when the great army of the Vikings controlled large parts of Flanders. However, these
comital ringforts, like Bourbourg, Sint-Winoksbergen, Veurne and Oostburg, can also
be viewed as an attempt to gain control over older estates and the activities and trade
of the free coastal maritime societies, after, and maybe also before the stabilization of
the situation in the coastal plain around ad 890. This is supported by the fact that
powerful counts like Baldwin II (ad 879–918) and Arnulfus I (ad 918–968) succeeded
in the creation of enormous territorial possessions in the same coastal plain, next to
MAM 37: 10-tys – Press

126 maritime societies of the viking and medieval world

the free lands of the older population. The development and acquisition of power
was strongly connected with the acquisition and employment of land and property in
order to raise the means to invest in a military force, socio-political networks, a court
and central administrative framework, and cultural capital (such as newly founded
collegiate churches) (Bisson 1995; Davies and Fouracre 1995). The counts did so by
gaining control over important resources through the usurpation of older estates
from the crown or ecclesiastical institutions, probably close to the ringforts and at the
edge of the coastal plain, near Bruges and Oudenburg, where we find a former
Carolingian fisc in Snellegem. They also exercised the regal right to waste grounds in
the coastal plain, which were to be found in the lower salt marshes in the estuaries of
the open tidal channels and rivers, such as the Zwin, the Yzer and the Gersta. For
example, on the east bank of the estuary of the River Yzer, 16km2 of salt marsh
became the property of the counts of Flanders who established large sheep domains
with a total capacity of 10,000 sheep.
The counts organized their property in the lower salt marshes into so-called
terrae ad oves. The actual sheep domains were not only organized as administrative
units or ministeria under responsibility of a preco, but they also formed distinc-
tive estates in the landscape, as is evidenced by several oval enclosures found in the
centres of these domains. These embankments were probably erected to create pro-
tected hayfield areas for the sheep as a kind of infield, most likely before the end of
the 10th century. This means that the counts started the intensification of sheep
estates through technological innovations. The next step in this process was the
systematic embanking of the comital estates by the construction of longitudinal
defensive dykes next to the last open tidal channels (end of the 10th–end of the
11th century). These dykes enclosed large areas of several square kilometres, which
were called ‘old lands’ in the historical sources. Similar situations are encountered in
most of the north-west European coastal plains that were reclaimed in the medieval
period (Eddison 1995; Gottschalk 1955; Hallewas 1984; Meier 2001; Rippon 2001;
Silvester 1999 and many others). It thus seems that the counts made use of the
inherent ecological possibilities of the coastal plain to organize specialized sheep
domains, which delivered large quantities of wool to the comital storehouses in the
different comital castra (eg Bruges, Veurne, Sint-Winoksbergen and Oudenburg).
This act of embanking had several consequences and implications. First of all, it
caused an environmental shock to the former salt marshes and salty meadows. The
salt marshes and mudflats were transformed into brackish secondary wetlands, free
from sedimentation and the permanent influence of salt water. Salt marsh plants
were replaced by rushes. Stilt walkers were replaced by birds like swans, ducks and
herons, while the new hydrologic situation favoured the reproduction of eels and
freshwater fish (Hoffmann 1996). The embanking also meant that the natural drain-
age of the former tidal environment, based on gravitational flows along salt-marsh
creeks, had to be modified and improved by a large-scale artificial network of drain-
age ditches. Through the act of drainage the soil also became dryer and minerals
oxidized, providing nutrients for richer pastures and agriculture (Cook and
Williamson 1999). Historical data show that agriculture and grain cultivation
became significant in coastal Flanders during the same period as the construction of
MAM 37: 10-tys – Press

the management of water and environment 127

the defensive dykes alongside the tidal channels (Verhulst 1998). Another implication
is that the counts must have had the ability to organize manpower on a large scale
and the financial capacity to carry out expensive flood control and drainage on this
large scale.
The drainage of the embanked area also required the construction of sluices and
outlets. Sluices were used to lead drainage waters at low tide through the dykes into
the open tidal channels, but also opening sluices at high tides and closing them at low
tides enabled fields to be irrigated in the summer. They were also used to regulate the
water level in the trading canals that ran through the domains. The outlets were
hollow trees placed through the dykes with a flap on the outside that was closed by
the pressure of the flood water, and which could be opened by drainage water at low
tide. These systems had been used from the Roman period (see an example from
Vlaardingen: de Ridder 1999) to at least the 17th century (Coles and Hall 1998, 78; see
also Brandt 1984 for a medieval example in north Germany). Although we do not
have archaeological evidence, some place names ending with ‘gote’ can be found
where watercourses crossed the 10th-/11th-century defensive dykes, suggesting that
similar constructions must have been present.
An environmental consequence of the embanking was that it caused rising storm
surge levels in the tidal sea-water channels, since the water had less space to spread
out (Baeteman 1998; Vos and van Heeringen 1997). Higher storm-surge levels implied
greater pressure on the dykes, and an increase of the risk of dykes bursting. Historical
sources indeed describe an increase in the number of disastrous floods in the 11th
and 12th centuries, with a lot of casualties (Verhulst 1959). This could have been the
result of similar dyke bursts, although we have to keep in mind that the embankments
must have made people feel more secure behind the dykes, so that settlement in the
embanked areas increased and therefore unexpected floods caused much more
damage and casualties compared to the situation before the embanking. In the
11th-century defensive dykes we also find so-called ‘wielen’, which are semi-circular
restorations of burst dykes (Allen 1997, 12–16).
In order to solve this problem, but also to increase comital land even further, the
decision was taken to embank the last remaining tidal channels by means of expen-
sive and complex hydraulic technologies, in particular the construction of large dams,
sluices and drainage channels (Tys 2001). In the 12th century, the estuaries of the
partially silted-up channels were blocked by large, stone constructions like the novum
dam on the River Yzer (around ad 1160: Tys 2001; 2003), the dam of Damme on the
Zwin near Bruges (ad 1180), the Hooghen Steendam near Bergues on the Gersta-
stream and so on. The channels themselves were replaced by smaller drainage canals
in order to be reclaimed and transformed into ‘nieuwland’ (new lands) of great socio-
political value. This practice entailed the construction of larger sluices, leading the
drainage water of large areas through the dams into the estuary of the River Yzer,
and then into the sea. An unforeseen consequence of all this was that the drainage
function of the former tidal channels had to be replaced by artificial systems. Only at
the end of the 12th century were new drainage canals dug to prevent inundations by
rain floods from the sandy inland region and to lead the inland waters safely to the
sea. The old and new drainage canals led vast amounts of water from inland areas
MAM 37: 10-tys – Press

128 maritime societies of the viking and medieval world

through the polders to the sea. This in turn led to the construction of larger sluices.
As such, these sluices were the crowning achievement of the embanking, flood control
and complex organization of the watershed area of the Flemish coastal plain.2
This comital hydraulic policy was clearly part of a larger spatial programme,
including the foundation of ports under direct control of the counts at the mouth of
the main watercourses in the coastal plain. These watercourses were the main trade
axes with the inland centres and the main axes of the drainage of the coastal area
(Verhulst 1967). The new ports (Nieuwpoort, Damme, Duinkerke, Gravelines and
Biervliet) were thus centres of comital control of water, hydraulic technology and
trade.
The counts thus created for themselves a dominant ‘market position’ in the
supply and trade of bulk commodities, which was organized during the 10th and 11th
centuries via the castral warehouses and marketplaces in, amongst other places, the
ringfort castra in Veurne, Bergues and in other castral centres like Bruges and Ypres.
Through this reorientation of economic activities the counts were responsible for an
important social transformation in the 10th- and 11th-century coastal plain, in which
the older free inhabitants lost their initial intra-regional position and came under
comital social and economic control. This is also shown in the landscape, not only by
the presence of the impressive, complex and large-scale hydraulics and embankments,
but also by the control of traffic and the exploitation of the environment as a kind of
feudal capital.
This new hydrological and technological organization of the coastal area meant
that the whole landscape depended on and had to participate in comital initiatives
and investments. Moreover, trade routes and communication were now centred on
the comital estates. The counts established a toll system on the entrances to the
waterways and droveways in the comital estates, often places where we also find
sluices and so called ‘overdragen’ (a kind of early ship-lift where river vessels could
cross dykes: see Pirenne 1925), thus cutting off the older society from their openings
towards the sea.

fisheries
The counts and their administration also controlled the fishing rights on the
(embanked) channels, which were often given in fief or donated to ecclesiastical insti-
tutions (see also de Boer 1997). These fishing rights were most often specifically
applied to eels and called ‘palincsetes’. Eel fishing in medieval Europe is often men-
tioned in a feudal context (Darby 1983, 23–24; Rippon 2001, 221–229).3 Feudal lords,
like the counts of Flanders, took advantage of the inherent environmental conse-
quences of embanking coastal waters and channels, which seems to have favoured the
conditions for eel reproduction and fishing in coastal Europe (Hoffmann 1996;
van Dam 1999). Thus fishing rights in the entire coastal area, including the region of
the early medieval free farmers, became ‘enclosed’ and subjected to comital control,
so that once more the former subsistence lifestyle of the free farmers, with free access
to resources, could no longer be maintained without comital control and surplus
extraction.4
MAM 37: 10-tys – Press

the management of water and environment 129

The Flemish fishery witnessed its heyday in the second half of the 12th and 13th
centuries, as shown by historical data studied by Degryse (1939). For example, at the
end of the 13th century, the fishing fleet of Blankenberge had more than 60 ships
(while Calais had 30 ships around the same time: Degryse 1939, 186–190). The grow-
ing group of consumers in medieval urban centres like Ghent and Bruges was prob-
ably the main incentive for the development and intensification of the Flemish sea
fishery, and consequently the dense pattern of coastal settlement (Degryse 1939,
195–196; Ervynck et al 2004). Archaeological and archaeozoological research in
Ghent and Tournai has shown that the ‘rise of frequency of herring on the inland
markets’ started in the early 11th century (Ervynck et al 2004, 233; see also Van Neer
and Ervynck forthcoming). According to historical sources, it seems that during the
11th and 12th centuries the supply of sea fish was at least partly organized from
within the large comital estates in coastal Flanders. There are indeed several indica-
tions that sea fish belonged to the market-oriented ‘produce’ of these coastal comital
estates (Thoen 2004; Tys 2004). Herring fishers around Veurne are mentioned in the
oldest-known general account of the revenues of the comital territorial estates dating
from ad 1187 (Verhulst and Gysseling 1962, 179–181).5 Herring are also mentioned
as revenue from the (large) comital estates in the areas around Bruges, Saint-Omer,
Bergues and especially Mardyck, which supplied no less than 130,000 herring
(Verhulst and Gysseling 1962, 151, 152, 181, 189).6 Herring (as well as eels and flat-
fish) were also listed in these accounts in relation to salt and peat that was used to
produce salt (through peat burning) (Verhulst and Gysseling 1962, 37, 74, 151–152).
These data show that herring were caught on a scale exceeding the level needed for
the rural communities to be self-sufficient. Furthermore, they show that these com-
munities were not only involved in fishing, but in several aspects of the ‘chaı̂ne opéra-
toire’ of fish processing, probably including salt production. This does not necessarily
mean that we are already dealing with specialized fishing communities, since it may
well have been a part-time activity of the people living on the comital estates that
were leased out from the end of the 11th century. This image of comital coastal
fisheries reflects the social and political context of fisheries elsewhere in Europe
between the 9th and 12th centuries. During this period fisheries became important
economic and feudal resources, owned by large landowners and powerful central
authorities (de Boer 1997; Squatriti 1998, 105–112).7
The sea fish caught by part-time specialist fishermen in the feudal context of the
comital estates were to be sold in the early urban centres, which might have been a
first step in the development of commercial fishing (Hoffmann 2000). In this respect,
it is remarkable that the oldest markets in cities like Ghent and Antwerp were fish
markets (Verhulst 1999, 78 and 100). These fish markets probably date from the 10th
and early 11th centuries, and were related to the comital castles that played an impor-
tant role in the formation and development of these cities during this period. These
castles were also the depots where produce from the comital manors and estates was
stored and sold on the town markets, leading to important interaction between castle
and trade settlement. Fish were apparently one of the products from the comital
estates to be sold at the castral markets, which explains the early frequent appearance
of herring in Ghent.
MAM 37: 10-tys – Press

130 maritime societies of the viking and medieval world

This origin of the development of sea fishing out of the comital domain system
also explains the mid-12th-century position of comital towns like Nieuwpoort and
Grevelingen as coastal fish markets. Both towns were comital creations on comital
terrains (the dunes were entirely comital property), where fish were collected and
where, from the middle of the 12th century, the counts’ officials collected tolls on the
traded sea fish (Degryse 1939, 193). The fish that were brought to these centres came
most probably from the domainal fisheries. The actual 11th- and 12th-century fisher-
men’s settlements that correspond with these fisheries are not well known, apart from
some fishing ports, such as Mardyck, sometimes with early medieval antecedents, like
Wissant. Several places between the rivers Aa and Yzer that had to pay herring tithes
at the end of the 12th century may go back to domainal fisheries, while several of the
known late medieval fisheries in the castellany of Bruges may also have antecedents in
older comital estate fisheries.
By choosing a market-oriented approach to resources and land, the general
comital policy thus also had a great impact on the development of coastal fisheries.
These fisheries were semi-specialized and their fish and salt production were to a
large extent market-driven, with consequences for the concurrent development of
urban market economies. Although archaeological evidence concerning these early
fisheries is still rare, the historical-geographical data show that fish, peat and salt
became important coastal commodities, as a result of the investment by the Flemish
counts in the maritime environment and landscape of coastal Flanders (Tys 2006).
Moreover, the comital landscape subsumed the former fishing rights and organi-
zations of the older maritime communities. Comital manipulation of the coastal
environment of Flanders thus seems to have resulted in the spatial, economic and
social ‘closing’ of the landscape.

social and spatial consequences, ad 1200 – 1500

In the late medieval period, the inhabitants of the maritime region had to pay
contributions to maintain the (comital) hydraulic system and the drainage, protection
and also irrigation of their lands. These contributions were called ‘geschot’ and were
raised by particular micro-regional institutions called ‘watering’ (Fockema-Andreae
1960; Soens 2001). The ‘wateringen’ were organized on the level of the micro-regional
public institutions called ‘ambachten’ (most often in accordance with 10th- to 11th-
century mother parishes). The head of each ‘ambacht’ (‘ambachtsheer’ or ‘amman’)
was an agent of central juridical power and often belonged to the group of comital
knights (Koch 1951, 125–129). The count’s main coastal knights also held positions as
scabini of the coastal area (burggraafschap Brugge), which had, at least in the
13th century, but probably earlier, jurisdictional power in watershed and drainage
matters. These functionaries also represented the central jurisdictional power for the
organization of the watershed and drainage, meaning that they mediated in juris-
dictional conflicts inside or between the ‘wateringen’.8 Jurisdiction and organization
of these institutions may therefore be of comital origin (Fockema-Andreae 1960,
190–191). It seems clear that, through use of his knights and vassals, the count tried to
control the economic, strategic and social importance of the water-control systems.
MAM 37: 10-tys – Press

the management of water and environment 131

The ‘wateringen’ themselves, although local public authorities, were organized

as an association and assembly of the landowners in the territory of the ‘ambacht’
and ‘watering’. The assembly was called ‘the meentucht’, and it decided on watershed
measures, the appointment of the executive officers like the ‘sluismeesters’ or sluice
masters (who belonged to the families of the comital estate knights and ‘ambacht-
sheren’) and the amount of the contribution (‘geschot’). The origin of the ‘water-
ingen’ as such probably goes back to the 12th century. After that the comital estates
were gradually leased out to individual landholders, who became the landowners in
real terms.
Tim Soens (2001) has shown that the democratic appearance of these institutions
was merely an idealistic anachronism. Decisions were controlled and steered by the
small group of larger landowners and by the executive officials, who belonged to the
network of the regional elite, with antecedents in the group of comital knights
involved in many (feudal and other) offices and functions in the region. Their policy
did not necessarily benefit the smaller landowners, and the financial contributions
were apparently not always well spent. Many sources indicate that the hydraulic
system was often neglected, especially during the 14th century (Augustyn 1992).
Between the end of the 13th and the 14th centuries, heavy storm floods caused the
drowning and loss of innumerable square kilometres of coastal lands, amongst others
the island of Wulpen, villages (such as Walraversyde and Scarphout) and even towns
(75% of Ostend, for example, disappeared into the sea during the Sint-Vincentius
flood of January 1394). According to geologists, this was a natural result of coastal
erosion caused by the embankment of the coastal plain between the 10th and 12th
centuries (Baeteman 1999; see Tys 2001). Nevertheless, the studies of Augustyn (1992;
see also Soens 2001) have shown that these disastrous floods were also a consequence
of the neglect of the maintenance of the dune sea belt and the sea dykes by the
waterboards.
The damage caused by these floods worsened the socio-economic situation of
the large group of small commercial holdings that had originated in the coastal plain
during the 13th and 14th centuries. This situation reached a crisis moment at the end
of the 14th century, especially in the surroundings of Ostend, where the costs of the
storm flood of ad 1394 were most likely unbearable for local community members.
In answer to the flood, the waterboard had to construct new sea dykes behind the
weak dune belt. The historical sources mention the construction of new dykes near
Walraversyde, west of Ostend, in ad 1399, 1404 and 1427.9 These dykes were part
of a large-scale, systematic construction programme of sea dykes on the coast of
Flanders between the end of the 14th century and the first half of the 15th century
(Rijckaert 1980).10 The construction of these dykes was organized and financed by
the ‘wateringen’. The local ‘watering’ of Kamerlings Ambacht most probably asked
for a significant additional ‘geschot’ from each landowner, which must have been an
extra heavy burden for the large group of impoverished small landowners in the
area.11 The accounts of the abbey of St Peter’s confronts us with the fact that between
ad 1394 and ad 1417 many landowners of the surrounding area of Walraversyde
could not pay their relatively low fixed rent.12 In ad 1401, arrears had risen to 75% of
the rents owed.13
MAM 37: 10-tys – Press

132 maritime societies of the viking and medieval world

The inhabitants of the domain clearly faced financial crisis in the years around
ad 1400. It seems that the financial demands of reconstructing the embankments fol-
lowing the ecological catastrophe of ad 1394, in the context of worsened general
socio-economic conditions at the end of the 14th century, impoverished most of the
landowners such that they could no longer pay their contributions. Smaller land-
owners of the area who could not pay the extra costs of construction could even have
been confronted with the right of ‘abandon’, through which those unable to pay for
dyke works could be expropriated (Soens 2001). The insolvency that occurred
amongst the landowners of the area around Walraversyde would have resulted in
expropriations, desertions and land sales, which could have triggered the start of a
concentration process of coastal properties into the hands of a small group of large
landowners during the 15th century (Verhulst 1990). Thus the watershed and drain-
age institutions acted as surplus extraction systems, with an increased social and
economic control of the rural community. They had important consequences for
the development of a rural capitalist landscape from the late 15th century onwards, a
development that still determines to a large extent the environment of coastal
Flanders.

conclusions
By choosing to invest in technological innovation, the counts of Flanders created
a comital hydraulic landscape and society. Hydraulic technology was aimed at the
management and manipulation of the environment of the comital landed possessions.
It integrated advanced technical skills, knowledge of natural history and the ambi-
tious neo-Carolingian policy of the counts of Flanders in the period ad 900–1200. It
allowed the counts to develop and control a valuable power base for their new prin-
cipality and to organize a landscape and society in the coastal plain that depended
almost completely on comital initiatives, with long-term consequences in the late
medieval period (Brenner 2001; Soens 2001). Hydraulic technology (including dams,
sluices, dykes, ports, canals, etc) formed a distinct material culture of power in the
coastal landscape, with its most direct and influential significance in the period
between the 10th and 12th centuries.
The counts thus created a landscape in which the ordering principles of the ruler
were articulated with environmental conditions in such a way that these principles
had to be accepted ‘without any conscious thought or consideration as to the way
things might otherwise be’ (Miller 1996, 404). By regulating the watershed, the counts
created a geographical dialogue of power, materialized in the comital landscape of
the coastal plain (Allen 2003, 2–3). Hydraulic technology was used to expand the
framework of conceptual possibilities of environment and space (see also Squatriti
1998, 72–75).
The success and large-scale application of water technology in medieval coastal
Flanders was intrinsically connected to ‘the social networks that employed them’ and
to ‘the economic system that made them viable’ (Squatriti 2000, xvi). Nevertheless, the
application of the technology itself also created economic opportunities and influ-
enced the intensive commercial economy that characterized medieval coastal Flanders
MAM 37: 10-tys – Press

the management of water and environment 133

under the influence of market-oriented comital policies (Brenner 2001; Thoen 2001;
Tys 2005). This intensive urban commercial economy, which goes back to the 10th–
12th centuries, changed the landscape drastically, with environmental consequences
that were not always intended. The new relationship with the environment placed
heavy demands on those who wanted to control it – on both the technological and
socio-economic levels (TeBrake 2000, 112). A question remains regarding the origin
of the technological knowledge behind this comital spatial policy. It is clear, how-
ever, that the risks of adopting new water technology were subordinate to the benefits
of accumulating power and developing a princedom.

notes
1 See also TeBrake (2000, 106): ‘Many cultures century eel fisheries at Wisbech in The Fens,
have found ways of living from such bounty where 33,000 eels were caught in ad 1086.
without actually draining them’. 8 This function was later assigned to the so-called
2 The coastal plain is still drained by the medi- ‘waesschout’ (water sheriff), the forerunner of
eval drainage system and organization without the ‘dijkgraaf’.
pumps. 9 ‘zyn gheghaen in den nieuwen dyc ghemaect
3 On the east bank of the Yzer, eel fisheries were int jaer 1404 metgaders pitlande daer huut
organized by the comital feudal courts of Bruges, ghenomen also ’t blyct 5 ghemten 1 line 37
Oudenburg and Veurne, amongst others (Tys roeden lands’ (Rijksarchief Gent, Fonds Sint-
2003, appendix II-6), or were given to abbeys Pietersabdij, Rek. 832c: fo23ro).
such as Oudenburg and Ter Duinen. 10 These dykes were constructed in the part of
4 In ad 1258, a case was made against Willardus coastal Flanders that in this period seems to have
Bateghoet, who apparently transgressed the fish- been most vulnerable to land loss and dune drift
ing rights of the monastery of Oudenburg on the (Augustyn 1992). From the 17th century, the sea
Lekeleed (one of the most important drainage dykes that were the result of these actions were
canals in the coastal plain), granted to them by called Gravejansdijk, after Duke Jan zonder
the Lord Chamberlain. The relevant charter Vrees.
states that he is not allowed to fish for eels ‘ad 11 Unfortunately, the sources available do not tell
anguillas sive capiendi anguillas in aqueductibus us the actual sum of this extra geschot, nor the
de Leka et de Ghistella, iacentibus inter locum normal sum of money that the landowners had
qui dicitur Boenburg in officio Camerarii, neque to pay annually for their property. Information
extra sclusas neque infra sclusas’. This source for the year ad 1682, however, throws some light
indicates the importance of sluice fishing in on this matter. In that year the waterboard asked
Flanders in the 13th century and probably also for an extra geschot of 1 schelling and 8 pen-
earlier (Rijksarchief Brugge, Blauwe nummers, ningen for each gemet (0.4ha) from every land-
6728, 6734). owner on top of the normal geschot of 3 schel-
5 Some piscatori de haringa delivered certain lingen 1 penning for each gemet in order to
quantities of herring (20,000) and flatfish (750), repair the sea dyke (Rijksarchief Brugge, regis-
to the count’s officials, which were listed under ters Brugse Vrije, 4039: Rekeningen van de kerk
the revenues of the comital lardarium and spic- van Middelkerke, ad 1670–84: 72ro). The price
arium of Veurne (ee Tys 2004). for an entirely new dyke undoubtedly must have
6 These amounts are to be considered as a rent in been much higher!
kind and are therefore merely a reflection of the 12 Rijksarchief Gent, Fonds Sint-Pietersabdij, Rek.
total amount caught by the fishermen of the 832a: fo8vo.
coastal estates. 13 Rijksarchief Gent, Fonds Sint-Pietersabdij, Rek.
7 See also Darby (1983, 23–24) on the royal 11th- 832b.

archival sources
Rijksarchief Brugge: Blauwe nummers, 6728, 6734: registers Brugse Vrije, 4039: Rekeningen van de
kerk van Middelkerke, ad 1670–84: 72ro
Rijksarchief Gent: Fonds Sint-Pietersabdij Rek. 832a: fo8vo, Rek. 832b, Rek. 832c: fo23ro
MAM 37: 10-tys – Press

134 maritime societies of the viking and medieval world

references
Allen, J, 2003 Lost Geographies of Power, Wiley-Blackwell, London
Allen, J R L, 1997 ‘The geoarchaeology of land-claim in coastal wetlands: a sketch from Britain
and the north-west European Atlantic-North Sea coasts’, Archaeological Journal 154, 1–54
Allen, J R L, 2000 ‘Morphodynamics of Holocene salt marshes: a review sketch from the Atlantic
and southern North Sea coasts of Europe’, Quaternary Science Reviews 19, 1155–231
Augustyn, B, 1992 Zeespiegelrijzing, transgressiefasen en stormvloeden in maritiem Vlaanderen tot
het einde van de 16e eeuw: Een landschappelijke, ecologische en klimatologische studie in
historisch perspectief, Algemeen Rijksarchief, Brussels
Baeteman, C, 1998 ‘Is het de natuur of de mens die een bedreiging vormt voor kusten en polders?’,
Mededelingen van de Koninklijke Academie voor Overzeese Wetenschappen 44, 257–273
Baeteman, C, 1999 ‘The Holocene depositional history of the Ijzer palaeovalley (western Belgian
coastal plain) with reference to the factors controlling the formation of intercalated peat
beds’, Geologica Belgica 2, 39–72
Baeteman, C, Scott, D B and van Strydonck, M, 2002 ‘Changes in coastal zone processes at high-
sea-level stand: a late Holocene example from Belgium’, Journal of Quaternary Sciences 17,
547–559
Beekman, A A, 1932 Nederland als Polderland, Nabu Press, Zutphen
Bisson, T N, 1995 ‘Introduction’, in T N Bisson (ed), Cultures of Power: Lordship, Status, and
Process in Twelfth-century Europe, University of Pennsylvania Press, Philadelphia, 1–8
Blanchard, R, 1906 La Flandre: étude géographique de la plaine flamande en France, Belgique et
Hollande, Nabu, Handzame
de Boer, D E H, 1997 ‘‘‘Roerende van der visscheryen’’: Enkele aspecten van de visvangst in
Holland en Zeeland tot de Sint Elisabethsvloed van 1421’, in J J J M Beenakker (ed), Holland
en het water in de middeleeuwen: Strijd tegen het water en beheersing en gebruik van het
water, Verloren, Hilversum, 115–140
Brandt, K, 1984 ‘Der Fund eines mittelalterlichen Siels bei Stollhammer Ahndeich Gem. Butjad-
ingen, Kr. Wesermarsch, und seine Bedeutung für die Landtschaftsentwickelung zwischen
Jadebusen und Weser’, Probleme der Küstenforschung im südlichen Nordseegebiet 15, 51–64
Brenner, R, 2001 ‘The Low Countries in the transition to capitalism’, in P Hoppenbrouwers and
J L van Zanden (ed), From Peasants to Farmers? The Transformation of the Rural Economy
and Society in the Low Countries (Middle Ages–19th Century) in the Light of the Brenner
Debate, Brepols, Turnhout, 275–338
Brown, A G, 1997 Alluvial Geoarchaeology, Floodplain Archaeology and Environmental Change,
Cambridge University Press, Cambridge
Coles, J and Hall, D, 1998 Changing Landscapes: The Ancient Fenland, Wetland Archaeology
Research Project, Exeter
Cook, H and Williamson, T, 1999 ‘Introduction: landscape, environment and history’, in H Cook
and T Williamson (ed), Water Management in the English Landscape, Field, Marsh and
Meadow, Edinburgh University Press, Edinburgh, 1–14
Darby, H C, 1983 The Changing Fenland, Cambridge University Press, Cambridge
Davies, W and Fouracre, P, 1995 ‘Introduction’, in W Davies and P Fouracre (ed), Property and
Power in the Early Middle Ages, Cambridge University Press, Cambridge, 1–16
Degryse, R, 1939 ‘Vlaanderen’s haringvisscherij in de middeleeuwen’, Annales de la Société d’emu-
lation de Bruges 82, 185–204
de Langen, G, 1992 Middeleeuws Friesland: De economische ontwikkeling van het gewest
Oostergo in de vroege en de volle middeleeuwen, Noordhoff, Groningen
MAM 37: 10-tys – Press

the management of water and environment 135

de Ridder, T, 1999 ‘De oudste deltawerken van West-Europa: Tweeduizend jaar oude dammen en
duikers te Vlaardingen’, Tijdschrift voor Waterstaatgeschiedenis 8, 10–22
Eddison, J, 1995 Romney Marsh: The Debatable Ground, Oxford University Committee for
Archaeology, Oxford.
Ervynck, A, van Neer, W and Pieters, M, 2004 ‘How the North was won (and lost again): historical
and archaeological data on the exploitation of the North Atlantic by the Flemish fishery’, in
R A Housley and G Coles (eds), Atlantic Connections and Adaptations: Economies, Environ-
ments and Subsistence in Lands Bordering the North Atlantic, Oxbow Books, Oxford,
230–239
Fockema-Andreae, S, 1960 ‘L’eau et les hommes de la Flandre Maritime’, Tijdschrift voor
Rechtsgeschiedenis 28, 181–196
Gottschalk, M K E, 1955 Historische geografie van westelijk Zeeuws-Vlaanderen, Volume I: Tot
de St-Elisabethsvloed van 1404, Van Gorcum, Assen
Gottschalk, M K E, 1980 ‘Subatlantische transgressiefasen en stormvloeden’, in A Verhulst and
M K E Gottschalk (ed), Transgressies en occupatiegeschiedenis in de kustgebieden van Neder-
land en België: Colloquium Gent 5–7 september 1978, Belgisch Centrum voor Landelijke
Geschiedenis, Ghent, 21–27
Halbertsma, H, 1963 Terpen tussen Vlie en Eems: Een geografisch-historische benadering,
J B Wolters, Groningen
Hallewas, D P, 1984 ‘Mittelalterliche Seedeiche im holländischen Küstengebiet’, Probleme der
Küstenforschung im südlichen Nordseegebiet 15, 9–27
Heidinga, H A, 1997 Frisia in the First Millennium: An Outline, Matrijs, Utrecht
Hoffmann, R C, 1996 ‘Economic development and aquatic ecosystems in medieval Europe’,
American Historical Review 101, 631–669
Hoffmann, R, 2000 ‘Medieval fishing’, in P Squatriti (ed), Working with Water in Medieval
Europe: Technology and Resource-use, Brill, Leiden, 331–393
Koch, A, 1951 De rechterlijke organisatie van het graafschap Vlaanderen tot in de 13de eeuw,
N V Standaard-Boekhandel, Antwerp
Loveluck, C and Tys, D, 2006 ‘Coastal societies and identity, ad 600–1200: rethinking the social
distribution of material culture along the Channel and North Sea coasts’, International
Journal for Maritime Archaeology 1, 140–169
Meier, D, 2001 Landschaftsentwicklung und Siedlungsgeschichte des eiderstedter und dithmar-
scher Küstengebietes als teilregionen des Nordseeküstenraumes, Dr Rudolf Habelt, Bonn
Miller, D, 1996 ‘Artefacts and the meaning of things’, in T Ingold (ed), Companion Encyclopedia
of Anthropology, Routledge, London, 396–419
Pirenne, H, 1925 ‘Les ‘‘overdragh’’ et les ‘‘portes d’eau’’ en Flandre au XIIIe siècle’, in A G Little
and F M Powicke (ed), Essays in Medieval History presented to Thomas Frederick Tout,
Manchester University Press, Manchester, 139–140
Rijckaert, M, 1980 ‘Resultaten van het historisch-geografisch onderzoek in de Belgische
kustvlakte’, in A Verhulst and M K E Gottschalk (ed), Transgressies en occupatiegeschiedenis
in de kustgebieden van Nederland en België, Belgisch Centrum voor Landelijke Geschiedenis,
Ghent, 75–92
Rippon, S, 2001 The Transformation of Coastal Wetlands: Exploitation and Management of
Marshland Landscapes in North West Europe during the Roman and the Medieval Periods,
Oxford University Press, Oxford
Schmid, P, 1988 ‘Die mittelalterliche Neubesiedlung der niedersächsischen Marsch’, in M Bierma,
O H Harsema and W van Zeist (ed), Archeologie en landschap: Bijdragen aan het gelijk-
namige symposium gehouden op 19 en 20 Oktober 1987, ter gelegenheid van het afscheid van
MAM 37: 10-tys – Press

136 maritime societies of the viking and medieval world

H T Waterbolk, Biologisch-Archaeologisch Instituut, Rijksuniversiteit Groningen, Gronin-

gen, 133–164
Silvester, R, 1999 ‘Medieval reclamation of marsh and fen’, in H Cook and T Williamson (ed),
Water Management in the English Landscape, Field, Marsh and Meadow, Edinburgh
University Press, Edinburgh, 122–140
Soens, T, 2001 ‘Het waterschap en de mythe van democratie in het Ancien Régime: Het voorbeeld
van de Vlaamse kustvlakte in de late middeleeuwen’, Jaarboek voor Ecologische Geschiedenis
2001, 39–56
Squatriti, P, 1998 Water and Society in Early Medieval Italy, ad 400–1000, Cambridge University
Press, Cambridge
Squatriti, P, 2000 ‘Introduction’, in P Squatriti (ed), Working with Water in Medieval Europe:
Technology and Resource-Use, Brill, Leiden, xv–xx
TeBrake, W T, 2000 ‘Hydraulic engineering in the Netherlands during the Middle Ages’, in
P Squatriti (ed), Working with Water in Medieval Europe: Technology and Resource-Use,
Brill, Leiden, 101–127
Thoen, E, 2001 ‘From a medieval peasant economy to the period of industrialisation: the
countryside and the evolution towards capitalism in Flanders’, in P Hoppenbrouwers and J L
van Zanden (ed), From Peasants to Farmers? The Transformation of the Rural Economy and
Society in the Low Countries (Middle Ages–19th Century) in the Light of the Brenner Debate,
Brepols, Turnhout, 102–157
Thoen, E, 2004 ‘‘‘Social agrosystems’’ as an economic concept to explain regional differences: an
essay taking the former county of Flanders as an example (Middle Ages–19th century)’, in
B J P van Bavel and P Hoppenbrauwers (ed), Landholding and Land Transfer in the North Sea
Area (Late Middle Ages–19th Century), Brepols, Turnhout, 47–66
Tys, D, 2001 ‘De verwerping van het zgn. Duinkerke-transgressiemodel en nieuwe inzichten in de
vroegste bedijking van de kustvlakte’, in E Huys and M Vandermaesen (ed), Polders en water-
ingen: Studiedag georganiseerd te Damme op 19 mei 2000, Algemeen Rijksarchief, Brussels,
17–53
Tys, D, 2003 Landschap als materiële cultuur: De interactie tussen macht en ruimte in een
kustgebied en de wording van een laatmiddeleeuws tot vroegmodern landschap. Kamerlings
Ambacht, 500–1200/1600, Unpublished PhD dissertation, Vrije Universiteit Brussels, Brussels
Tys, D, 2004 ‘De inrichting van een getijdenlandschap. De problematiek van de vroegmidde-
leeuwse nederzettingsstructuur en de aanwezigheid van terpen in de kustvlakte: Het voorbeeld
van Leffinge (gemeente Middelkerke, provincie West-Vlaanderen)’, Archeologie in Vlaande-
ren 8, 257–279
Tys, D, 2005 ‘Domeinvorming in de ‘‘wildernis’’ en de ontwikkeling van vorstelijke macht: Het
voorbeeld van het bezit van de graven van Vlaanderen in het IJzerestuarium tussen 900 en
1200’, Jaarboek voor Middeleeuwse Geschiedenis 7, 34–87
Tys, D, 2006 ‘Expressions of power in a marginal landscape? The social, economic and political
context of the 15th century fishermen’s village of Walraversyde’, in M Pieters, F Verhaeghe
and G Gevaert (ed), Fishery, Trade and Piracy: Fishermen and Fishermen’s Settlements in and
Around the North Sea Area in the Middle Ages and Later, Institute for Archaeological
Heritage, Brussels, 19–40
Tys, D, 2007 ‘La formation du littoral flamand et l’intervention humaine’, in L Verslype (ed), Villes
et campagnes en Neustrie (IVe–Xe s.): Sociétés, économies, territoires, christianisation,
Mergoil, Saint-Germain-en-Laye, 211–220
Tys, D, 2013 ‘The medieval embankment of coastal Flanders in context’, in E Thoen, G J Borger,
A M J d Kraker, T Soens, D Tys, L Vervaet and H J T Weerts (ed), Landscapes or Seascapes?
MAM 37: 10-tys – Press

the management of water and environment 137

The History of the Coastal Environment in the North Sea Area Reconsidered, Brepols,
Turnhout, 199–239
van Dam, P, 1999 Het succes van de paling in de Hollandse wateren, 1300–1600, Verloren, Utrecht
van der Leeuw, S E, 1990 ‘Rythmes temporels, espaces naturels et espaces vécus’, in J L Fiches and
S E van der Leeuw (ed), Archéologie et espaces, Xe rencontres internationales d’archéologie et
d’histoire, Antibes, Octobre 1989, Association pour la promotion et la diffusion des connais-
sances archéologiques, Juan-les-Pins, 299–346
Van Neer, W and Ervynck, A, forthcoming ‘The rise of sea-fish consumption in inland Flanders
(Belgium)’, in J Barrett and D Orton (ed), Cod and Consequences: The Archaeology and
History of Medieval Sea Fishing, Oxbow, Oxford
Verhaeghe, F, 1977 De middeleeuwse landelijke bewoningssites in een deel van Veurne-Ambacht:
Bijdrage tot de middeleeuwse archeologie, Unpublished PhD dissertation, Rijksuniversiteit
Gent, Ghent
Verhaeghe, F, 2006 ‘Living on the edge of the land: a few preliminary questions’, in M Pieters,
F Verhaeghe and G Gevaert (ed), Fishery, Trade and Piracy: Fishermen and Fishermen’s
Settlements in and around the North Sea Area in the Middle Ages and Later, Instituute for
Archaeological Heritagevoor het Archeologisch Patrimonium, Brussels, 19–40
Verhulst, A, 1959 ‘Historische geografie van de Vlaamse kustvlakte tot omstreeks 1200’, in
Bijdragen voor de Geschiedenis der Nederlanden 14, 1–37
Verhulst, A, 1967 ‘Initiative comtale et développement économique en Flandre au XIIe siècle: le
rôle de Thierry et de Philippe d’Alsace (1128–1191)’, in Miscellanea Mediaevalia in
Memoriam Jan Frederik Niermeyer, J B Wolters, Groningen, 227–240
Verhulst, A, 1990 Precis d’histoire rurale en Belgique, Éditions de l’Université de Bruxelles, Brussels
Verhulst, A, 1995 ‘Economic organisation’, in R McKitterick (ed), The New Cambridge Medieval
History II c.700–c.900, Cambridge University Press, Cambridge, 481–509
Verhulst, A, 1998 ‘Keulse handelaars in het Zwin tijdens de 12de eeuw en de vroegste ontwikkeling
van de Vlaamse zeehavens’, Bijdragen tot de Geschiedenis 81, 351–358
Verhulst, A, 1999 The Rise of Cities in North-west Europe, Cambridge University Press,
Cambridge
Verhulst, A, 2002 The Carolingian Economy, Cambridge University Press, Cambridge
Verhulst, A and Gysseling, M, 1962 Le compte général de 1187, connu sous le nom de ‘Gros brief’
et les institutions financières du Comté de Flandre au XIIe siècle, Koninklijke Commissie voor
Geschiedenis, Brussels
Vos, P C and van Heeringen, R M, 1997 ‘Holocene geology and occupation history of the Province
of Zeeland (SW Netherlands)’, in M M Fischer (ed), Holocene Evolution of Zeeland (SW
Netherlands), Netherlands Institute of Applied Geoscience, Haarlem, 5–110
MAM 37: 11-deckers – Press

chapter 11
THE MARITIME CULTURAL LANDSCAPE
OF EARLY MEDIEVAL NORTHUMBRIA:
SMALL LANDING PLACES AND THE
EMERGENCE OF COASTAL URBANISM

By Pieterjan Deckers

The maritime cultural landscape of early medieval Northumbria was characterized

by a recurring spatial and functional relationship between centre and landing place.
Central places were usually set back from the waterfront and served by spatially
distinct landing places, consistently situated at the closest navigable point to the
centre. Although a number of coastal central places in Northumbria at first sight do
not seem to conform to this model, it is argued that many of these still reflect the
basic features of spatial separation between centre and landing place. The emergence
of coastal urbanism from the 11th century onwards implied a drastic reorganization
of this settlement configuration, reflected in a process of (functional) nucleation and
reversal of the hierarchical relation between centre and landing place.
The debate concerning early medieval exchange and maritime activity has long
been preoccupied with the major, ‘urban’ emporia (eg Clarke and Ambrosiani 1991;
Hill and Cowie 2001; Hodges 1982). Only recently has the study of (micro-)regional
economic networks and rural trade centres come to the fore (eg Davies 2010; Naylor
2004; Pestell and Ulmschneider 2003; Ulmschneider 1999; 2000; 2005). Maritime
activity was often important to these regional economies, at least as far as England is
concerned (eg Gardiner et al 2001; Loveluck and Tys 2006; Loveluck et al 2007). Thus
this paper attempts to reconstruct the spatial organization of the interface between
rural centres and the maritime sphere, and how it changed through time. It does so
using a series of local case studies from Middle Saxon to early Norman Northumbria,
each focusing on a single central place in the coastal area.

the maritime cultural landscape of early medieval northumbria

A wide variety of imported goods, ranging from Continental pottery, quern-
stones from the Rhineland and Scandinavian whetstones, to more exotic items such
as silk constitute undeniable evidence of the overseas connections of early medieval
MAM 37: 11-deckers – Press

the emergence of coastal urbanism 139

Northumbria (eg Naylor 2004; Ferguson 2011). The economic centrality of the trade
centre of York, as attested by excavations at Fishergate (Kemp 1996) and Coppergate
(Hall 1984; 2000), is reflected in its status as a redistribution centre for certain
categories of imports, such as Mayen quernstones (Parkhouse 1997). The distribution
of other objects, however, indicates that York was not the only point-of-entry for
overseas imports. Many coin find spots are situated at 15km or less from the shore
(Naylor 2004, 48–51). The settlements within this ‘monetized zone’, known from
surface finds and/or excavations, mostly fall into the category of ‘productive’ sites,
usually interpreted as centres of religious or secular power, performing adminis-
trative and market functions for a regional hinterland (Ulmschneider 2000, 66–70).
Although subsistence was primarily terrestrial (Huntley and Stallibrass 1995,
157–168), there is evidence suggesting the exploitation of marine resources. At
Paddock Hill (Thwing), excavations yielded a shell midden of probable Anglo-Saxon
date, while oyster shells were found during field-walking at Cottam (Manby 1988;
Richards 1994). Also at Cottam, two stone net sinkers were found, one with an
engraved depiction of a ship (Richards 1994).
Therefore, it is probable that imports found at rural centres such as these
(partly) derived from active, direct involvement in maritime activity, rather than
overland redistribution from York. On the other hand, none of these sites is actually
coastal, a pattern which is also recorded in Domesday Book.1 This raises the ques-
tion of how this maritime activity was organized spatially. Where are the landing
places serving these inland centres?

localizing maritime activity

Landing places provide the opportunity for interaction between land and sea,
and thus constitute the defining element of any maritime cultural landscape. Early
medieval landing places are notably difficult to locate for a number of reasons.
Harbour infrastructure was usually unnecessary before the introduction of the cog
around ad 1200, as ships could simply be drawn on shore (McGrail 2001, 232).
Temporary beach markets leave ephemeral traces at best. Substantial evidence for
trade and craft activity at waterfront locations, although a reasonable indicator for
maritime activity (see Ulriksen 1994, 803; contra Ilves 2009, 159-160), only occurs at
the more intensively used sites.
Inferences can be made from the topographical qualities of a littoral location.
Different coastal localities offer varying degrees of shelter from winds and currents
and suitability for landing. Along the Northumbrian coast, a distinction can be made
between river mouths, protected coastal havens and unprotected coastal havens
(Makepeace 1995, 19). In spite of their limited and short-term protection, even the
latter, most numerous and least easily recognizable, category may be used as landing
places. Nevertheless, the interpretation of even a sheltered site as a landing place on
the basis of natural topography will remain just that — an interpretation.
An additional indicator for the perceived suitability of a littoral location as a
landing place is its use as a harbour in later medieval times. The topographical
requirements for landing ships may have changed through time, but the shelter
MAM 37: 11-deckers – Press

140 maritime societies of the viking and medieval world

offered was a constant factor. Furthermore, the location of certain medieval harbour
towns was arguably (partly) predetermined by their previous successful use as a
landing place (Deckers 2007).

a model for the maritime cultural landscape

The organization of the early medieval landscape profoundly differed from later
settlement configurations. Contrary to medieval towns, early medieval centres
usually consisted of a cluster of localities performing various functions (eg Fabech
1999). In coastal areas, the landing place often formed such an integral, yet distinct
element of the central place. In Denmark, the inland royal estate centre of Lejre
appears to have been served by a distinct landing place at Gevninge. This landing
place was situated about 2km from the mouth of the River Lejre Å, which connected
Gevninge with Lejre further upstream (Ulriksen 1994, 802–803). On the Isle of Wight,
the inland ‘productive’ site near Carisbrooke had a postulated landing place near
Newport on the River Medina (Ulmschneider 2005, 523). One of the few beach
markets to be archaeologically attested, Sandtun (Kent) has been linked to the inland
Middle Anglo-Saxon monastery of Lyminge by an 8th-century grant (Gardiner et al
2001; Thomas 2013).
A similar pattern of separation between centre and landing place has been noted
along the cliff coast of North Yorkshire (Pearson 1999). Several early medieval high-
status sites were situated at the heads of small river valleys and gorges leading
towards the coast. Not only did the river mouths provide shelter and fresh water,
they also formed the easiest points of access to the hinterland along the cliff coast.
These valleys provided natural routes towards the landing places serving the inland
centres. Unfortunately, there is so far no archaeological evidence corroborating these
conclusions based principally on an assessment of the natural topography.
Within the area of study, roughly the coastal region between Humber and Tyne,
there is one archaeological site shedding further light on the issue. Excavations at
Skerne in the upper Hull Valley brought to light the timber remains of either a bridge
or a jetty protruding into the water, preceded by an earlier, smaller jetty, dated
between the mid-7th and 11th centuries ad (Dent et al 2000). A nearby artificial canal
may have served as a dock (Leifsson 2005, 3). This site, located at the highest navi-
gable point on the River Hull, probably functioned as the landing place for the royal
estate of Driffield c5km upstream (Dent et al 2000, 242; Loveluck 1996, 43–45).
Skerne is an illuminating case in more than one sense. Firstly, it confirms that,
within the area of study, the observed pattern is not restricted to the particular
coastal topography of North Yorkshire, but also applies to the low-lying wetlands of
the East Riding. Secondly, it sheds additional light on the topographical factors deter-
mining the location of landing places. Set at the highest navigable point on the river
(as is also supposed for the landing places of Lejre and Carisbrooke; more generally
see Mägi 2004, 137–138, 143; Sindbæk 2009, 103), Skerne allows one to view the
postulated landing places along the North Yorkshire cliff coast in a different light.
They are not necessarily coastal; they are at the nearest navigable point to the central
place.
MAM 37: 11-deckers – Press

the emergence of coastal urbanism 141

Figure 11.1 Schematic

representation of a
heuristic model for the
spatial organization of
the maritime cultural
landscape of early
medieval Northumbria
(P Deckers)

The shared characteristics of these cases can be distilled into an idealized model
of coastal settlement (Figure 11.1), which will serve as a heuristic device, helping to
isolate patterns in the organization of the landscape. In this model, the centre is situ-
ated at some distance from the coast and served by a landing place which is located as
nearby as the navigability of the local waterway allows. This limit of navigability
may be determined by natural factors or infrastructural works, such as a dam or
bridge. The landing place therefore acts as a transhipment point where goods were
transferred to other modes of transport.
This model reconciles features of two early medieval settlement types that are
sometimes regarded as diametrically opposed: the central place and the trading site.
Although this distinction is usually made in reference to settlements of supra-regional
or international importance (emporia and civitates), many elements of the model
remain valid. Sindbæk (2007; 2009), for example, contrasts regional central places
with ‘nodal points’ (eg the major trading places of early medieval Scandinavia),
observing that, while the former depend on local traffic and are positioned to maxi-
mize their hinterlands, the latter’s locations are often determined by topographical
barriers and profit from the flow of longer-distance traffic.
Furthermore, several authors (eg Verhulst 2000, 115) point to the functional dif-
ferentiation between the two categories of settlement. The best case in point is
offered by the two principal settlements in the kingdom of Wessex, Hamwic and
Winchester (Biddle 1976; Hill 2001, 79, 80). The locations of both sites, one coastal,
the other inland but connected by river, neatly conform to the model. In addition,
both sites were functionally complementary: while governmental functions (adminis-
tration, religion, and jurisdiction) were concentrated at Winchester, Hamwic ful-
filled mainly commercial functions (exchange, craft production). Although the scale
of activities at these two sites is proportionate to the size of their hinterland, a similar
form of functional differentiation may be imagined for more local centres and land-
ing places, with the latter occasionally serving as foci of small-scale exchange (‘beach
markets’).
MAM 37: 11-deckers – Press

142 maritime societies of the viking and medieval world

an early medieval settlement configuration

The early development of a number of medieval harbour towns in Northumbria
will be tested against this model. Criterion for selection was inclusion in the so-called
‘fifteenth’ of ad 1203–04. This tax on imported and exported goods was recorded in
the form of a list of the 35 largest trade ports of the east and south coasts of England
(Lloyd 1977, 10–11). Of the nine Northumbrian ports on the list, four are examined
in this paper: Wyke-upon-Hull, Newcastle-upon-Tyne, Hedon and Scarborough.
A fifth port, York, is discussed further on as one of the early medieval sites deviating
from the model. The selected ports are situated in a range of natural landscapes: from
the low-lying wetlands of the East Riding to the cliff coasts of North Yorkshire and
the steep banks of the River Tyne.
Scarborough
Scarborough is an unambiguous example of the proposed settlement model.
The Norman castle on the headland was founded in the late ad 1130s. The earliest
development of the town at the foot of the headland probably started in this period as
well. A Middle Saxon chapel on the headland, about which little is known, is the only
indication of pre-Norman activity in the town area. South Bay at Scarborough, how-
ever, is one of the best natural harbours along the Northumbrian coast as it offers
both shelter from prevailing winds and currents and a gently sloping, sandy beach
ideal for landing (Bell 1998, 310–311; Pearson 2005).
Applying the heuristic criteria of the model, which requires an awareness of the
natural and cultural topography of the surrounding landscape, helps to shed addi-
tional light on the origins of Scarborough. As stated above, the steep coastal slopes of
the North Yorkshire coast are pierced by small valleys allowing access to the interior.
At Scarborough, Ramsdale valley runs inland from South Bay towards the Domesday
estate centre of Falsgrave about 2km away (Pearson 1999, 162). Thus, Scarborough’s
South Bay is the most likely location of a landing place serving the manor at
Falsgrave, which was itself not directly accessible from the sea.
Hedon
Hedon is situated in the south-west of Holderness, on the north bank of Hedon
Haven, a small creek flowing towards the Humber. More precisely, it lies at the
confluence of two becks, at what was presumably the limit of tidal influence and
the highest navigable point on the creek, as well as the location of a road crossing
(Allison 1984, 168; Kent 1984, 111; Slater 1985, 32, 37) (Figure 11.2). The earliest core
of the medieval town lay on Market Hill, a slight rise in a low-lying and, before the
onset of land reclamation in the 12th century, often flooded landscape (English 1979,
203–204; van de Noort 2004; Sheppard 1966; Slater 1985, 29–32).
The first recorded activity at Hedon occurs in ad 1115, when William of Aumale,
the lord of Holderness and owner of the nearby manor of Burstwick, granted a hos-
pitum to the monks of Aumale (Normandy) (English 1979, 213, 214). The earliest
archaeological evidence stems from the late 12th century, when Hedon had grown
into an important craft and trade centre (eg Slater 1985, 26–27, 34). Nevertheless,
MAM 37: 11-deckers – Press

the emergence of coastal urbanism 143

Figure 11.2 Map showing the locations of Hedon (approximate extent of the 12th-century
town shown by a dotted line) and Burstwick in relation to the hydrography of the area
(after Slater 1985, fig 1)

there are some indications that Hedon Haven was already a focus of maritime traffic
before then. There are 12th-century references to a ferry and beacon at Paull at the
mouth of Hedon Haven (English 1979, 214; Kent 1984, 114), and the Old English
place-name Paull refers to a ‘stake’ or landmark near the river (Craven 1972, 18;
Smith 1937, 36–37).
A more important argument for maritime activity is Hedon’s location in relation
with the local waterways and the manor of Burstwick. This Domesday estate centre
stood on a slightly raised gravel deposit about 3km further inland from Hedon along
one of the tributaries of Hedon Haven, Humbleton Beck (English 1979, 112, 206).
The latter waterway was made navigable in the second half of the 12th century
(Allison 1984, 168–169; English 1979, 213–214), implying that before that time,
Burstwick was inaccessible by boat.
Situated at the limit of navigability and at a road crossing, and linked to a not
directly accessible central place further inland, Hedon conforms to the proposed
settlement model. Clearly, and not surprisingly, the lords of Holderness chose an
already active landing place as the location for their principal harbour town.
The Lower Hull Valley
The somewhat enigmatic site of Wyke near the mouth of the River Hull ranks
as the sixth most lucrative harbour recorded in the ‘fifteenth’ of 1203–1204. Despite
this apparent commercial importance, its early history is virtually unknown. The
little written evidence from this period alludes to its involvement in the export of
wool and the import of wine, but contains no references to other central functions
beside trade (Allison 1969, 13; Gillett and MacMahon 1980, 2). Pre-Conquest finds
MAM 37: 11-deckers – Press

144 maritime societies of the viking and medieval world

from the area are negligible (Evans 2000, 200). The earliest archaeological evidence
from the present-day town area dates to the mid-13th century, and by the time
Edward I acquired the settlement from the monks of Meaux (renaming it Kingston-
upon-Hull) in ad 1293, it counted only about 60 households (Evans 2000, 202). This
discrepancy between its economic significance (implied by the ‘fifteenth’ records) and
its insignificance as a central place and settlement has given rise to the suggestion that
early Wyke was situated outside the medieval town area and served as a customs port
for all goods entering and leaving the Hull Valley, regardless of the actual location of
exchange (Evans 1997; D Evans pers comm).
That traffic on the Hull was already important in Middle Anglo-Saxon times can
be inferred from the presence in the Hull Valley not only of a royal vill (see above),
but all the more of Beverley, a central place of at least regional importance from the
8th century onwards. Founded as a monastery and (partly) abandoned in the mid-
9th century, Beverley was revived in the 10th and 11th centuries as a major settlement
performing commercial and artisanal as well as religious functions for a wide hinter-
land. Despite its increasing autonomy, the Archbishop of York maintained a strong
interest in the town (eg Miller et al 1982, 17, 19) because it acted as the principal
outlet for his Holderness estates. Evidence for the importation of overseas objects and
the consumption of marine foodstuffs is persistently present throughout the time
period under study, albeit in small quantities (eg Armstrong et al 1991; Evans and
Tomlinson 1992). Historically, the town was known for its wool industry, attracting
foreign merchants (Horrox 1989, 39). Given Beverley’s own religious significance and
its archiepiscopal connections, it may be assumed that wine was a principal import.
Situated about a mile away from the Hull banks, medieval Beverley was served
by two nearby landing places. Beckside came into use only after canalization of
Walker Beck connected it with the Hull in the first half of the 12th century (Miller
et al 1982, 18–19). Therefore, the riverside suburb of Grovehill, connected to Beverly
by a major road crossing the Hull Valley, is the more likely location of a pre-
Conquest landing place. A 1321 lawsuit states this was the closest large boats could
come to the town (Horrox 1989, 35–36; Miller et al 1982, 31).
There are few hints of a formal connection between Wyke and Beverley.
Evidence for a link between the two sites may be found in the fact that the Arch-
bishop, who held privileges relating to wine trade at Wyke by the late 12th century,
presumably also granted the harbour its borough status in ad 1239 (Beresford 1967,
509–510; Gillett and MacMahon 1980, 2). On the other hand, Wyke was not the
nearest navigable point for Beverley, and the case remains difficult to reconcile with
the model. A partial solution may be offered by the fact that, due to geomorpho-
logical developments (Metcalfe et al 2000), the Hull mouth and any associated land-
ing place were originally situated closer to Beverley (Evans 2000, 193; Sheppard 1958,
1–3). Furthermore, the Hull mouth was the best natural anchorage along the
Humber, a vital transport route connecting the Midlands with the North Sea (Evans
2000, 200). Some form of involvement in, or control over, activity here by the monks
and townspeople of Beverley may be presumed, given their economic dependence on
exported and imported goods. One may further hypothesize that, as the estuarine
landscape developed and the anchorage for Humber traffic shifted southwards, the
MAM 37: 11-deckers – Press

the emergence of coastal urbanism 145

inhabitants of Beverley attempted to preserve this vital link with the international
trade network.
Newcastle-upon-Tyne
A next case demonstrates the utility of the heuristic model as a useful frame-
work for the analysis of archaeological and historical information regarding the
Anglo-Saxon origins of Norman towns, even when this information is too limited to
draw definite conclusions (Figure 11.3). Newcastle-upon-Tyne was the location of a
Roman bridge forming the lowest crossing on the Tyne. Possibly in a ruined state in
the early Middle Ages, the bridge’s low arches presumably still hindered upstream
traffic (Harbottle and Clack 1976, 118). Therefore a landing place and transhipment
site may have been situated just downstream of the bridge. Here, the low-lying area at
the foot of the high river banks widened into tidal mudflats through which a small
tributary, Lort Burn, flowed into the Tyne. Following 12th-century reclamation
works (eg O’Brien et al 1989), this area became the core of the medieval harbour.
Newburn, about 11km upstream, is the first pre-Conquest central place that can
be linked to this landing place. This Northumbrian royal centre and later residence of
the earls of Northumbria lost its central status in the late 11th century. Fords at
Newburn provided a natural crossing point, but rivercraft plied this far at least until
the 17th century (Dodds 1930, 117–118, 130). The relation between Newburn and the
postulated landing place at Newcastle thus fits the model’s criteria.
Two further Anglo-Saxon central places have been proposed within the town
area, albeit on the basis of inconclusive evidence.2 Amongst other arguments, the
Anglo-Saxon origins of the church of St Andrew are hinted at by its dedication and the
presence of probable pre-Conquest cross-slab (Deckers 2007, 67; Honeyman 1941,
117–118; Ryder 2002, 90). Located at the northern edge of the medieval town,
St Andrew’s was linked to the postulated landing place by a road along the western
edge of Lort Burn’s valley — a route which became one of the medieval town’s prin-
cipal thoroughfares. In addition, the 7th- to 12th-century cemetery3 on Castle Hill (the
location of the Norman ‘new castle’) potentially points to the presence of (another?)
settlement of some importance. Whereas a centre at St Andrew’s is reconcilable with

Figure 11.3 Map of the Lower Tyne Valley, showing the principal places mentioned in the
text and the location of the Roman bridge at Newcastle. The approximate extent of the
medieval walled town of Newcastle is depicted by a dotted line (P Deckers)
MAM 37: 11-deckers – Press

146 maritime societies of the viking and medieval world

our model, a central place on Castle Hill, immediately overlooking the landing place,
seems a less easy fit. However, several comparable sites exist in Northumbria, and the
next section will reveal that these are not as deviant as may seem.

exceptions to the ‘rule’?

So far, it has been argued that in a significant proportion of cases the North-
umbrian shore is characterized by physical separation between centre and landing
place. However, some of the best-known central places of early medieval Northum-
bria are set in a littoral location.
Elevated sites
Many of these ‘exceptional’ sites, including the royal residence at Bamburgh and
the majority of the monasteries (including Whitby, Hartlepool, Monkwearmouth
and Tynemouth) are situated on top of headlands or in a similar elevated position,
looking out over the sea. A closer look at the monastery of Tynemouth illustrates
that the model retains some of its validity. The monastery lies on the northern bank
of the mouth of the River Tyne, on a 30m high headland protruding into the sea.
Excavations brought to light only limited material remains of the Anglo-Saxon occu-
pation (Fairclough 1983; Jobey 1967). Its involvement in maritime activity is difficult
to assess, but at least overseas contacts with nearby monasteries can be presumed.
Given its elevated position, access to the monastery must have been mediated by
a distinct landing place at the foot of the cliffs. Makepeace (1995, 35) suggests that a
sandy cove on the south side of the headland served this purpose. This was far from
an ideal situation: besides being useable only in fair weather, the goods and people
disembarked here faced a steep climb towards the top of the headland.
The small inlet known on later maps as ‘Low Lights’, almost a mile further
inland along the Tyne, probably acted as the monastery’s principal landing place. It
was the first safe anchorage when entering the Tyne, offering greater and less
weather-dependent safety than the aforementioned cove. The original main road of
the Norman village formed a direct connection between the headland and the inlet
(Hart 1997, 101–104), allowing for a less steep, easier ascent.
York
Anglo-Saxon York was the political and commercial heart of northern England.
Although not actually coastal, it was a supra-regional central place situated at the
confluence of the rivers Foss and Ouse, immediately accessible to sea-going ships, and
therefore warrants discussion.
York is usually described as a town, meeting the traditional criteria (eg Biddle
1976, 99–100) used to define urban settlements. As elsewhere in early medieval
Europe, the presence of a Roman substratum greatly determined the further develop-
ment of the settlement, both through the availability of certain infrastructural
elements (notably, the Roman defences) and through the reproduction of the concept
of urbanism (Biddle 1976, 103–112; Carver 1997, 382–384).
MAM 37: 11-deckers – Press

the emergence of coastal urbanism 147

Nonetheless, this impression of nucleation and centralization must be nuanced.

Middle Saxon York consisted of two distinct parts: the old, walled archiepiscopal
centre and an extramural trading site (Hall 1999). The latter was separated from the
former not only by the Roman walls, but also by the River Foss. Only in the
9th century, when the Middle Saxon emporium was abandoned and a new trading
site appeared adjacent to the walled area, the term ‘urban’ is rightly applicable to
York. In this bipolar layout, which occurs at other emporia too (Hodges 2000, 86–87;
Hill 2001, 79; McCormick 2007, 47), the two main features of the model — physical
separation and functional differentiation between a commercial and a governmental
focus — are clearly discernible. Although not inaccessible to sea-going traffic, the
episcopal see of Middle Saxon York was separated from the trading site by the River
Foss and the Roman wall.
Comparison with the elevated sites mentioned above raises an intriguing pos-
sibility. Was it the availability of a (naturally or artificially) defensible position that
allowed a central place to be littoral? Perhaps central places in what would otherwise
be exposed littoral locations required a degree of (cognitive and/or physical) separa-
tion and protection, thus necessitating the mediating presence of an external landing
place.
Jarrow
Lastly, a few littoral central places defy the model altogether. These sites are
directly accessible to seaborne traffic, and thus lack a distinct landing place through
which access is guided. One such site, the monastery of Jarrow, is briefly examined
here.
The monastery stands on a low hill overlooking the Slake, a tidal mudflat on
the southern banks of the Tyne estuary. The River Don flows along the southern
boundary of the complex, access to which was provided by a bridge (Cramp 2005,
8–13). Imports (Cramp 2005, 343–344) and small quantities of fish remains demon-
strate the monastery’s involvement in maritime activity (Jones and Hutchinson 2006,
576), although the intensity of that activity should not be overestimated (Jenner et al
2006, 407).
The majority of objects pointing to exchange (eg coins and Continental pottery)
were found in the south of the complex, on the slope towards the River Don, which
formed the entry point to the monastery. The most southerly stone structure, build-
ing D, was presumably used as a guesthouse, conveniently situated near the riverside
landing place (Cramp 2005, 219–241). Although less well known, the monastic settle-
ment on Lindisfarne (O’Sullivan and Young 1995) arguably had a comparable lay-
out. The landing places of these monasteries formed an integral part of the monastery
complex, and thus did not conform to the model.

explanations
Why so many central places were set back or separated from the waterfront and
required ‘mediation’ by an external landing place remains an unanswered question
thus far. Several factors potentially contributed to this recurring settlement pattern.
MAM 37: 11-deckers – Press

148 maritime societies of the viking and medieval world

The natural environment is, of course, a prime factor determining settlement loca-
tion. Suitable locations for permanent settlement in the low-lying East Riding of
Yorkshire were limited to high ground, whereas small-scale landing places required
only periodically dry land. This may (partly) explain the settlement pattern in the
Lower Hull Valley. Conversely, the natural topography factored less in the spatial
relation between Burstwick and Hedon, as both were situated on raised gravel
deposits. Inland central places may have been favoured for a combination of environ-
mental and economic reasons, such as the maximization of access to agricultural and
other resources (eg Ulmschneider 2005, 523) and integration into regional transport
networks. Beverley, for instance, was located near one of the few easy crossings on
the River Hull.
Additionally, central settlements may have been set back from the shore for
defensive reasons. In literary support of this obvious point, the Anglo-Saxon poem
Beowulf recounts that, upon landing on the Danish coast, the hero and his entourage
were confronted by a guard on the sea-cliff before being allowed to proceed inland to
the King’s mead hall (Heaney 1999). The setting of littoral central places on top of
sea-cliffs or behind Roman walls may have been motivated by the same need for
protection against and/or early warning of seaborne attackers, although this was
clearly not very effective against the Viking raids of the late 8th and 9th centuries.
Moreover, the coastal settlement configuration observed in Northumbria may
have wider relevance as the expression of a more general aversion regarding the sea.
For instance, sea fishing as an economic activity was negligible before ad c1000
(Barrett et al 2004). Early medieval literature usually associates the sea and the
coastal landscape with danger and evil (eg Corbin 1988, 11–32; McCormick 2001,
402–404; Rippon 2009). Closely related is the Gaelic Christian notion of ‘desertum in
oceano’, inspiring monks to deliberately seek danger and hardship by establishing
their hermitages on small isolated islands and in other exposed locations (eg Dumville
2002). Although this is a subject that requires further research, these observations
feed the impression that the sea had many negative connotations in the early medi-
eval mind. These connotations may well have affected decisions regarding settlement
location and landscape organization, consciously or subconsciously, thus contribut-
ing to the observed pattern in the arrangement of maritime cultural landscapes in
Northumbria and beyond.
Despite these many factors favouring inland central places, a few sites were
definitely ‘atypical’. Why were Jarrow and Lindisfarne built in such exposed loca-
tions? The choice was presumably influenced by the patrons donating the land.
Alternatively, the Gaelic eremitic ideology that also affected the locations of other
coastal Northumbrian monasteries was of primary importance to the founders of
Lindisfarne and Jarrow. Or perhaps the placement of some of these monasteries was
determined by their postulated function as navigation beacons (Rainbird 2007, 55)?

urbanization and the transformation of the littoral landscape

The emergence of Northumbrian harbour towns from the late 11th century
onwards signified a radical change in the coastal settlement pattern. Many of the
MAM 37: 11-deckers – Press

the emergence of coastal urbanism 149

functions previously concentrated around the inland centres were transferred to the
landing places, which became the sole foci of administration, jurisdiction, religion,
commerce and artisanal production. As a result, the former centres lost their impor-
tance and fell into decline. This process of relocation and centralization, as Hutcheson
(2006, 74) terms comparable developments in 11th-century King’s Lynn, was a gradual
and differential process, only intermittently illuminated by the available historical
evidence.
Falsgrave was still a base for royal officials under Henry II (ad 1154–89). The
soke was last mentioned in ad 1190, but had expired by ad 1256, when the manor
was granted to the townspeople of Scarborough (Pearson 2001, 80; Rowntree 1931;
Russell 1923, 554). The shift of centrality from Burstwick to Hedon was less linear,
because the lords of Holderness established their principal residence and military
strongpoint at Burstwick in the late 12th or early 13th century. The estate centre also
continued to fulfil a judicial role as wapentake court (Dalton 2001, n45; English 1979,
36, 112, 206). Nonetheless, Hedon was the economic and administrative centre of
the wapentake, as sheriff’s seat and a (short-lived) mint. The wapentake court also
convened in the town occasionally, and was finally transferred from Burstwick in the
13th century (Allison 1984, 174–175; English 1979, 72–73, 112; Jones 1949).
The urbanization of Wyke, starting only in the 13th century, is harder to under-
stand in these terms. The process began with the establishment of an agricultural
grange by the monks of Meaux (an abbey founded ad c1150 about 12km to the
north), followed by drainage and reclamation works (Burton 1999, 121, 262–263). A
chapel is mentioned around ad 1200 (Beresford 1967, 169–170). Perhaps the landing
place, which previously shifted continually through the unstable landscape, now for
the first time was formalized in a fixed location. Notwithstanding its considerable
commercial importance, the town failed to attract other central functions. The emer-
gence of Wyke could thus best be described as a change in scale, rather than in nature,
of the landing place. Increasing Humber traffic, as well as the growing output from
Beverley and Meaux, brought about expanding activity at the landing place, which
enabled the establishment of a small permanent settlement slowly acquiring a limited
number of urban characteristics. At least until the end of the 13th century, the early
medieval configuration prevailed here: inland centres (Beverley and Meaux) served by
a separate, albeit elaborate, landing place/transhipment site.
Elsewhere, the spatial shift of centrality occurred even later. Along the North
Yorkshire cliff coast, many Norman castles (eg Mulgrave and Skelton) were built
at the locations of early medieval estate centres, rather than at associated landing
places (l’Anson 1913; Pearson 1999, 162). Even in fully developed harbour towns the
dichotomy between landing place and centre, commerce and government was some-
times reproduced in a cognitive sense. In Scarborough, the castle and the parish
church of St Mary’s, the material expressions of secular and religious authority, are
elevated above and perceptually distinct from the less-monumental harbour. At
Newcastle, the Norman castle on Castle Hill and the parish church of St Nicholas
on the high banks of the Tyne similarly dominated the reclaimed harbour area of
Sandhill below. This layout, not unlike the Middle Saxon coastal monasteries of
MAM 37: 11-deckers – Press

150 maritime societies of the viking and medieval world

Northumbria, can be interpreted as the reproduction of old notions of landscape

organization within the new socio-economic context of the medieval town.

conclusions
The cases presented in this paper, with a few notable exceptions, all display
features that may be considered constants in the coastal settlement of early medieval
Northumbria: spatial separation and functional differentiation between centre and
landing place. Although most rural central places were set back from the coast,
excavations have demonstrated that these centres were involved in maritime activity.
Therefore they must have been served by landing places on the waterfront. These are
difficult to detect archaeologically, but a heuristic model is proposed to aid the iden-
tification of likely locations. This model postulates that landing places were situated
as near to the centre as the navigability of waterways allowed, and thus served as
transhipment points.
Mainly applicable to rural central places, the model’s basic features are also
recognizable in the placement of a group of central places in a decidedly littoral
setting. Notable among these are the coastal monasteries, which were mostly set on
cliff-tops or high headlands overlooking the sea and therefore required a landing
place down near the waterfront. Similarly, the Middle Saxon emporium of York was
a distinct settlement, detached from the centre by Roman walls and the River Foss.
The avoidance of littoral locations for permanent settlement had diverse causes,
including the reliance on inland resources, the need for integration with hinterland
networks and the threat of overseas attack. In addition to such mundane consider-
ations, however, one might wonder whether there existed a broader cultural aversion
towards the sea, with its connotations of danger and evil. Only a few sites, notably
the coastal monasteries of Jarrow and Lindisfarne, were different. Here the landing
places were directly integrated into the settlement complexes, possibly due to the
donation policies of royal patrons or to a stricter adherence to eremitic ideologies.
‘The urban community of Northern Europe was a product of two different
developments . . .: that of the centre and that of the trading place’ (Callmer 1994, 52).
This statement perfectly summarizes the process that led to the emergence of the medi-
eval harbour towns of Northumbria, a process that entailed the nucleation of central
functions and a shift of centrality towards formerly peripheral landing places. Simul-
taneously, early medieval centres lost their importance and declined. Although certain
features could be reproduced in the new towns, the coalescence of centre and landing
place represented a drastic departure from the old maritime cultural landscape.4

notes
1 The important estate centres (with three soke both known from historical sources to have lain
lands or more) of the coastal wapentakes of in the vicinity of later Newcastle, remains un-
North Yorkshire and the East Riding are all set certain (Graves 2002, 178; Harbottle and Clack
back from the shore, with the sole exception of 1976, 117).
Whitby. 3 Short notes may be found in the ‘Medieval Britain
2 The relation of these with the monastery of and Ireland’ sections of Medieval Archaeology 23
Monkchester and the royal estate of Ad Murum, (1979), 26 (1982), 27 (1983) and 37 (1993).
MAM 37: 11-deckers – Press

the emergence of coastal urbanism 151

4 This paper began with my MA dissertation at the Trevor Pearson (English Heritage) and David
University of York (Deckers 2007). Thanks are Evans (Humberside Archaeology Partnership) for
due to James Barrett (my former supervisor), their suggestions and support.

references
Allison, K, 1969 ‘Medieval Hull’, in K Allison (ed), A History of the County of York, East Riding,
Volume 1: The City of Kingston upon Hull, Oxford University Press, Oxford, 11–89
Allison, K, 1984 ‘Hedon’, in K Allison (ed), A History of the County of York, East Riding, Volume
5: Holderness: Southern Part, Oxford University Press, Oxford, 168–186
L’Anson, W, 1913 ‘The castles of the North Riding’, Yorkshire Archaeological Journal 22,
338–399
Armstrong, P, Tomlinson, D and Evans, D, 1991 Excavations at Lurk Lane Beverley, J R Collis,
Beverley
Barrett, J H, Locker, A M and Roberts, C M, 2004 ‘‘‘Dark Age economics’’ revisited: the English
fish bone evidence ad 600–1600’, Antiquity 78, 618–636
Bell, T, 1998 ‘A Roman signal station at Whitby’, Archaeological Journal 155, 303–322
Beresford, M, 1967 New Towns of the Middle Ages: Town Plantation in England, Wales and
Gascony, Lutterworth Press, London
Biddle, M, 1976 ‘Towns’, in D Wilson (ed), The Archaeology of Anglo-Saxon England, Methuen
& Co, London, 99–150
Burton, J, 1999 The Monastic Order in Yorkshire 1069–1215, Cambridge University Press,
Cambridge
Callmer, J, 1994 ‘Urbanization in Scandinavia and the Baltic region c. ad 700–1100: trading places,
centres and early urban sites’, in B Ambrosiani and H Clarke (ed), The Twelfth Viking
Congress: Developments Around the Baltic and the North Sea in the Viking Age, Birka
Studies 3, Riksantikvarieämbetet and Statens Historiska Museer, Stockholm, 50–90
Carver, M, 1997 ‘Town and anti-town in the first millennium ad’, in G De Boe and F Verhaeghe
(eds), Urbanism in Medieval Europe: Papers of the ‘Medieval Europe Brugge 1997’ Confer-
ence, Volume 1, I A P Rapporten, I A P, Zellik, 379–389
Clarke, H and Ambrosiani, B, 1991 Towns in the Viking Age, Leicester University Press, Leicester
Corbin, A, 1988 Le territoire du vide: l’Occident et le désire du rivage, Abier, Paris
Cramp, R, 2005 Wearmouth and Jarrow Monastic Sites, English Heritage, Swindon
Craven, M, 1972 A New and Complete History of the Borough of Hedon, Ridings Publishing,
Driffield
Dalton, P, 2001 ‘The foundation and development of Scarborough in the twelfth century’, in
A D Crouch (ed), Medieval Scarborough: Studies in Trade and Civic Life, Yorkshire Archaeo-
logical Society Occasional Paper 1, Yorkshire Archaeological Society and Scarborough
Archaeological and Historical Society, Leeds, 1–6
Davies, G, 2010 ‘Early medieval ‘‘rural centres’’ and west Norfolk: a growing picture of diversity,
complexity and changing lifestyles’, Medieval Archaeology 54, 89–122
Deckers, P, 2007 The maritime cultural landscape of Northumbria, ad 700–1200: early medieval
settlement configuration and the emergence of coastal urbanism, Unpublished MA disserta-
tion, University of York, York
Dent, J, Loveluck, C and Fletcher, W, 2000 ‘The early medieval site at Skerne’, in R van de Noort
and S Ellis (ed), Wetland Heritage of the Hull Valley: an Archaeological Survey, Humber
Wetlands Project, Kingston upon Hull, 217–242
Dodds, M, 1930 A History of Northumberland, Volume XIII, Andrew Reid, Newcastle-upon-
Tyne
MAM 37: 11-deckers – Press

152 maritime societies of the viking and medieval world

Dumville, D, 2002 ‘The North Atlantic monastic thalassocracy: sailing to the desert in early medi-
eval insular spirituality’, in B Crawford (ed), The Papar in the North Atlantic: Environment
and History, St John’s House Papers 10, The Committee for Dark Age Studies, University of
St Andrews, St Andrews, 121–130
English, B, 1979 The Lords of Holderness 1086–1260: A Study in Feudal Society, Oxford
University Press, Oxford
Evans, D, 1997 ‘Archaeological work in the medieval port of Kingston-upon-Hull’, in M Gläser
(ed), Lübecker Kolloquium zur Stadtarchäologie im Hanseraum, Volume 1: Stand, Aufgaben
und Perspektiven, Schmidt-Römhild, Lübeck, 35–49
Evans, D, 2000 ‘Archaeology in the modern city of Kingston upon Hull, and recent research at
Kingswood’, in R van de Noort and S Ellis (ed), Wetland Heritage of the Hull Valley: an
Archaeological Survey, Humber Wetlands Project, Kingston upon Hull, 193–216
Evans, D and Tomlinson, D, 1992 Excavations at 33–35 Eastgate, Beverley, 1983–86, J R Collis,
Sheffield
Fabech, C, 1999 ‘Organising the landscape: a matter of production, power and religion’, in
T Dickinson and D Griffiths (ed), The Making of Kingdoms, Anglo-Saxon Studies in
Archaeology and History 10, Oxford University School of Archaeology, Oxford, 37–47
Fairclough, G, 1983 ‘Tynemouth Priory and Castle: excavation in the outer court, 1980’,
Archaeologia Aeliana (5th series) 11, 101–131
Ferguson, C, 2011 ‘Re-evaluating Early Medieval Northumbrian Contacts and the ‘‘Coastal High-
way’’’, in D Petts and S Turner (ed), Early Medieval Northumbria. Kingdoms and Com-
munities, ad 450–1100, Studies in the Early Middle Ages 24, Brepols, Turnhout, 283–302
Gardiner, M, Cross, R, MacPherson-Grant, N and Riddler, I, 2001 ‘Continental trade and non-
urban ports in Mid-Anglo-Saxon England: excavations at Sandtun, West Hythe, Kent’,
Archaeological Journal 158, 161–290
Gillett, E and MacMahon, K, 1980 A History of Hull, Oxford University Press, Oxford
Graves, C, 2002 ‘The development of towns in the north’, in C Brooks, R Daniels and A Harding
(ed), Past, Present and Future: The Archaeology of Northern England, Research Report 5,
Architectural and Archaeological Society of Durham and Northumberland, Durham, 177–184
Hall, R, 1984 The Viking Dig: The Excavations at York, The Bodley Head, London
Hall, R, 1999 ‘York 700–1050’, in C Karkov (ed), The Archaeology of Anglo-Saxon England: Basic
Readings, Garland Publishing, New York, 119–136
Hall, R, 2000 ‘Excavation at 16–22 Coppergate’, in A Mainman and N Rogers (ed), Craft, Industry
and Everyday Life: Finds from Anglo-Scandinavian York, Council for British Archaeology,
York, 2455–2461
Harbottle, B and Clack, P, 1976 ‘Newcastle upon Tyne: archaeology and development’, in P Clack
and P Gosling (ed), Archaeology in the North: Report of the Northern Archaeological Survey,
HMSO, Newcastle, 111–131
Hart, C, 1997 ‘Excavations in Tynemouth, 1995’, Archaeologia Aeliana (5th series) 25, 87–108
Heaney, S (ed), 1999 Beowulf, Faber, London
Hill, D, 2001 ‘End piece: definitions and superficial analysis’, in D Hill and R Cowie (ed), Wics:
The Early Medieval Trading Centres of Northern Europe, Sheffield Archaeological
Monographs 14, Sheffield Academic Press, Sheffield, 75–84
Hill, D and Cowie, R (ed), 2001 Wics: The Early Medieval Trading Centres of Northern Europe,
Sheffield Archaeological Monographs 14, Sheffield Academic Press, Sheffield
Hodges, R, 1982 Dark Age Economics: The Origins of Towns and Trade ad 600–1000, Duck-
worth, London
Hodges, R, 2000 Towns and Trade in the Age of Charlemagne, Duckworth, London
MAM 37: 11-deckers – Press

the emergence of coastal urbanism 153

Honeyman, S, 1941 ‘The Church of St Andrew, Newcastle upon Tyne’, Archaeologia Aeliana
(4th series) 19, 117–161
Horrox, R, 1989 ‘Medieval Beverley’, in K Allison (ed), A History of the County of York, East
Riding, Volume 6: The Borough and Liberties of Beverley, Oxford University Press, Oxford,
2–62
Huntley, J and Stallibrass, S, 1995 Plant and Vertebrate Remains from Archaeological Sites in
Northern England: Data Reviews and Future Directions, Architectural and Archaeological
Society of Durham and Northumberland, Durham
Hutcheson, A, 2006 ‘The origins of King’s Lynn? Control of wealth on the Wash prior to the
Norman Conquest’, Medieval Archaeology 50, 71–104
Ilves, K, 2009 ‘Discovering harbours? Reflection on the state and development of landing site
studies in the Baltic Sea region’, Journal of Maritime Archaeology 4, 149–63
Jenner, A, Mills S and Burke, B, 2006 ‘The Anglo-Saxon and medieval pottery’, in R Cramp (ed),
Wearmouth and Jarrow Monastic Sites, Volume 2, English Heritage, Swindon, 327–428
Jobey, G, 1967 ‘Excavation at Tynemouth Priory and Castle’, Archaeologia Aeliana (4th series) 45,
33–104
Jones, A and Hutchinson, A, 2006 ‘The fish remains from Jarrow’, in R Cramp (ed), Wearmouth
and Jarrow Monastic Sites, Volume 2, English Heritage, Swindon, 575–581
Jones, F, 1949 ‘Hedon near Hull — a new Norman mint’, British Numismatic Journal 26, 28–30
Kemp, R, 1996 Anglian Settlement at 46–54 Fishergate, Council for British Archaeology, York
Kent, G, 1984 ‘Paull’, in K Allison (ed), A History of the County of York, East Riding, Volume 5:
Holderness: Southern Part, Oxford University Press, Oxford, 111–127
Leifsson, R, 2005 The Skerne bones: processual means to interpretive ends, Unpublished MA
dissertation, University of York, York
Lloyd, T, 1977 The English Wool Trade in the Middle Ages, Cambridge University Press,
Cambridge
Loveluck, C, 1996 ‘The development of the Anglo-Saxon landscape, economy and society ‘‘on
Driffield’’, East Yorkshire’, in D Griffiths (ed), Anglo-Saxon Studies in Archaeology and
History, Volume 9, Oxford University Committee for Archaeology, Oxford, 25–48
Loveluck, C and Tys, D, 2006 ‘Coastal societies, exchange and identity along the Channel and
southern North Sea shores of Europe, ad 600–1000’, Journal of Maritime Archaeology 1,
140–169
Loveluck, C, Dobney, K, and Barrett, J, 2007 ‘Trade and exchange: the settlement and the wider
world’, in C Loveluck (ed), Rural Settlement, Lifestyles and Social Change in the Later First
Millennium ad at Flixborough, Lincolnshire: Anglo-Saxon Flixborough in its Wider Context,
Excavations at Flixborough 4, Oxbow Books, Oxford, 99–111
Mägi, M, 2004 ‘‘‘. . . Ships are their main strength.’’ Harbour sites, arable lands and chieftains on
Saaremaa’, Estonian Journal of Archaeology 8, 128–155
Makepeace, J, 1995 Early medieval harbours in Northumbria, Unpublished MA dissertation,
University of York, York
Manby, T, 1988 ‘Excavation and field archaeology in eastern Yorkshire: the Thwing project
1973–1987’, CBA Forum 1988, 16–18
McCormick, M, 2001 Origins of the European Economy: Communications and Commerce,
ad 300–900, Cambridge University Press, Cambridge
McCormick, M, 2007 ‘Where do trading towns come from? Early medieval Venice and the
northern emporia’, in J Henning (ed), Post-Roman Towns, Trade and Settlement in Europe
and Byzantium, vol. 1: The Heirs of the Roman West, Walter de Gruyter, Berlin, 41–68
McGrail, S, 2001 Boats of the World: From the Stone Age to Medieval Times, Oxford University
Press, Oxford
MAM 37: 11-deckers – Press

154 maritime societies of the viking and medieval world

Metcalfe, S E, Ellis, S, Horton, B P, Innes, J B, McArthur, J, Mitlehner, A, Parkes, A, Pethick, J S,

Rees, J, Ridgway, J, Rutherford, M M, Shennan, I and Tooley, M J, 2000 ‘The Holocene
evolution of the Humber estuary: reconstructing change in a dynamic environment’, in
I Shennand and J E Andrews (ed), Holocene Land–Ocean Interaction and Environmental
Change around the North Sea, Geological Society, London, 97–118
Miller, K, Robinson, J, English, B and Hall, I, 1982 Beverley: An Archaeological and Architectural
Study, HMSO, London
Naylor, J, 2004 An Archaeology of Trade in Middle Saxon England, British Archaeological
Reports, British Series 376, Archaeopress, Oxford
van de Noort, R, 2004 The Humber Wetlands: The Archaeology of a Dynamic Landscape,
Windgather Press, Macclesfield
O’Brien, C, Brown, L, Dixon, S, Donel, L, Gidney, L, Huntley, J P, Nicholson, R and Walton, P,
1989 ‘Excavations at Newcastle quayside: the Crown Court site’, Archaeologia Aeliana
(5th series) 17, 141–205
O’Sullivan, D and Young, R, 1995 Lindisfarne Holy Island, B T Batsford/English Heritage, London
Parkhouse, J, 1997 ‘The distribution and exchange of Mayen lava quernstones in early medieval
northwestern Europe’, in G De Boe and F Verhaeghe (ed), Exchange and Trade in Medieval
Europe: Papers of the ‘Medieval Europe Brugge 1997’ Conference, Volume 3, I A P Rapporten,
I A P, Zellik, 97–106
Pearson, T, 1999 ‘Early medieval settlement on the Yorkshire coast’, Northern Archaeology 17/18,
159–164
Pearson, T, 2001 ‘Falsgrave soke and settlement’, in A D Crouch (ed), Medieval Scarborough:
Studies in Trade and Civic Life, Yorkshire Archaeological Society Occasional Paper 1,
Yorkshire Archaeological Society and Scarborough Archaeological and Historical Society,
Leeds, 79–84
Pearson, T, 2005 The Archaeology of Medieval Scarborough: Excavation and Research 1987–
2004, Scarborough Archaeological and Historical Society, Scarborough
Pestell, T and Ulmschneider, K (ed), 2003 Markets in Early Medieval Europe: Trading and
‘Productive’ Sites, 650–850, Windgather Press, Macclesfield
Rainbird, P, 2007 The Archaeology of Islands, Cambridge University Press, Cambridge
Richards, J, 1994 ‘An incised chalk weight from Cottam’, Medieval Archaeology 38, 167–169
Rippon, S, 2009 ‘‘‘Uncommonly rich and fertile’’ or ‘‘not very salubrious’’? The perception and
value of wetland landscapes’, Landscapes 10, 39–60
Rowntree, A, 1931 ‘The borough’, in A Rowntree (ed), The History of Scarborough, J M Dent &
Sons, London, 120–140
Russell, A, 1923 ‘The Borough of Scarborough’, in W Page (ed), The Victoria History of the
County of York, North Riding, Volume 2, Constable, London, 538–560
Ryder, P, 2002 ‘Medieval cross slab grave covers in Northumberland, part 2: Newcastle and south
east Northumberland’, Archaeologia Aeliana (5th series), 30, 75–137
Sheppard, J, 1958 The Draining of the Hull Valley, The East Yorkshire Local History Society,
York
Sheppard, J, 1966 The Draining of the Marshlands of South Holderness and the Vale of York, The
East Yorkshire Local History Society, York
Sindbæk, S, 2007 ‘Networks and nodal points: the emergence of towns in early Viking Age
Scandinavia’, Antiquity 81, 119–132
Sindbæk, S, 2009 ‘Open access, nodal points, and central places: maritime communication and
locational principles for coastal sites in south Scandinavia, c. ad 400–1200’, Estonian Journal
of Archaeology 13, 96–109
MAM 37: 11-deckers – Press

the emergence of coastal urbanism 155

Slater, T, 1985 ‘Medieval new town and port: a plan-analysis of Hedon, East Yorkshire’, Yorkshire
Archaeological Journal 57, 23–41
Smith, A, 1937 The Place-names of the East Riding of Yorkshire and York, Cambridge University
Press, Cambridge
Thomas, G, 2013 ‘Life before the minster: the social dynamics of monastic foundation at Anglo-
Saxon Lyminge, Kent’, The Antiquaries Journal 93, 109–145
Ulmschneider, K, 1999 ‘Archaeology, history, and the Isle of Wight in the Middle Saxon Period’,
Medieval Archaeology 43, 19–44
Ulmschneider, K, 2000 ‘Settlement, economy and the ‘‘productive’’ site: Middle Anglo-Saxon
Lincolnshire ad 650–780’, Medieval Archaeology 44, 53–79
Ulmschneider, K, 2005 ‘The study of early medieval markets: are we rewriting the economic
history of Middle Anglo-Saxon England?’, in H Hüssler, Neue Forschungsergebnisse zur
nordwesteuropäischen Frühgeschichte unter besonderer Berücksichtigung der altsächsischen
Kultur im heutigen Niedersachsen, Isensee, Oldenburg, 517–531
Ulriksen, J, 1994 ‘Danish sites and settlements with a maritime context, ad 200–1200’, Antiquity
68, 797–811
Verhulst, A, 2000 ‘Roman cities, emporia and new towns’, in I Hansen and C Wickham (ed), The
Long Eighth Century, Brill, Leiden, 105–120
MAM 37: 12-skre – Press

chapter 12
POST-SUBSTANTIVIST PRODUCTION
AND TRADE: SPECIALIZED SITES FOR
TRADE AND CRAFT PRODUCTION IN
SCANDINAVIA ad c700 – 1000

By Dagfinn Skre

When Karl Polanyi published his first book The Great Transformation in 1944 he
was 60 years old. He had served and been imprisoned in the First World War, he had
seen the economic depression of the 1930s, and as he was writing the book the world
was falling apart again in the terrors of the Second World War. One may see the book
as his attempt to answer the question: ‘Why did it all go so terribly wrong?’ In his
early years Polanyi was strongly influenced by Karl Marx’s theory on capitalist
society, although never becoming a true Marxist. Polanyi never believed in the deter-
minism built into Karl Marx’s economic theory. Politically, he is best characterized as
a humanist socialist. This background directed his search for the reasons for the
disasters of the early 20th century. It was all because of the market economy. In The
Great Transformation he developed the theory that the role of economy in pre-
capitalist society was drastically different from that which it had in capitalist society.
Capitalism is a historical anomaly because, while the economy in pre-capitalist
society was ‘embedded’ in social relations, in capitalism, the situation was reversed
— social relations were defined by economic relations.
In Polanyi’s view, before capitalism, transactional forms like reciprocity and re-
distribution were far more frequent than market transactions. However, capitalism
destroyed these other forms irreversibly and market transactions became totally domi-
nant. The ‘great transformation’ of the Industrial Revolution was to completely
replace all previous forms of transactions with market exchange. The economy
became separated from society and established its own reality, no longer serving the
needs and interests of people. The two World Wars and the economic recession
which Polanyi had seen and experienced were the results of this new role of the
economy.
From Marx he inherited the notion that mankind is capable of living in harmony
— actually, that is our natural disposition. His apparent need to find a point in time
where it all went wrong may have led him to ‘purify’ and glorify earlier periods. If
MAM 37: 12-skre – Press

post-substantivist production and trade 157

Polanyi’s theory was an attempt to understand the problems of his own time, one
should perhaps be somewhat cautious in using his substantivist theory as a tool for
understanding early or prehistoric societies.
When, in connection with the Kaupang project, I was working on economy and
trade in the 8th–10th centuries, I felt a need to explore the reasons for the discomfort
which I felt with Polanyi’s substantivist approach (Skre 2007; 2008). Was it one of the
rather numerous theories which disguises in the garments of the past a more or less
romantic dream about the present and future? Was it in reality a critique of Western
civilization and not a theory about the economy of ancient societies? The fruits of this
exploration will be presented in the following. First I will discuss some theoretical
issues which led to an alternative to Polanyi’s substantivist approach. Thereafter I
will apply this alternative approach to trade and economy in the maritime markets
and towns of 8th- to 10th-century Scandinavia.

the post-substantivist approach

It is truly remarkable that, even though Polanyi declared formalist economic
theory to be irrelevant to understanding pre-capitalist societies, he adopted the
formalist view of the market and market trade, even if it occurred in pre-capitalist
situations. Polanyi’s idea of the market is completely anti-contextual. The market
obeys only mechanical economic laws, such as that of supply and demand, and is free
of influence from social relations (Polanyi 1957, 256–257). Market transactions
are undertaken solely with a view to economic profit. Just like formalist economists,
Polanyi bases his understanding of the market on one single proposition concerning
humanity: that individuals follow a purely economic rationality in trying to satisfy
their preferences.
This, in reality, rather metaphysical premise has been challenged over the past
few decades by anthropologists, economists and sociologists who have been explor-
ing the territory between those three disciplines. Through empirical analysis and
theoretical discussions they have shown that this assumption about human economic
rationality is inadequate for understanding not only the economies of pre-capitalist
societies, but also the economies of modern societies. As the sociologists Richard
Swedberg and Mark Granovetter have written (1992, 9), with reference to modern
society: ‘Economic action is socially situated and cannot be explained by reference to
individual motives alone. It is embedded in ongoing networks of personal relation-
ships rather than being carried out by atomized actors.’ Swedberg and Granovetter
thus make use of Polanyi’s term ‘embedded’ to characterize the modern market
economy. The same or similar terms have been employed by a number of other
scholars within sociology, economics and anthropology (eg Davis 1992; Granovetter
1985; Gudeman 1986, 44; 2001; Hirsch et al 1987; Lie 1991; 1992, 509; 1997, 347–348;
Swedberg and Granovetter 1992; Wilk 1996; Zafirovski 2001).
Based on this research the notion of an ‘embedded’ economy cannot be reserved
for the economies of pre-capitalist societies alone, as Polanyi would have it. The
foundations of his sharp distinction between the market economies of capitalist
societies and the embedded economies of pre-capitalist societies therefore vanish.
MAM 37: 12-skre – Press

158 maritime societies of the viking and medieval world

This distinction was the main plank of Polanyi’s argument that an economic ratio-
nality is of no significance in pre-capitalist societies, and that economic dealings in
those contexts have to be understood with reference not to economic mechanisms,
but to social relations alone.
Ancient transactions and socio-economic formations may then, just like con-
temporary ones, be understood in both social and economic terms. The sociologist
John Lie (1997, 350) characterizes this approach thus:
In avoiding the oversocialized (eg the substantivist school in economic anthro-
pology) and undersocialized (eg the economic approach) approaches, it seeks to
strike a correct balance in analyzing markets and other economic phenomena and
institutions.
For Polanyi, the free determination of price on the basis of supply and demand is
diagnostic of market trade. However, empirical analyses of modern markets clearly
demonstrate that, since these markets are embedded in society, the determination
of price never takes place completely independently of the social context. In pre-
capitalist as well as in modern societies, all exchanges of goods will also have an
economic element, in other words, an element of the type that Polanyi finds only in
market exchange. An economic way of thinking therefore cannot be associated with
modern society alone, as Polanyi would have it. It was a factor in any transaction
in any period, including the Viking Age and Middle Ages (cf Barrett et al 2000, 15;
2004, 631).
This does not in any way mean that the mechanisms of the market economy,
such as determination of price by supply and demand, play an equally significant role
in every transaction in both modern and pre-capitalist societies. The social circum-
stances will always constrain the impact of mechanisms of this kind. As Swedberg
and Granovetter put it (1992, 10, their italics), the ‘level of embeddedness [of the
economy] varies considerably — both in industrial and pre-industrial societies’.
A number of the fundamental concepts of substantivism are challenged by this
view: for instance, the distinction drawn by Polanyi between market trade and
reciprocity. Reciprocal exchanges — exchanges between parties of equal status at
normalized prices, which should appear within closely integrated societies (Polanyi
1957, 252–253) — can be considered as trade embedded in, variously, kinship, tribal
or neighbourly structures. Polanyi’s idea of redistribution can be challenged in the
same manner; it may be more productive to regard transactions of this sort as trade
between individuals who are at different levels within a hierarchical social structure,
where the prices of goods normally are determined by tradition and relations of
power.
Polanyi’s notion of administered trade (1963, 30) is inadequate in that it directs
attention to the administrative framework and away from the economic aspects of
the trade in question. The economic aspects of trade in primitive societies are like-
wise defined out of significant existence by the term peripheral market exchange
(Bohannan and Dalton 1962).
If every transaction is regarded as embedded, having both an economic and a
social component, it is not necessary to make fundamental assumptions about
MAM 37: 12-skre – Press

post-substantivist production and trade 159

people’s innate tendencies in the way that the formalists do, or to postulate absolute
connections between types of transactions and social constellations in the manner of
the substantivists. Rather, it can be profitable to start from the basis that it is within
the compass of human possibility to produce and exchange goods as both an eco-
nomic and a social agent at the same time. The relative weight of these two com-
ponents is dependent on the interplay between social and cultural conditions. Such
conditions will give different groups and individuals opportunities to act within the
space marked out between economic motives and social obligations: as Richard Wilk
expresses it (1996, 146–153), between short-sighted, individual economic motives and
far-sighted, altruistic, social norms. Pierre Bourdieu (1990, 113–116) adopts the same
stance when he writes that economic agency is situated in the area between a ‘man of
good faith’ who bases all exchange on trust and generosity and ‘the shady dealer’,
who always allows ‘interested calculation’ to govern his business.
Bourdieu, however, goes one interesting step further when he points out that
‘interested calculation . . . is never absent from the generous exchange’ (1990,
114–115). The real contrast to the transactions of the shady dealer is that good-faith
economy is ‘based on a set of mechanisms tending to limit and disguise the play of
(narrowly) ‘‘economic’’ interests and calculation’. Bourdieu thus introduces a dis-
tinction between the individual’s own, private assessment and calculation of the
transaction and the positions acted out in the public domain vis-à-vis the parties to
the trade and others. Economic agency may thus clothe itself with hints of social
agency, and so become socially accepted, and play a different social role than it
otherwise would have.
Equally of interest is the importance Bourdieu attaches to whether the parties to
the transaction are connected in dense or loose social networks. Intimate and long-
term social relations give either party little scope for economic agency (Bourdieu
1990, 115). Trade with outsiders has an element of mistrust and conflict (Gustin 2004,
166–174; Skre 2007, 450–452) which may mean that honour does not lie in the
demonstration of a normative approach, restraint and generosity, but rather in
coming out of the transaction with an advantage. There may have been more honour
to be gained amongst one’s own in deceiving a stranger by selling him or her poor
goods at full price or by defrauding him or her over weight (Bourdieu 1990, 115).
A number of general observations concerning the analysis of economy and
society can be abstracted from the above. The connections between society, culture,
social norms and economy are so complex that absolute linkages between specific
combinations of the basic conditions cannot be postulated. Each individual economic
history has to be examined empirically. The market is not an ahistorical entity:
markets should be regarded as ‘historically variable social organisations constituted
by traders’ (Lie 1991, 227). Investigations of markets, on the one hand, have to map
out their social parameters, which the substantivists prioritize and, on the other, have
to explore the economic agency acted out there. In this way, ancient markets will
not appear as theoretical abstractions based upon assumptions about the profit-
maximizing tendencies of humankind, but rather as concrete, historical formations
with real actors who exploited economic opportunities with reference to cultural
norms, social relations and conflicts, structures of power, and laws. As Lie puts it
MAM 37: 12-skre – Press

160 maritime societies of the viking and medieval world

(1991, 230): ‘Rather than assuming the invisible hand, we should investigate the
concrete social relations of those who buy and sell; the visible hand of the market.’
Various terms have been used to label the direction within sociology, economics
and anthropology that has been outlined here. Swedberg and Granovetter’s term the
new economic sociology (1992) does not fit the archaeological context well. Lie’s
term the embeddedness approach (1997, 349–351) works best in sociology and eco-
nomics, as it emphasizes its contrast with the formalist approach that formerly
dominated analyses of the economies of modern societies within these disciplines.
Within archaeology and history, it is Polanyi’s substantivism that is the point of
departure, and the post-substantivist approach may therefore be an appropriate
label.

scandinavian specialized sites for trade and craft

Supplied with this perspective, I will turn to the Scandinavian sites from the
1st millennium ad which display substantial traces of trade and craft production.
Trade and craft production also took place on a smaller scale at manors and magnate
farms, but such sites are excluded from my analysis. Such production is aimed at the
residents of these farms. I will be looking at the sites which people travelled to for
trade purposes — I suggest calling them specialized sites for trade and craft produc-
tion. My ambition is to view production and exchange on these specialized sites for
trade and craft from both an economic and a social perspective. Social, cultural and
economic conditions on these sites seem to have presented new opportunities for
various social groups to display economic agency.
Criteria for sorting these sites into types can be drawn from the analysis above. I
have chosen the factors that seem decisive for the opportunities to display economic
agency. The criteria are thus:
1. Seasonal or permanent settlement. A permanent community of traders and
craftsmen will develop a different economic reality than occurs at seasonal
sites.
2. Inter-regional, intra-regional or long-distance trade. The opportunity to dis-
play economic agency is highly dependent on whether the two parties in the
exchange are connected in a dense or a loose social network.
3. The primary context of the site (see below). This criterion is also related to
the social connections between the two parties in the exchange.
Prior to ad c700 all specialized sites of trade and craft were located in what one
would call high-ranking central places. Early examples are Gudme/Lundeborg at
Fyn, Helgö at the mouth of Lake Mälaren, Uppåkra in Skåne and Sorte Muld at
Bornholm. In the 6th and 7th centuries a new generation of central places was estab-
lished, and some of these, like Tissø in Sjælland and Old Uppsala in Uppland, have
seasonal market sites. Judged from the artefactual evidence the trade there was
mainly inter-regional, to a small extent also intra-regional. I call these market sites
central-place markets (Figures 12.1 and 12.2).
MAM 37: 12-skre – Press

post-substantivist production and trade 161

Figure 12.1 Some specialized sites for trade and craft in Scandinavia ad c200–1000
(D Skre)
MAM 37: 12-skre – Press

162 maritime societies of the viking and medieval world

Figure 12.2 The chronological distribution of the four Scandinavian types of specialized sites
for trade and craft production (D Skre)

From ad c700 seasonal market sites were established in a different context. They
seem to have had no connections to central places. Johan Callmer (1995) comes to
the conclusion that goods produced at most of these sites had a primarily regional
distribution. Søren Sindbæk (eg 2005) has analysed the artefactual evidence from
such sites, and he concludes that there are few traces of long-distance trade. The craft
production conducted there, such as comb making and ironworking, was based upon
local raw materials. I call these sites local markets (Figures 12.1 and 12.2).
However, two of the seasonal sites that emerged in the early 8th century had a
different character. In Åhus, Skåne and Ribe in western Jutland there were a high
proportion of goods from long-distance trade, as well as copious traces from crafts
based on raw materials brought in through long-distance trade (eg large-scale glass-
bead production and metal-casting). Ribe was located in a context well suited for
long-distance trade, in the border-zone between Danes and Frisians. This was pos-
sibly also a political border-zone at the time. Åhus may have had a similar location,
in the border-zone between south-western Scandinavia, possibly an early Danish
kingdom, and the Baltic zone with Slavs, Swedes and other potential trading partners
(Figures 12.1 and 12.2).
Around ad 800, four specialized sites of trade and craft with permanent settle-
ment were established — Hedeby, Birka, Ribe and Kaupang — all in maritime
settings (Figures 12.1 and 12.2). They also have several other similarities, like plot
division, house types, layout, etc. I have chosen to call them ‘towns’. Sindbæk (2007)
has analysed the artefactual assemblages from a number of specialized sites of trade
and craft, and I agree with him that there are only these four sites in Scandinavia that
show a significant and lasting connection with long-distance trade systems before the
11th century. This connection is represented by their large quantities of goods
imported from outside Scandinavia. Similarly, these four sites have a broad range of
craftwork that made use of imported raw materials, like metal-casting and glass-bead
production. Furthermore, the quantity of balances, weights and coins, and evidence
MAM 37: 12-skre – Press

post-substantivist production and trade 163

of the use of silver as a currency, occurred earlier and is much more substantial at
these four sites than at the seasonal sites.
I have analysed the three towns in south-western Scandinavia and all of these
seem to have been founded within one decade on either side of the year ad 800, and in
the border-zone of the Danish kingdom. At this time the kingdom was being more
firmly established by Sigfred and his son Godfred. It appears that first, Ribe was
converted from a seasonal to a permanently settled site in the ad 790s. Thereafter,
Kaupang was established around ad 800, and then Hedeby was founded in ad 808.
As mentioned earlier, Ribe lies on the border with the Frisians. Kaupang lies on the
northern fringe of the kingdom, on the border to the Northmen. Hedeby lies on the
border with the Slavs (Figure 12.1). This border location of the three towns follows
the pattern of contemporary and earlier town foundations in English kingdoms and
in the Carolingian empire. It seems that Sigfred and Godfred had attempted to build a
kingdom modelled on those.

money
Before starting the discussion of the opportunities for economic agency at each
of these four types of sites, a brief discussion on the nature of money is needed.
Money has three functions. It can serve as currency, as a measure of value, and as a
means of saving. In modern societies these functions are commonly unified in just one
form, coinage, but in earlier times they were most commonly divided amongst vari-
ous media. In the Scandinavian Middle Ages commodities were commonly used as
currency (Skre 2011). As a measure of value, on the other hand, coined or weighed
silver was the norm. Relative values amongst a large number of different kinds of
goods were much more easily kept track of when they were defined according to a
common measure. In transactions, silver could serve as the common measure used to
value the goods involved, although no silver was exchanged.
As on the Continent, the common Scandinavian measure of value before the age
of silver was gold, which was weighed in units of the early øre: a term that etymo-
logically is derived from the Latin word for gold, aurum. The existence of weight
standards and gold objects adjusted by weight from the 3rd–6th centuries shows that
gold was then valued according to weight (Bakka 1978; Brøgger 1921; Herschend
1980; Munksgaard 1980; Skre 2007, 448–450; Steinnes 1936). Both the weight stan-
dards and the terminology of weight thus point towards gold having served a func-
tion as a measure of value in this period. There are some indications in the form of
cut pieces of gold found at places like Helgö and Lundeborg that gold was also used
as currency. However, the very limited quantities of cut-up gold indicate that com-
modities were the currency in the main bulk of transactions. Hoards demonstrate
that gold and other valuables were the main means of saving.
The earliest use of silver as currency in Scandinavia is represented by the first
striking of sceattas in Ribe a few years after the site was established around ad 710
(Feveile 2006, 31; Metcalf 1984). This introduction of a new form of currency was
part of a monetization of trade between the towns and market sites which had grown
up along the Carolingian and Anglo-Saxon North Sea coasts (Metcalf 2007, 1–2).
MAM 37: 12-skre – Press

164 maritime societies of the viking and medieval world

Beyond Ribe, silver took on the role of a measure of value in the rest of Scandinavia
in the 9th and 10th centuries. This development started in the towns in the second
quarter of the 9th century, when it was first used as a currency — as coins in Hedeby
and as hacksilver at Kaupang. In the mid-10th century numerous hoards show that
silver was then in rural areas a means of saving (Hårdh 2008). It must therefore also
have been the dominant measure of value. In the 11th century, coinage became widely
accepted as currency in most of Scandinavia. Still, commodities seem to have con-
tinued to be the most common form of currency, and weighed silver the most com-
mon measure of value until the end of the Middle Ages. After these excursions into
the types of sites and the functions of money, I return to the main question: How was
trade put into practice in 7th- to 10th-century Scandinavia, and what opportunities
were there for economic agency?

central-place markets
As Callmer interprets the finds from southern Sweden (1995, 65–66), craftsmen
in this area prior to ad c700 produced mainly one-off items, presumably to speci-
fications given directly by their customers. Craftsmen were primarily situated within
regions, within the parameters of dense social networks. Production largely took
place on manors and magnate farms where the customer lived, although there was
also large-scale production at central-place markets (Hjärthner-Holdar et al 2002,
164–169).
Since production was done to order, both the production and trading of the
products of craft would have taken place within a relationship between the producer
and the customer. This relationship must at least have lasted from the first contact
concerning what was to be produced, through the stages of specification and the
sequence of manufacture, on to delivery and payment. The duration of this relation-
ship indicates that trust was established which probably involved some element of
power. Close and durable social relations left the two parties little scope for eco-
nomic agency (Bourdieu 1990, 115).
As is well documented for later periods, prices during trade in dense networks
were strongly controlled by traditional valuations (Lunden 1972; Naumann 1987;
Sawyer 1990; Steinnes 1936). This stability of the values of commodities was not a
reflection of any external power sanctioning breaches of the norm, nor was it
dependent on whether the relationships between the trading partners were
hierarchical or not. The reason for the stability was that excessive pricing was
heavily sanctioned by the community — the man who showed himself to be greedy
was no longer considered to be a man of honour. Both parties were obliged to respect
a social norm. They had to place ‘honour before profit’, as Helgi Þorláksson has put
it (1992, 233, 242).
Probably all transactions in central-place markets respected prices fixed by tradi-
tion, and that gave little or no scope for economic agency. Beyond that, however,
markets coincided with sacrificial feasts and thing assemblies which met at the central-
place. Those meetings between the leading aristocrats of the region provided numer-
ous opportunities to exchange gifts. In gift exchange the precise calculation of the gift
MAM 37: 12-skre – Press

post-substantivist production and trade 165

received was essential. The return gift should relate to the value of the gift received.
In some relations giving a less valuable and a more valuable gift in return would be a
grave insult, while other relations demanded a deliberate difference between the values
of the gifts exchanged. Gift-giving was therefore an arena where the precise calcula-
tion of value was even more important than in trade. But the scope for economic
agency was non-existent.

local and nodal markets in the 8th century

In the nodal markets of the 8th century, the presence of long-distance tradesmen
changed this picture drastically. The scope for economic agency at nodal markets
must have been much greater. Two aspects of long-distance trade are the reason for
this. Firstly, foreign items whose value could not be defined by tradition were brought
in. Secondly, the trading partners lacked social ties beyond those fleeting connections
implicit in the trading itself. The absence of social ties stripped the transaction of
the social pretence which otherwise disguised its economic character, so that this
character was all the more bare to the eye than earlier. Display of economic agency in
trade with foreigners was not subject to social sanctions.
This opportunity for economic agency emerged primarily in the context of trans-
actions between Scandinavians and foreign traders, probably mostly Frisians and
Slavs. However, the public revelation of the economic character of these transactions
would have had consequences for the other forms of trade and production that took
place at nodal markets. This is most evident in the change from the production of
one-off items to the mass production of various kinds of goods. At the nodal markets
of the 8th century, Ribe and Åhus, the first evidence of the mass production of
identical items and the standardization of specific types is found. This was a style
of production in which the consumer, rather than specifying a personal order and
then awaiting manufacture before the final transaction could take place, was faced
with finished goods which he or she was invited to buy.
This change is evident earliest and most clearly in the production of glass beads
and copper-alloy jewellery (Callmer 1995, 53–57; Feveile and Jensen 2000, 17, 22).
Both are examples of production in which craftsmen had to deal with long-distance
traders in order to obtain their raw materials. This involvement with proficient
market traders may have imbued craftsmen with a market rationality that they
carried into their own production for, and trade with, Scandinavian customers. In
this way, one can suggest that these craftsmen literally embodied the change in eco-
nomic mentality that took place at the nodal markets. They must have been at the
very core of the transmission of greater economic agency into the wide variety of
transactions that took place in a nodal market.
The trade that was going on at the local and central-place markets in Scandi-
navia in the 8th century must also, essentially, have been conducted using commod-
ities as currency, which would presumably then have been valued according to the
traditionally accepted scales. On the other hand, there was a certain, wide-ranging
standardization of the products of craftwork, such as combs (Callmer 1995, 65–66),
indicating that craftsmen generally discovered opportunities for economic agency,
MAM 37: 12-skre – Press

166 maritime societies of the viking and medieval world

and that the production and marketing of their products grew less dependent upon
social relations even at the level of intra-regional production and trade. This change
probably occurred because the blatant exercise of economic agency amongst crafts-
men working with imported raw materials at nodal markets was adopted in other
craft-environments too. These two groups of craftsmen would have been in close
contact at the nodal markets. This behaviour thus became socially more acceptable
even within dense social networks; it was less stigmatized than previously.

towns in the 9th and 10th centuries

Mark Blackburn (2008, fig 3.28a–b) has suggested two models for coin-loss at
Kaupang, one of them reproduced here (Figure 12.3). The earliest finds of silver
currency from Kaupang are imported Western and Scandinavian coins which must
have arrived in the period ad 800–840. Only six of these have been found (Blackburn
2008, 56–57; Rispling et al 2008, nos 6–11). The other silver coins from Kaupang are
all Islamic dirhams which reached the site from just after the middle of the 9th century
until sometime before the early ad 930s (Blackburn 2008, 54; Kilger 2008, 245). A few
Islamic coins found their way to Kaupang just after ad 950.
In addition, there is hacksilver from objects and ingots, which first appeared in
the second quarter of the 9th century. The dirhams were also cut up, and the use

Figure 12.3 A model

for the estimated rate of
coin loss at Kaupang
(after Blackburn 2008,
fig 3.28b)
MAM 37: 12-skre – Press

post-substantivist production and trade 167

of coined and uncoined hacksilver as currency at Kaupang must have been rather
extensive in the second half of the 9th century, sharply falling towards the end of
permanent settlement there some time close to ad 930. The post-ad 950 activity
probably represents seasonal markets on the site.
The foundation of towns meant that craftsmen could now be permanently
settled elsewhere than at aristocratic residences (cf Hjärthner-Holdar et al 2002).
Their production thus became more independent of the elite, and the power-relation
that had probably existed between craftsmen and the lords who housed them disap-
peared or at least became weaker. This independence came with a price: craftsmen
became more dependent upon making their productivity and sales sufficient to sup-
port themselves. As the inhabitants of Kaupang did not produce their own food apart
from occasional fishing and perhaps some hunting, they depended on acquiring it
from farms in the agrarian hinterland.
These two factors — less aristocratic dominance and more dependence on
buying food — must be the two main reasons for the further surge in the stan-
dardized mass production that began at the nodal markets. Greater dependence of
the marketing of their products will have led to greater display of economic agency
amongst the permanently settled craftsmen. Owing to this dependence on selling and
buying — on kaup — Kaupang’s inhabitants performed a large number of trans-
actions. This alone will have increased awareness of the economic aspect of the deals.
Nevertheless, the situation could hardly have been one of purely economic enterprise.
Even within the town, the exercise of economic agency had to be balanced against
social norms for relationships with other townsfolk, the town authorities, the sup-
pliers of agricultural produce in the hinterland, and with customers and suppliers
from further afield. Still, the many transactions of town-life, and the activity of the
residents within loose networks, must have put those conventions under pressure,
and they must have changed, as a result, quite significantly during the lifetime of
Kaupang.
Because of the frequency of transactions in everyday life, the town’s population
also had a great need for currencies that all were willing to accept and for a common
measure of value that various forms of craftwork and other goods could be priced in.
Silver, in various forms, seems to have been an accepted currency throughout the
lifetime of Kaupang. The first generation of Kaupang’s population were familiar with
silver as currency only in the form of Western coins, the second generation adopted
the use of hacksilver (in the form of cut-up rings and ingots) and the third generation
used cut-up Islamic coins. It was with this third and subsequent generations that the
quantity of finds reaches a level which implies that silver was a widespread form of
currency within the town.

economic agency
It was not before the 10th century that greater quantities of hacksilver started to
appear in hoards in rural areas in the vicinity of towns. The amounts increased over
the course of the century, but in the parts of Scandinavia that lie far from towns, such
as the west and north of Norway, hacksilver seems still to have been little used
MAM 37: 12-skre – Press

168 maritime societies of the viking and medieval world

(Hårdh 2008). In those areas, whole silver artefacts are predominant in the hoards.
The appearance of weight-adjusted silver objects all over Scandinavia shows never-
theless that silver was established as the most common measure of value and means
of saving in Scandinavia by the end of the 10th century.
The social sanctions on economic agency were reduced already in the nodal
markets of the 8th century. However, the development of silver as currency and a
generally accepted measure of value and means of saving in the 9th and 10th centuries
demonstrated that the economic life of the towns was far more expansive and
dynamic. The transformative significance of the early towns regarding the economies
of the Scandinavian societies was not, however, primarily a matter of currencies and
measures of value. It lay rather in the opportunity that the loose social networks
and high frequency of transactions in the towns, created by long-distance trade and
the urban way of life, provided for the growth of economic agency. Karl Polanyi’s
theories on the economies of former times were a formidable achievement in that they
enabled us to see them as distinct in character and different from the economies of
our own time. As I see it, however, he made them too different, leaving virtually no
room for any dynamic within the economy itself. To catch that dynamic we need to
acknowledge that peoples of the past acted not only as social beings, but also as
economic agents.

references
Bakka, E, 1978 ‘Two aurar of gold: contributions to the weight history of the Migration Period’,
Antiquaries Journal 58, 279–298
Barrett, J, Beukens, R, Simpson, I, Ashmore, P, Poaps, S and Huntley, J, 2000 ‘What was the
Viking Age and when did it happen? A view from Orkney’, Norwegian Archaeological
Review 33, 1–39
Barrett, J H, Locker, A M and Roberts, C M, 2004 ‘‘‘Dark Age Economics’’ revisited: the English
fish bone evidence ad 600–1600’, Antiquity 78, 618–636
Blackburn, M, 2008 ‘The coin-finds’, in D Skre (ed), Means of Exchange: Dealing with Silver in the
Viking Period, Kaupang Excavation Project Publication Series vol 2, Aarhus University Press,
Aarhus, 29–74
Bohannan, P and Dalton, A (ed), 1962 Markets in Africa, Northwestern University African Studies
vol 9, Northwestern University Press, Evanston
Bourdieu, P, 1990 The Logic of Practice, Polity Press, Oxford
Brøgger, A W, 1921 Ertog og øre: den gamle norske vegt, Dybwad, Kristiania
Callmer, J, 1995 ‘Hantverksproduktion, samhällsförändringar och bebyggelse. Iakttagelser från
östra Sydskandinavien ca. 600–1100 e.Kr’, in H G Resi (ed), Produksjon og samfunn: om
erhverv, spesialisering og bosetning i Norden i 1. årtusen e.Kr., Universitetets Oldsaksamling,
Varia vol 30, Universitetets Oldsaksamling, Oslo, 39–72
Davis, J, 1992 Exchange, University of Minnesota Press, Minneapolis
Feveile, C (ed), 2006 Det ældste Ribe: Udgravninger på nordsiden af Ribe Å 1984–2000, Ribe
Studier vol 1.1, Jysk Arkæologisk Selskab, Højbjerg
Feveile, C and Jensen, S, 2000 ‘Ribe in the 8th and 9th century: a contribution to the archaeological
chronology of north western Europe’, Acta Archaeologica 71, 9–24
Granovetter, M, 1985 ‘Economic action and social structure: the problem of embeddedness’,
American Journal of Sociology 91, 481–510
MAM 37: 12-skre – Press

post-substantivist production and trade 169

Gudeman, S, 1986 Economics as Culture: Models and Metaphors of Livelihood, Routledge &
Kegan Paul, London
Gudeman, S, 2001 The Anthropology of Economy: Community, Market, and Culture, Blackwell,
Malden
Gustin, I, 2004 Mellan gåva och marknad. Handel, tillit och materiell kultur under vikingatid,
Almqvist & Wiksell, Lund
Hårdh, B, 2008 ‘Hacksilver and ingots’, in D Skre (ed), Means of Exchange: Dealing with Silver in
the Viking Age, Kaupang Excavation Project Publication Series vol 2, Aarhus University Press,
Aarhus, 95–118
Herschend, F, 1980 Myntat och omyntat guld. Två studier i öländska guldfynd, Gustavianum,
Uppsala
Hirsch, P, Michaels, S and Friedman, R 1987 ‘‘‘Dirty hands’’ versus ‘‘clean models’’: is sociology in
danger of being seduced by economics?’, Theory and Society 16, 317–336
Hjärthner-Holdar, E, Lamm, K and Magnus, B, 2002 ‘Metalworking and central places’, in
B Hårdh and L Larsson (ed), Central Places in the Migration and Merovingian Periods: Papers
from the 52nd Sachsensymposium Lund, August 2001, Uppåkrastudier vol 6, Almqvist &
Wiksell, Lund, 159–183
Kilger, C, 2008 ‘Kaupang from afar: aspects of the interpretation of Dirham finds in northern and
eastern Europe between the late 8th and early 10th centuries’, in D Skre (ed), Means of
Exchange: Dealing with Silver in the Viking Age, Kaupang Excavation Project Publication
Series vol 2, Aarhus University Press, Aarhus, 199–252
Lie, J, 1991 ‘Embedding Polanyi’s market society’, Sociological Perspectives 34, 219–235
Lie, J, 1992 ‘The concept of mode of exchange’, American Sociological Review 57, 508–523
Lie, J, 1997 ‘Sociology of markets’, Annual Review of Sociology 23, 341–360
Lunden, K, 1972 Økonomi og samfunn. Synspunkt på økonomisk historie, Universitetsforlaget,
Oslo
Metcalf, D M, 1984 ‘A note on sceattas as a measure of international trade, and on the earliest
Danish coinage’, in D Hill and D M Metcalf (ed), Sceattas in England and on the Continent,
British Archaeological Reports, British Series 128, Archaeopress, Oxford, 159–164
Metcalf, D M, 2007 ‘Regions around the North Sea with a monetised economy in the pre-Viking
Age’, in J Graham-Campbell and G Williams (ed), Silver Economy in the Viking Age, Left
Coast Press, Walnut Creek, 1–11
Munksgaard, E, 1980 ‘Justerede ringe af ædelmetal fra germansk jernalder og vikingetid’,
Aarbøger for Nordisk Oldkyndighed og Historie 1978, 150–154
Naumann, H-P, 1987 ‘Warenpreise und Wertverhältnisse im alten Norden’, in K Düwel,
H Jahnkuhn, H Siems and D Timpe (ed), Untersuchungen zu Handel und Verkehr der vor-
und frühgeschichtlichen Zeit in Mittel- und Nordeuropa. Teil IV: Der Handel der Karolinger-
und Wikingerzeit, Vandenhoeck & Ruprecht, Göttingen, 374–389
Polanyi, K, 1944 The Great Transformation, Farrar & Rinehart, New York
Polanyi, K, 1957 ‘The economy as instituted process’, in K Polanyi, C M Arensberg and H W
Pearson (ed), Trade and Markets in the Early Empires: Economies in History and Theory,
Free Press, New York, 243–270
Polanyi, K, 1963 ‘Ports of trade in early societies’, The Journal of Economic History 23, 30–45
Rispling, G, Blackburn M and Jonsson K, 2008 ‘Catalogue of the coins’, in D Skre (ed), Means of
Exchange: Dealing with Silver in the Viking Period, Kaupang Excavation Project Publication
Series vol 2, Aarhus University Press, Aarhus, 75–93
Sawyer, P, 1990 ‘Coins and commerce’, in K Jonsson and B Malmer (ed), Sigtuna Papers: Proceed-
ings of the Sigtuna Symposium on Viking-Age Coinage 1–4 June 1989, Kungl. vitterhets
historie och antikvitets akademien, Stockholm, 283–288
MAM 37: 12-skre – Press

170 maritime societies of the viking and medieval world

Sindbæk, S M, 2005 Ruter og rutinisering: Vikingetidens fjernhandel i Nordeuropa, Multivers

Academic, Copenhagen
Sindbæk, S M, 2007 ‘Networks and nodal points: the emergence of towns in early Viking Age
Scandinavia’, Antiquity 81, 119–132
Skre, D, 2007 ‘Towns and markets, kings and central places in south-western Scandinavia
c. ad 800–950’, in D Skre (ed), Kaupang in Skiringssal, Kaupang Excavation Project Publica-
tion Series, vol 1, Aarhus University Press, Aarhus, 445–469
Skre, D (ed), 2008 Means of Exchange: Dealing with Silver in the Viking Period, Kaupang Exca-
vation Project Publication Series vol 2, Aarhus University Press, Aarhus
Skre, D, 2011 ‘Commodity money, silver and coinage in Viking-age Scandinavia’, in J Graham-
Campbell, S M Sindbæk and G Williams (ed), Silver Economies, Monetisation and Society in
Scandinavia, ad 800–1100, Aarhus University Press, Aarhus, 67–91
Steinnes, A, 1936 ‘Mål, vekt og verderekning i Noreg i millomalderen og ei tid etter’, in S Aakjær
(ed) Maal og Vægt, Nordisk kultur vol 30, Bonnier, Stockholm, 16–86
Swedberg, R and Granovetter, M, 1992 ‘Introduction’, in M Granovetter and R Swedberg (ed),
The Sociology of Economic Life, Westview Press, Boulder, 1–26
Þorláksson, H, 1992 ‘Social ideals and the concept of profit in thirteenth century Iceland’, in
G Pálsson (ed), From Sagas to Society: Comparative Approaches to Early Iceland, Hisarlik
Press, Enfield Lock, 231–245
Wilk, R R, 1996 Economies and Cultures: Foundations of Economic Anthropology, Westview
Press, Boulder
Zafirovski, M, 2001 Exchange, Action and Social Structure, Greenwood Press, London
MAM 37: 13-nordeide – Press

chapter 13
LATE IRON AGE BOAT RITUALS AND
RITUAL BOATS IN NORWAY

By SÆbj ø rg Walaker Nordeide

introduction
In recent studies of Late Iron Age graves in Norway I surveyed religious cult in
selected municipalities, aiming at providing a representative view of various parts of
southern and central Norway (eg Nordeide 2011b). Through this analysis I observed
considerable regional differences — one being the occurrence of a boat or boat parts
in various contexts. In some coastal areas boats occurred in graves, whereas in others
they did not. In yet other districts boats were found in bogs instead. This article will
discuss various aspects of these rituals.
Iron Age burials with boats is a Nordic tradition — found in Norway, Sweden,
Finland, Iceland and the Scandinavian settlements of the British Isles. Only a few Iron
Age boat graves have been found in Denmark (Børsheim 1999). The role of a boat1
in a grave has been discussed, and various explanations have been presented based on
the status of the deceased in this world or in the world of the dead. The boat is
sometimes seen as a kind of coffin, a protected room in which the deceased could be
placed, or as a vehicle to transport the deceased to the world of the dead, which
would have required a voyage (Curle et al 1954). It is also argued that the boat was,
just like other types of grave goods, a practical item which the deceased needed in his/
her new existence, or that it was part of the fuel for the cremation (Kivikoski 1964,
172; Nilsen 1997, 103). However, sometimes only parts of boats were placed in the
graves, or else the boat was broken at the time of the burial. Taking these circum-
stances into consideration, boats in graves should probably sometimes be interpreted
as a symbol for something the dead required (Næss 1969). There may, however, have
been various reasons for putting a boat in a grave that differed from place to place.
We should keep in mind the likelihood of diverse traditions.

iron age boats in norway

The total number of recorded graves and excavated graves in the selected
municipalities varies from place to place (Figure 13.1), but this observation is not so
important for the present discussion. The general picture from my results is similar to
MAM 37: 13-nordeide – Press

172 maritime societies of the viking and medieval world

Figure 13.1 The number of occurrences of boats in Late Iron Age graves found in selected
municipalities (S W Nordeide)

previous analyses. Boat graves have a relatively wide distribution, though this type
of burial was practised by people in coastal rather than inland districts (see, for
instance, Næss 1969). In some areas the frequency of boats is higher than in others.
For instance, c50% of the graves in Tingvoll included a boat, which is an unusually
high representation, while in other regions there are hardly any boats found in graves.
Such differences can be hard to explain. It is even more difficult to understand why
people in some coastal areas never seem to have used boats for burials. This is the
case, for instance, at Selje and Herøy on the extreme west coast where no boats were
found in the graves. If there was a connection between items put in the graves and
items used in daily life, it is impossible to explain the absence of burials with boats in
these municipalities. The only possible way to reach Herøy was by boat, and boats
must have been of general importance among people along the west coast.
It is a fact, however, that boats also had a meaningful religious position in Selje
and Herøy. They were used in rituals other than burial. Bog finds of boats, pieces
of boats and boat equipment (like oars) — finished or unfinished, fully sized or in
miniature — were found here and elsewhere along the entire western coast of
Norway (Figure 13.2). Conversely, only in one instance is this kind of bog find known
from eastern Norway.
Boat or boat parts have appeared occasionally in bogs and graves in Norway
from ad c200,2 but they became more frequent and typical during the Late Iron Age,
from the 7th century. There is no complete map of all the boat graves in Norway, but
based on a comparison of Figures 13.1 and 13.2, and a general knowledge of boat
graves in Norway, it is probable that boat burials have a much wider geographic
MAM 37: 13-nordeide – Press

late iron age boat rituals 173

Figure 13.2 Bog finds of boats/boat parts from the Iron Age in Norway
(after Nordeide 2011b, 250)

distribution than the bog finds, even if both categories represent a mainly coastal
phenomenon.
In this article I will compare the two ritual find categories: boats found in bogs
and boats used in burials. I will concentrate on probably the most famous example
from each category: Kvalsund from among the bog finds, and the Oseberg ship
barrow from among the burials (Figure 13.3). Each shares some typical character-
istics with other finds in their respective category. There are similarities between the
two rituals, but also some significant differences. One must ask whether or not we
should compare these finds at all?
Excavations took place in 1904 at Oseberg, and in 1920–23 at Kvalsund. The
archaeologist Håkon Shetelig, the botanist Jens Holmboe and ship engineer
MAM 37: 13-nordeide – Press

174 maritime societies of the viking and medieval world

Figure 13.3 The locations of

Kvalsund and Oseberg
(S W Nordeide and Vicki Herring)

Fr Johannessen took part in the site and/or post-excavation work in both excava-
tions. This eases comparison of the two sites, as the specialists used their experience
at Oseberg as a foundation for their work at Kvalsund, and made many of the same
kinds of observations in both places.
Both finds could be dated to early in the Norwegian Late Iron Age. The
Kvalsund ship was built in the period ad 651–869, but probably before ad 800.3 The
Oseberg burial occurred slightly later, in ad 834, although the boat had been con-
structed 10 to 15 years earlier based on dendrochronology (Bonde and Christensen
1993; Bonde 1994).

description of the selected material

Kvalsund
Kvalsund is located by a narrow and shallow inlet on the island Nerlandsøya in
Herøy municipality (Figure 13.4). The site is close to one of the hardest parts to cross
by boat on the north–south route along the west coast of Norway. The sound at
Kvalsund has been important as a sheltered part of the route, where people may have
had to wait for the wind to drop before continuing their voyage. Several burial monu-
ments and a circular enclosure were found on the same headland, dated from the
Bronze Age to the Iron Age.
MAM 37: 13-nordeide – Press

late iron age boat rituals 175

Figure 13.4 Kvalsund’s local position (S W Nordeide and Vicki Herring)

At least four boats were buried in the bog at Kvalsund on at least three different
occasions (Shetelig and Johannessen 1929). The site was originally carefully pre-
pared. The boats were deposited in an artificial pit dug almost to the bottom of the
bog. Half of the depth was filled in with a 0.5m thick and 15m long layer of turf,
moss, heather, wood chippings, etc. The plant remains used to fill the hollow showed
that this work was done during the spring or early summer, before midsummer.
At this level the parts of one big and one smaller boat were laid down, all
elements oriented c north–south in a layer of turf with white moss (Figure 13.5). At
the south end a square metre of turf was covered with nettles. On-site observations
indicate that the pit may have been filled with water whilst open during the summer,
thus creating a little pond. Some stones were put in, but not in an obvious pattern.
They have been interpreted as stepping stones for ease of access to the pit (Shetelig
and Johannessen 1929, 27). Many narrow wooden pieces like pointed branches, twigs
and pieces of the boat (such as oars) were inserted. They were forced with great
power more or less vertically through the white moss and turf until their lower ends
reached the bottom of the hollow. The sticks were not put down in any obvious
pattern. No body or classic grave goods were found, and there were definitely no
graves associated with the finds. The only additional object was a hollow piece of
wood, interpreted as a possible loud hailer.
MAM 37: 13-nordeide – Press

176 maritime societies of the viking and medieval world

Figure 13.5 The bog find at Kvalsund, plan (top) and section (bottom) (after Shetelig and
Johannessen 1929, 28)

At least three of the four boats (the fourth has not been preserved) were deliber-
ately and systematically destroyed, piece by piece, with pure manpower and no tools.
Some parts were partially burnt. Ultimately, the boat parts were covered up by moss,
turf, etc until the pit was filled to the top of the bog. Some of the white moss did not
grow at the site, but was collected from elsewhere. The pit was closed in the late
summer or autumn, so the activity (be it continuous or intermittent) continued for the
whole summer season.
Oseberg
Oseberg was a large burial mound, 40.5m in diameter and 6.4m high, situated by
a sheltered harbour in a fully cultivated region providing among the best conditions
for agriculture in Norway (Brøgger et al 1917, 134–135) (Figure 13.6). Compared to
the coastal climate at Herøy, this part of the country has colder winters and warm,
relatively dry summers. The only contemporary town in the country, Kaupang, was
located c45km south of Oseberg, along the coast.
As at Kvalsund, the burial site at Oseberg was carefully prepared (Brøgger et al
1917, 123–164). A ditch, 23m long, was constructed and roughly 100m3 of clay
removed to accommodate the keel of the ship. The ship was then placed in the ditch
and the mound partly constructed (W C Brøgger 1917; Gansum 1995, 171–172). The
vessel was tied with a rope to a large stone. Clearly, it was not supposed to sail. The
grave goods were placed carefully inside, and nothing was deliberately destroyed,
except for the animals that were killed (Brøgger et al 1917, 140–144). Among other
grave goods there was a carriage, four sledges, five decorated posts, pillows, textile
tapestries, fifteen horses, four dogs, two oxen and a wide spectrum of kitchen utensils,
agricultural tools and tools for textile production. Everything was exceptionally well
MAM 37: 13-nordeide – Press

late iron age boat rituals 177

Figure 13.6 The locations of the

Oseberg barrow and other well-
known burials at Gokstad, Borre
and Tune (after Christensen et al
1992, 11)

preserved, and many items were beautifully decorated (for example, the rich animal
ornament on wooden objects, including the ship). Most artefacts demonstrated an
overwhelming quality in all ways, but in addition to this, more ordinary, everyday
utensils were also present. Heaps of stones were also placed in the boat, before every-
thing was covered up by the mound.
Two women were buried in the ship, inside a wooden grave chamber (A W
Brøgger 1917, 146). Buckets of apples and blueberries indicate that the mound was
closed during the late summer or autumn (Holmboe 1917). The timber for the grave
chamber was felled during the summer of ad 834, which would also indicate this
burial season (Bonde 1994). The abaft part of the ship was filled in and sealed by the
mound first. Then the part from the front of the ship to the middle of the grave
chamber was closed. This means that the burial preparations and rituals occurred
during the late summer/autumn season, similar to Kvalsund, but the grave monument
at Oseberg was probably more quickly completed (Nordeide 2011a).

discussion
There are some clear similarities between the two sites. Both had a ship as a
central object and both were arenas for religious rituals in the summer season. This
justifies comparison, but there is also an obvious difference between them. Oseberg
MAM 37: 13-nordeide – Press

178 maritime societies of the viking and medieval world

was a grave for two women, including many grave goods, while there was no grave,
nor any grave goods, at Kvalsund. The question, then, is whether burial marks the
only real difference between these two places? Was it a coincidence that the two
women died at Oseberg, and could we have expected a grave at Kvalsund as well? To
the latter question, at least, I believe the answer is no. The Kvalsund find is typical of
many along the coast, the distribution of which is shown in Figure 13.2. In addition
there are other bog finds where, although no boat parts were present, there were
items like those found at Kvalsund. Examples include pointed wooden sticks, some-
times laid down in circular formations. Table 13.1 shows a selection of these bog
finds from the municipalities of Herøy, Selje and Gulen in western Norway. Not all
of the pointed-stick finds are dated, but the example of Kvalsund, and of Storhaug
(see below), suggest that they may belong to the Late Iron Age. Moreover, boats were
included among the grave goods of both cremations and inhumations, whereas the
bog finds are never burned (even though some charcoal is observed in association
with some of the bog finds, and some of the sticks at Kvalsund were charred). The

Table 13.1 Artefacts found in bogs in Herøy, Selje and Gulen

Museum
Municipality no. Location Period Find

Herøy B7600 Kvalsund ad 651–869 Four boats offered and destroyed,

pointed sticks. Found in bottom of
bog
Herøy B6979 Kopperstad ad 560–1050? Front part of boat, decorated, pine
Herøy B9384 Leikanger ad 560–1050? Two concentric circles of pointed
pine sticks, diameter 40cm and
175cm
Herøy B9399 ad 560–1050? A pine celt, 50cm deep in a bog
Herøy B9400 Leikanger ad 560–1050? A pine tool similar to a knife,
found 1m deep in bog (same bog as
above), in charcoal
Herøy B9402 Ytre Bø Iron Age Pine sword (peat knife?)
Herøy B10279 Kvalsund ad 560–1050? Pine clamp, found 50m from the
boats (B7600)
Herøy B12589 Remøy ad 560–1050? Two ‘bog sticks’
Selje B6613 Aarsheim Hove ad 800–1050 12 glass beads and two gilded
bronze fittings of Irish origin,
found 1m deep in bog
Gulen B7234 Byrknes, ad 800–1050 13 glass beads, found 2m deep in
Byrknesøy bog
Gulen B10221 Revurholmen, ad 560–1050? 20 pointed sticks, found 1.5m deep
Byrknesøy in bog, all pointed ends stuck
together, like a fan
MAM 37: 13-nordeide – Press

late iron age boat rituals 179

geographic distribution and the physical characteristics suggest a different interpreta-

tion for the bog-found boats and the boat graves.
Nevertheless, an assemblage of artefacts similar to Kvalsund is recorded, along
with rich grave goods, in the Storhaug ship barrow at Avaldsnes in south-west
Norway. There the oars were deliberately destroyed and one of them was placed
vertically in the mound. Pointed sticks or poles inserted into the ground were also
observed in the mound (Opedal 2005, 42–44). This barrow is dated to ad c779 (Bonde
and Stylegar 2009). The date of Storhaug is thus roughly the same as the bog find at
Kvalsund. Just as the Oseberg ship was moored and probably not meant to sail, some
of the other items among the bog and grave finds must have had a symbolic meaning
in the ritual.
The time of year was crucial for annual rites, but a burial had to correspond to
the death of one or more persons, whenever they died or were sacrificed during the
year. Though the season of other ritual sites is rarely known, it seems to have been an
important part of the activity at Kvalsund. The long duration of summer rituals there
could be interpreted as an indication of how important this season was, and still is, to
Scandinavians. The long, cold and dark winter makes the warm summer particularly
important, and it would not be unexpected to find arenas where religious rituals took
place to ensure favourable sowing, growth and harvesting. As such, these rituals
would provide an assurance of life, and the boat was a convenient symbol for this
(Eliade 1974 [1958], 188–212). The boat was used as a symbol as early as the Bronze
Age, as demonstrated by the agricultural rock carvings in Scandinavia.
While all the grave goods were laid down carefully at Oseberg, the Kvalsund
boat was deliberately destroyed and only included symbolic pieces of wood, some of
which were rammed violently into the ground. The fierce use of power and the forc-
ing of vertical elements into the ground from which life grows could also be under-
stood as a symbol of fertility. Thus, the bog finds at Kvalsund combine many vital
elements that could be understood in the context of annual rites: water, soil, fire, boat
and human vitality.
At Oseberg it appears that the burial took place in late summer/autumn, most
likely in September. As a burial, the ritual naturally concerned aspects of life and
death. The fruits and berries may thus have been symbols of the ongoing harvest, and
the ship and other objects may have served as fertility symbols, making the burial
similar to annual fertility rites in many ways (Ingstad 1995). The foodstuffs may
alternatively be understood as only one part of all the necessary supplies with which
the deceased were provided.
Each turf was carefully cut and placed in the Oseberg mound and much effort
was taken to organize the grave goods. Nevertheless, it is likely that the burial was
performed relatively rapidly. It may have been a coincidence that both women died at
the same time of the year, but it is relevant to ask whether both were dead before the
preparations started. Due to the wealth and care with which the burial was per-
formed, one could suspect that it had been planned for some time. The woman (or
women) may have been sacrificed, but this need not have been the case. Another
alternative is that they were ill and expected to die. The deceased or their kin may
MAM 37: 13-nordeide – Press

180 maritime societies of the viking and medieval world

have prepared their funeral in advance, just as the pharaohs of ancient Egypt
prepared their pyramids as many as 20 years before their deaths (Cook et al 1977, 8).
It is thus also relevant to consider what is known about the two women’s status and
destiny when the preparations began.
It is suggested that at least one of the women was a person of high rank: a queen,
a cult leader — a hofgyðja — and maybe a priestess for the cult of the god Frey
(Ingstad 1982). It appears likely that the death of at least one of the women was
expected. According to Per Holck’s recent results, one of the women was very elderly,
perhaps more than 80 years old, whereas the other was in her 50s (Holck 2008). The
older woman suffered from cancer, and the seeds of cannabis in a little bag were
possibly meant for use as painkillers. She may have had a hormonal syndrome that
made her look and sound like a man, and she would probably have had problems
getting pregnant. The cancer probably killed her. The cause of death for the other
woman is not known. She may have been sacrificed, as previously suggested.
To conclude, Kvalsund and Oseberg show some striking similarities while
nevertheless remaining very different — an observation also applicable to bog and
grave finds of boats more generally. The graves had to make burial the crucial event,
securing future social, religious and economic aspects both for the deceased and for
the living community. The funerary rite was a way to deal with the past, the present
and the future. Conversely, I would like to suggest that the bog finds represent more
specialized annual rites that had more to do with the present and the future than the
past, even if the rituals may have offered opportunities to re-experience myths about
the past as well. The ritual boats in both burials and bogs were thus associated with
aspects of life and death — with fertility in its wider sense — among the maritime
societies of various differing parts of Late Iron Age Norway.4

notes
1 As a general term, the word ‘boat’ will be applied a boat from Børøya, Nordland, is radiocarbon
rather than ‘ship’, regardless of the size of the dated (at 95.4% probability) to ad 333–596
vessel. (1600t60 BP).
2 For instance, a bog find from Mangersnes, 3 Nettle from the site was radiocarbon dated (at
Radøy produced the following radiocarbon 95.4% probability) to ad 651–799 (83.1%) or
dates at 95.4% probability: (ad 9–382 (T-7298, 790–869 (12.3%), (T-3755, 1290t50 BP).
1840t80 BP), ad 560–688 (Ua-703, 1415t90 BP) 4 I would like to thank Kari Hjelle, Niels Bonde,
and 190 bc–ad 03 (Ua-704, 1890t125). A crema- Arne Emil Christensen and Terje Gansum for
tion grave with boat at Valderøy, Giske, discussions regarding the history of the Oseberg
Sunnmøre is radiocarbon dated (at 95.4% prob- mound, and Jena Habegger-Conti for improving
ability) to ad 87–534 (173t80 BP). A bog find of my English.

references
Bonde, N, 1994 ‘De norske vikingeskipbsgraves alder: et vellykket norsk-dansk forsknings-
prosjekt’, Nationalmuseets Arbejdsmark 1994, 128–147
Bonde, N and Christensen, A E, 1993 ‘Dendrochronological dating of the Viking Age ship burials
at Oseberg, Gokstad and Tune, Norway’, Antiquity 67, 575–583
Bonde, N and Stylegar, F-A, 2009 ‘Fra Avaldsnes til Oseberg: dendrokronologiske undersøkelser
av skipsgravene fra Storhaug og Grønhaug på Karmøy’, Viking 72, 149–168
Børsheim, R L, 1999 ‘Gravene på Gausel’, Fra haug og heidni 4, 6–12
MAM 37: 13-nordeide – Press

late iron age boat rituals 181

Brøgger, A W, 1917 ‘Haugen’, in A W Brøgger, H Falk and H Shetelig (ed), Osebergfundet,

Volume 1, Den Norske Stat/Universitetets Oldsaksamling, Kristiania, 123–164
Brøgger, A W, Falk, H and Shetelig, H (ed), 1917 Osebergfundet, Volume 1, Den Norske Stat/
Universitetets Oldsaksamling, Kristiania
Brøgger, W C, 1917 ‘Oseberghaugens historie’, in A W Brøgger, H Falk and H Shetelig (ed),
Osebergfundet, Volume 1, Den Norske Stat/Universitetets Oldsaksamling, Kristiania,
165–197
Christensen, A E, Ingstad, A S and Myhre, B, 1992 Osebergdronningens grav: vår arkeologiske
nasjonalskatt i nytt lys, Schibsted, Oslo
Cook, J, Heidenstam, D, Maynard, C and Rowland-Entwistle, T, 1977 Arkeologi, Fakta i farger,
Chr. Schibsteds Forlag, Oslo
Curle, A O, M Olsen and H Shetelig, 1954, Civilisation of the Viking Settlers in Relation to their
Old and New Countries, Viking Antiquities in Great Britain and Ireland Volume 6,
Aschehoug, Oslo
Eliade, M, 1974 [1958] Patterns in Comparative Religion, New American Library, New York
Gansum, T, 1995 Jernaldergravskikk i Slagendalen: Oseberghaugen og storhaugene i Vestfold —
lokale eller regionale symboler? En landskapsarkeologisk undersøkelse, Unpublished Mag.Art
thesis, University of Oslo, Oslo
Holck, P, 2008 ‘The Oseberg ship burial, Norway: new thoughts on the skeletons from the grave
mound’, European Journal of Archaeology 9, 185–210
Holmboe, J, 1917 ‘Oseberghaugens torv’, in A W Brøgger, H Falk and H Shetelig (ed), Oseberg-
fundet, Volume 1, Den Norske Stat/Universitetets Oldsaksamling, Kristiania 201–205
Ingstad, A S, 1982 ‘Osebergdronningen — hvem var hun?’ Viking 1981, 49–65
Ingstad, A S, 1995 ‘The interpretation of the Oseberg-find’, in O Crumlin-Pedersen and B M Thye
(ed), The Ship as Symbol in Prehistoric and Medieval Scandinavia, National Museum of
Denmark, Copenhagen, 139–148
Kivikoski, E, 1964 Finlands förhistoria, Holger Schildts Förlag, Helsingfors
Næss, J-R, 1969 ‘Grav i båt eller båt i grav’, Stavanger Museums Årbok 1969, 57–76
Nilsen, R H L, 1997 Båtgravskikk? Om berettigelsen av å opprettholde begrepet ‘Båtgravskikk’
som analytisk kildekategori i arkeologien, MA thesis, NTNU, Trondheim
Nordeide, S W, 2011a ‘Death in abundance — quickly! The Oseberg ship burial in Norway’, Acta
Archaeologica 82, 7–15
Nordeide, S W, 2011b The Viking Age as a Period of Religious Transformation: The Christian-
ization of Norway from ad 560–1150/1200, Brepols, Turnhout
Opedal, A, 2005 Kongens død i et førstatlig rike. Skipsritualer i Avaldsnes-området og aspekter ved
konstituering av kongemakt og kongerike 700–950 e.Kr, Volume 228, Acta Humaniora, Det
humanistiske fakultet, Universitetet i Oslo, Oslo
Shetelig, H and Johannessen, F, 1929 Kvalsundfundet og andre norske myrfund av fartøier,
Volume 2, Bergen Museum, A/S John Griegs boktrykkeri, Bergen
MAM 37: 14-hansen – Press

chapter 14
BERGEN ad 1020/30 – 1170:
BETWEEN PLANS AND REALITY

By Gitte Hansen

In the Middle Ages, Bergen emerged as the most important town in Norway, serving
as a maritime gateway to the North Atlantic, Britain, Ireland and Continental
Europe. According to saga traditions, King Olav Kyrre (the Gentle, ‘the Peaceful’)
(ad 1066–93) founded the town, and for centuries researchers have studied early
Bergen and the king’s role. Until the middle of the 1990s, research was mainly based
on sparse written records, and the story of Bergen’s genesis was accordingly drawn
up in rather general terms. Only in recent years has the large body of archaeological
material been drawn into the discussion, forming the basis for new approaches. This
paper presents an attempt to populate the story about the rise of Bergen with actors/
people from different levels of the social hierarchy. It focuses in particular on the first
c150 years of Bergen’s history, between ad 1020/30 and ad c1170. Activities dating to
between ad c800 and ad c1020/30 serve as a foundation.

sources, theoretical and methodological approaches

I understand social change as a product of the interplay between people from
different levels of the social hierarchy and their wider historical context (cf Giddens
1984; Hansen 2000). Archaeological sources are tangible. Unlike ideas that only ever
existed within the mind of the planner (or the medieval historian), archaeological
sources have the advantage that they reflect the actual initiatives and activities carried
out by actors in the past — whether the initiatives were successful or not is another
story. I will pin actors, or groups of actors, to the concrete physical manifestations of
initiatives and activities reflected in the archaeological sources from early Bergen. In
this way a varied and dynamic picture of the rise of Bergen will be drawn up.
Somewhat simplified and schematically one may say that, in a stratified society,
like that of the early Norwegian central kingdom, initiatives were taken either from
the ‘top down’ or from the ‘bottom up’. In this model, top-down initiatives were
taken by resourceful actors with a central position in society. These individuals had
an opinion about how society was to be formed and perhaps also had the resources to
realize their ideas. The king or his representatives belonged to this group of actors,
and in Norway the church was closely tied to the king prior to the middle of the
MAM 37: 14-hansen – Press

between plans and reality 183

12th century (cf Andersen 1977, 311). Bottom-up initiatives were taken by actors with
fewer resources positioned at lower levels of the social hierarchy. I have chosen to call
this group of actors the townspeople and visitors in Bergen.
I will discuss the intended and actual functions of early Bergen. The functions
may have been complex and they may have changed as the historical context changed.
The intended functions are those that the founder(s) of Bergen had planned for the
place, whereas the actual functions are those that were carried into life by the users of
Bergen. I will mainly address the functional aspects involved and I hold as a premise
that activity in the very beginning reflected the functions intended for Bergen. The
physical layout as well as ‘rules’ for the use of the town probably constrained and
limited the users, but also presented assets and possibilities. In time, functions that
were not part of the original top-down plans may have been introduced and Bergen
may have evolved a life of its own. The contemporary sources about early Bergen
comprise a few written texts in addition to archaeological and botanical data from
excavations and masonry studies. Altogether c15,000m2 and 149 trench profiles have
revealed structures, layers and around 10,000 artefacts that can be dated to between
ad c800 and ad c1170. I have divided the sources into basic, supplementary and
general background sources according to their reliability regarding dates and location.
Basic sources are well dated and ideally should carry the weight in discussions; how-
ever, they are not always available. When only supplementary sources are available,
more than two independent supplementary sources must point in the same direction
when interpreting patterns in the material. General background sources cannot stand
alone when interpreting the material. Data has been studied spatially using the
production of maps and a qualitative and contextual approach. Plots and ‘pockets’ of
culture layers serve as analytic units for artefacts and other archaeological material.
Furthermore, early Bergen is studied through the narrow time scales of five archaeo-
logically defined horizons. This fine chronological resolution in some cases makes it
possible to link major initiatives with historically known actors. Horizon 1 covers the
period from ad c800–1020/30 and serves as a backdrop. Horizon 2 covers the period
between ad c1020/30 and ad c1070, horizon 3 the period between ad c1070 and
ad c1100, horizon 4 the period between ad c1100 and the late 1120s and horizon 5
covers the period between the late 1120s and ad c1170 (Hansen 2005, 35–53).
Medieval Bergen was located by Vågen, a bay that stretches north-west to south-
east into the land from the inner coast of western Norway. Around ad 1000 Vågen
was deeper and wider than it is today. The original northern shoreline ran as far as
130m north of the modern quay front and the bay extended some 300m further east-
wards than today (Hansen 2005, 53–55). Here Bergen is divided into: the Holmen
area, the town area (subdivided into northern, middle and southern town areas), the
Nonneseter area, and the Nordnes peninsula. An inlet — known in later sources as a
swampy area called Veisan — separated Holmen from the northern town area.

horizon 1 (ad c800 – 1020/30)

The sources from horizon 1 have been studied in order to establish whether or
not there was a non-agrarian settlement in the Bergen area prior to horizon 2. Only a
MAM 37: 14-hansen – Press

184 maritime societies of the viking and medieval world

few sources could be associated with horizon 1, they were located in two places: in
the northern part of Bergen in the Veisan inlet, and further south, between the middle
and the southern town areas.
Botanical samples from sediments in Veisan indicate that there was settlement in
the vicinity, perhaps as early as in the 9th or 10th century (Hjelle 1986; 2000). Based
on my evaluation of the available sources, I find that these traces are best explained as
representing a settlement where agricultural activities were carried out (for a full dis-
cussion, see Hansen 2005, 127–130; 2008, 22–23 and references therein; Hjelle 2000).
This settlement may have been located at Holmen and probably had fields in the
Bergen area. Between the middle and the southern town areas there was a pier dated
to c900. It was, however, not part of a wider built-up area. As a hypothesis it is sug-
gested that the pier served as a landing place for the royal estate at Alrekstad c2km
south of the mouth of Vågen. Thus, with the sources available today, I do not think
one can argue for a non-agrarian settlement in the Bergen area during horizon 1
(Hansen 2005, 127–131).

horizon 2 (ad 1020/30 – c1070)

Near the beginning of the 11th century, the northern town area was regulated
into a system of c11.5m wide plots demarcated by palisade fences (Figure 14.1). The
regulation can be tentatively dated to ad c1020/30 through activity that in all likeli-
hood was associated with the plots (Hansen 2005, 62ff, 67ff, 105ff, 140ff; 2008). The
Bergen area was most likely owned by the king before a town emerged there (Helle
1982, 77–79 with references), so it is likely that only the king possessed the authority
to divide the land into plots. If the suggested date for the regulation is correct, either
Olav Haraldsson (ad 1015–28, later St Olaf) or the Danish/English king Cnut
(ad 1028–34) would have been behind the initiative (Hansen 2005, 221–223).
Beside the plot fences, traces of activity assigned to horizon 2 are sparse. A build-
ing on site 7, two posts dated by dendrochronology (re-used in a building that was
assigned to the following horizon) and a jetty by plot 6/D on site 6 are the most
concrete sources. Except for the jetty, which is well dated, data assigned to horizon 2
are mostly from supplementary sources. Thus some reservations regarding their reli-
ability are appropriate when seen in isolation. Taken together, however, patterns in
the material imply that activity was sparse on the newly established plots. It seems
quite clear that there were people here, but it cannot be established whether settle-
ment was permanent (Hansen 2005, 145ff, 218ff, 224).
Based mainly on ownership of the land known from later medieval written
sources, it is likely that the plots in newly founded Bergen were given as gifts to the
king’s allies, probably land-owning magnates. Hence the first ‘townspeople’ were
most likely magnates, and it was probably magnates or members of their households
that carried out daily activities on the plots (Hansen 2005, 223).
According to the main trend in the material, occupation during horizon 2 was
mainly located by the shores of Veisan and Vågen. The jetty by plot 6/D indicates
that access across the tidal zone to the waterfront was considered important by the
users here. This investment in better working conditions by the waterfront may
MAM 37: 14-hansen – Press

between plans and reality 185

Figure 14.1 Activity in the Bergen area during horizon 2 (B, S and G in brackets refer to Basic,
Supplementary and General background sources as explained in the text; numbers refer to sites
or plots/analytic units) (G Hansen)

reflect at least some of the king’s initial plans for the town. Thus, the following can be
inferred about the intended function of the town plots in horizon 2. The plots were
probably going to be used by magnates, and good working conditions on the water-
front were important. In addition I hold as a premise that the king would benefit from
such activities if they were centred on Bergen. Nevertheless, patterns in the material
from horizons 2 to 5 show that settling the individual plots was not carried out
according to one overall plan. Rather plots were built on as and when the users
decided to do so. Activities on the plots can thus also be characterized as bottom-up
initiatives during horizons 2 to 5.
This is as far as the contemporary sources from Bergen can take us. Axel
Christophersen has suggested that the kings of the early Norwegian central monarchy
sought to control the redistribution of goods traditionally controlled by local elites.
MAM 37: 14-hansen – Press

186 maritime societies of the viking and medieval world

By transferring the redistribution of goods to new urban centres, the king would both
weaken local leaders and acquire a greater share of the wealth (Christophersen 1989,
129, 144). This hypothesis presupposes a strong central king and the use of negative
means of enforcement. Others have suggested that the king attracted merchants to
the early towns by offering plots and ensuring peaceful conditions in the market
place. The king would profit from this by collecting dues in return for protection
(eg Ros 2001, 19; Skovgaard-Petersen 1977, 140ff). This alternative hypothesis
assumes a positive means of enforcement. The two hypotheses entail different under-
standings of a king’s role and motives in founding urban centres and enforcing their
use, but both see the redistribution of goods or trade as an important function of
Norway’s early medieval towns.
A hypothesis that Bergen was founded by the king as a central place for long-
distance trade to be carried out by the local elite, finds an echo in the suggested private
ownership of the town plots, the bottom-up nature of activities, and in the towns-
people’s investments in improved working conditions on the waterfront. I suggest that
one of the king’s intentions for Bergen may have been to establish it as a central place
where goods supplied by individual plot owners (magnates) were collected and
entered into a national or international trade network. Soapstone vessels, dark grey
schist hones and Hyllestad quernstones from western Norway (Baug 2002; 2013;
Carelli 2001; Christophersen 1989; Jensen 1990; Mitchell et al 1984; Myrvoll 1986)
may have been trade goods. The king would benefit from this arrangement by col-
lecting dues on trade and for keeping the peace within the market place. In addition,
the king may have benefited by centralizing the collection of his own dues and veitsler
in Bergen. The surplus from such income could be entered into an international
trading network. The (horizon 1) pier between the middle and southern town areas
was still in use. If it did actually serve as a landing place for the royal estate at
Alrekstad, as suggested above, it may also have been meant to serve as the king’s
landing-place in Bergen. Unfortunately, the available sources cannot reveal the
intended or actual function of the pier, so its significance in relation to the king’s plans
can only be hypothetical. If we consider the location of Bergen on a macro scale, the
area was well located to serve as a central hub for long-distance maritime trade from
western Norway, being closer to England and the continent than Trondheim.
In summary, I suggest that around ad 1020/30 a king planned and founded
Bergen. The hypothesis that it was meant to be a centralized location where magnates
and the king could enter goods into a long-distance trading system finds some sup-
port within the scarce sources (Hansen 2005, 223, 230–231). However, how did the
townspeople receive the king’s suggested plans for Bergen? As already mentioned, the
sources indicate that after the establishment of demarcated plots hardly any major
initiatives were carried into life by the townspeople. Presumably this means that the
king’s plans were not yet well received.

horizon 3 (ad c1070 – 1100)

During horizon 3, several new initiatives were undertaken (Figure 14.2). If the
dates of the material are correct, the middle town area was now also parcelled into
MAM 37: 14-hansen – Press

between plans and reality 187

Figure 14.2 Activity in the Bergen area during horizon 3 (B, S and G in brackets refer to Basic,
Supplementary and General background sources as explained in the text; numbers refer to sites
or plots/analytic units) (G Hansen)

plots. In the northern town area the old plot system was slightly reorganized; boun-
daries were respected in terms of the width of the plots, but Vågen plots were now
lengthened and reached the waterfront. Generally, the focus seems to have shifted
towards the shore of Vågen Bay. The new layout of the northern town area may have
included space for a church (site 23) and a thoroughfare (site 14), but this is not very
well substantiated and is thus merely a hypothesis. Again, as the king owned the land
(Helle 1982 and references therein), he should be seen as being behind the initia-
tives in the town areas. The time span represented by horizon 3 makes Olav Kyrre
(ad 1066–93) the likely initiator. Indeed, according to Heimskringla, Olav Kyrre
founded Bergen. However, the Old Norse term used for founding, setja, is ambiguous
as it can be used either in the sense that an already established settlement is given
jurisdiction or demarcated topographically, or in the sense that something like a
church or a town is actually founded on a virgin site (Bjørgo 1971, 69–73; Helle 1982,
87–90). The archaeological sources suggest that Olav did not found Bergen on a new
MAM 37: 14-hansen – Press

188 maritime societies of the viking and medieval world

site. Rather he invested further in an already existing settlement by reorganizing

the northern town area and including the middle town area within the townscape.
Furthermore, Olav built Christchurch minor (site 3) and founded Christchurch
Cathedral (site 2) at Holmen, thus including this area in Bergen (Hansen 2005,
224–225).
As with the previous horizon, activity traces on the plots are vague during
horizon 3. However, the northern and middle town areas were settled. The distribu-
tion of occupied plots indicates that Veisan-bound plots were second choice to
Vågen-bound plots, and that better landing conditions were considered important
by the townspeople. No artefacts can be assigned to horizon 3 with confidence, but
one can speculate that activities on each plot were carried out as bottom-up initia-
tives. Furthermore, sea transport in general was probably part of the townspeople’s
strategy (Hansen 2005, 225).
Olav’s investments at Holmen have been seen as part of a plan for Bergen to
become the bishop’s residence and an ecclesiastical centre for western Norway (Helle
1982, 90 with references). It has also been suggested that Olav had intentions of
building a royal residence at Holmen, given the combination of royal and ecclesi-
astical centres common to Nordic countries and Ireland (Lidén and Magerøy 1990,
10). The archaeological sources may lend some tentative support to the latter hypoth-
esis, but are too vague for any real conclusions to be made (Hansen 2005, 232).
Olav’s investments in laying out plots and possibly also in infrastructure within the
town area show that he must have had plans for this part of Bergen. The focus on
favourable landing conditions implies that Olav, as suggested for his predecessors,
intended Bergen to be a central place where goods, disposed by magnates, could be
entered into a long-distance trading network. According to ecclesiastical rules,
bishops should have their seat in a town (Helle 1982, 111) and Olav’s investments in
the secular parts of Bergen thus correspond well with a greater plan for Bergen to
become a bishop’s residence.
How did the townspeople receive the king’s plans? The apparently limited
extent of activities in the town area implies that the royal plans were unsuccessful.
The townspeople apparently did not have strong incentives to use the plots in Bergen.
Either the king did not follow through with his plans for activities in the town area
with the necessary enforcement, or he did not possess the resources to apply new
rules and develop new traditions for using the plots (Hansen 2005, 233).

horizon 4 (ad c1100 – 1120s)

The period covered by horizon 4 (Figure 14.3) coincides more or less with the
reign of the joint kings Øystein (ad 1103–23), Sigurd (ad 1103–30) and Olav
(ad 1103–15) Magnusson. The foundation of the Church of the Apostles (site 4) and
of a large timber hall (site 5) at Holmen have been ascribed to Øystein. He is also
thought to be the founder of the Munkeliv abbey at Nordnes (site 43) and possibly of
the Church of St Nicholas in the middle town area (site 32). With the foundation of
Munkeliv one may argue that the Nordnes area was added to the townscape. More-
over, a possible first phase of St Mary’s (site 23) and a pier at site 14 may both be
MAM 37: 14-hansen – Press

between plans and reality 189

Figure 14.3 Activity in the Bergen area during horizon 4 (B, S and G in brackets refer to Basic,
Supplementary and General background sources as explained in the text; numbers refer to sites
or plots/analytic units) (G Hansen)

tentatively associated with royal initiative. St Nicholas’, the pier and St Mary’s are
assigned to horizon 4 by supplementary sources. Even if one or more of these sources
is erroneously attributed to horizon 4, it is still well-documented that major invest-
ments in monuments, institutions and infrastructure were made in Bergen on the
king’s initiative (Hansen 2005, 225–226).
MAM 37: 14-hansen – Press

190 maritime societies of the viking and medieval world

Regarding major initiatives taken by the townspeople during horizon 4, some

plots along Veisan were now occupied, and along the Vågen shoreline occupation is
evident on most documented plots. There was not much pressure on building space,
but townspeople still invested in improved harbour facilities along the Vågen.
Artefacts and pollen show that several plots were permanently settled (Hansen 2005,
218–220). Objects of Norwegian, Low Countries and Germanic origin show that
townspeople of Bergen now had contact with domestic and foreign harbours (Hansen
2005, 210–212). The same can be inferred from Orkneyinga saga, which tells us that
Rognvald Kali from Agder, who later became earl of Orkney, visited Bergen on his
way back from England around ad 1115–20. Moreover, he later went on to Trond-
heim (Holtsmark 1970, 93–94). One tool of trade has also been assigned to horizon 4.
Trade, understood as the exchange of goods in return for material or social profit,
thus seems to have been a real activity on plots in the town area (Hansen 2005,
210–212). The sources also suggest that professional shoemakers, metalworkers and
artisans working with antler, bone (including walrus bone) and horn sporadically
were producing affordable objects in Bergen (Hansen 2005, 234–235).
The great hall at Holmen may indicate that Øystein planned for Bergen to
become a royal residence, probably with royal administrative functions. He thus intro-
duced, or perhaps strengthened, already existing royal functions. Øystein’s founda-
tions also suggest that he wished to strengthen Bergen as an ecclesiastical centre on a
general level. The location of the Church of St Nicholas in the geographical centre of
the existing town area may be Øystein’s contribution to Bergen’s plan as the church
ties the northern and middle areas together across a natural topographical barrier by
Vågen (Hansen 2005, 231–233). St Nicholas was the patron of sailors and merchants,
so this choice of dedication may also imply that Øystein planned to encourage trade
in Bergen (Helle 1982, 116; KLNM 1956–78, XII 288–291).
Is it possible to be more certain as to why Øystein and, as we shall see below,
other 12th-century kings invested so heavily in ecclesiastical institutions in Bergen? It
has been argued that the many 12th-century churches bear witness to the physical
expansion of the town during this period (Helle 1982, 149; 1992, 26). Suggestions
have also been made that the churches were built mainly to serve as places of worship
for different groups of people (Lidén 1993; Lidén and Magerøy 1990, 17ff). These
suggestions presuppose that the churches were built as ‘parish’ churches to serve an
already existing population. The archaeological sources suggest that the churches
were instead built while the town’s physical extent was still quite limited. Accord-
ingly, they cannot have been established to serve an already existing population. It is
likely that there were additional or alternative motives — beyond faith and the aim of
providing places of worship.
During the 12th century several joint kings, rival kings and pretenders to the
throne fought for their positions and must have had a strong desire to show off their
social capacities. Bergen’s churches may therefore have been the result of conspicuous
consumption (Brendalsmo 2001, 262ff). Yet another and more practical issue, how-
ever, is that the foundation of ecclesiastical institutions may have been a positive way
to strengthen Bergen as a central hub for western Norway. Until at least the middle of
the 12th century a church’s founder/patron had a strong influence on matters
MAM 37: 14-hansen – Press

between plans and reality 191

concerning the disposal of land and income under the church (Helle 1995, 22–23;
Krag 1995, 201–203). The churches may have had income from land (Helle 1982, 151;
Helle and Nedkvitne 1977, 221) and after the first decades of the 12th century perhaps
also tithes (Andersen 1977, 335ff). If the king established churches and other ecclesi-
astical institutions in Bergen and gave them land to collect income from, he would
create a new group of landowners. Since these landowners had their base in Bergen,
they would probably use the town as a central location from which to trade their
surplus. This would in turn strengthen Bergen’s position as a centre of commercial
activities. Establishing churches here could thus be the kings’ means to make the
wheels go round (Hansen 2005, 234).
This explanation finds parallels in Anders Andrén’s ‘congested countryside’
theory which, on a more general level, applies to Scandinavian towns between
ad c1000 and ad c1150. Andrén sees the erection of so many churches in early medi-
eval Scandinavian towns as a product of the central kings’ de-centralization of the
right to execute sovereignty (Andrén 1985, 77–81; 1989). The kings’ investments in
Bergen, including the foundation of the many churches and other ecclesiastical insti-
tutions, should not be seen in isolation. On a more general level they should be seen
as investments made in connection with overall strategies, where probably a strength-
ened position for the Norwegian central kingdom is crucial.
Returning to Bergen and the period represented by horizon 4, Øystein is also
known as the founder of churches and other monuments along the sailing route
between Bergen and the fishery districts in northern Norway. This has been seen as
an investment in facilities and infrastructure between the rich fisheries in the north
and Bergen (Helle 1982, 116, note 78). Øystein and his brother Sigurd taxed people
who went fishing in Vågan in northern Norway (Helle 1982, 116; Keyser and Munch
1846, 257–258) and obviously had vested interests in developing fishing activities. In
the light of this, Øystein’s investments in Bergen may also be seen as part of a plan to
strengthen and control the export of stockfish to Europe. In addition, following the
arguments presented for the earlier horizons, it may still have been in the king’s
interest that Bergen functioned as a centre for goods in the hands of magnates.
Does the actual function of the town area correspond with Øystein’s suggested
plans? As already seen, pressure was not intense on building space on the plots; sub-
stantial efforts to improve working conditions on the shore, and access to the water-
front are, however, discerned as bottom-up activities. Furthermore, several plots
were permanently settled. The sources indicating trade imply that goods were now
transferred through the magnates’ plots in Bergen and into a long-distance system.
The townspeople of Bergen were clearly now involved in a wider international net-
work where commodities were exchanged. The king’s plan for the town area to be a
central market place for magnates thus seems to have been fairly successful. Øystein’s
investments in Bergen, and between Bergen and the fisheries to the north, may perhaps
have triggered growing interest among the magnates to use town plots. Conversely,
the scarce sources for horizon 4 cannot throw light upon whether the craftsmen that
produced affordable goods on a sporadic basis were active in Bergen as a direct
response to an overall royal plan. This issue will be discussed further below.
MAM 37: 14-hansen – Press

192 maritime societies of the viking and medieval world

horizon 5 (ad 1120s – c1170)

In ad 1170 the relics of St Sunniva were transferred to Bergen and placed in
Christchurch Cathedral (Storm 1880, 151–152). The construction of the cathedral
may then have been concluded and the bishop’s seat formally attached to Bergen
(Helle 1982, 92, 146; Lidén 1993, 10). During horizon 5 seven churches and two mon-
asteries were erected (Figure 14.4). The churches are St Mary’s (2nd phase?) (site 23),
St Olav’s on the Hill (site 25), St Olav’s in Vågsbunnen (site 39), St Cross (site 40), All
Saints (site 45), St Peter’s (site 24) and St Columba (site 33). The monasteries are
Nonneseter (site 46) and St John’s Abbey on Nordnes (site 44). With these major
initiatives the southern town area and the Nonneseter areas became part of the built-
up area. On the basis of studies of land ownership before and after the establishment
of Bergen, later written sources, and patterns in the archaeological sources, I argue
that the erection of five of the churches and the two monasteries, in addition to the
inclusion of the southern town area, and the Nonneseter area into the townscape, in
all likelihood took place through royal initiative. The transfer of St Sunniva’s remains
to Bergen must also have been carried out through a top-down initiative. The two
remaining churches, St Peter’s and St Columba, may possibly have been erected
through bottom-up initiatives. They are placed within horizon 5 by supplementary
sources, and their presence here should thus be treated with caution (Hansen 2005,
226ff). Even if some of the abovementioned foundations were the result of bottom-up
activities, it is evident that considerable royal investments were made in Bergen
during horizon 5 (Hansen 2005, 226–227). It corresponds more or less with the time
of the civil wars in Norway, where joint kings and claimants to the throne fought
each other. Hence it is not possible to pin a historical name to every initiative; the
dates of the source materials are too wide.
During horizon 5, major investments were also made on the plots by the towns-
people. Plots that have been fully excavated show rows of buildings extending along
the entire length of the property. Substantial investments in harbour facilities sug-
gest that good landing conditions were considered important for the townspeople.
There was now permanent settlement and pressure on building land on almost all
investigated plots along Vågen and Veisan. Vacant building land did, however,
remain by St Nicolas and along the Fløyfjellet Hill, and in the southern town area
settlement also seems to have been sparse. In the latter area excavations are sporadic,
so the sources are not very comprehensive as to the extent of activities. Nevertheless,
the two churches that were erected here were both oriented east to west according to
the Christian standard. This is taken as an indication that land was ample, and
settlement sparse, when the churches were built.
Artefacts indicating direct or indirect contact with eastern and western Norway
were found in 16 and 7 of the 24 artefact-yielding plots/analytic units respec-
tively. Moreover, finds of Greenlandic/Icelandic, English, German, French and Low
Countries origin were retrieved from 22 of the 24 artefact-yielding plots/units.
Altogether these finds make trade a visible part of the townspeople’s strategy. Trade
is also witnessed by evidence of coastal shipping from Bergen to Trondheim during
horizon 5 (Helle 1982, 162; Jónsson 1893–1901, iii 438–439).
MAM 37: 14-hansen – Press

between plans and reality 193

Figure 14.4 Activity in the Bergen area during horizon 5 (B, S and G in brackets refer to Basic,
Supplementary and General background sources as explained in the text; numbers refer to sites
or plots/analytic units) (G Hansen)
MAM 37: 14-hansen – Press

194 maritime societies of the viking and medieval world

The sources also show that a broad spectrum of artisans now worked in Bergen.
They produced affordable objects in antler, bone, leather and fine metal. The distri-
bution and amount of production waste implies that the artisans did not work in
Bergen on a permanent basis. The spatial distribution of finished products shows that
they sold their merchandise to townspeople from many different households (Hansen
2005, 226–228). Service trades are also discerned in the archaeological sources. Pro-
fessional sausage makers were probably active in Bergen during horizon 5 and repre-
sent a new urban trade aimed at serving townspeople or visitors. Innkeepers and
people who had premises to let to, for example, travelling artisans may also belong to
a new group of professionals in Bergen (Hansen 2005, 226–228).
Why did the kings invest further in Bergen? The transfer of the relics of
St Sunniva to Bergen and the foundation of St John’s Augustinian abbey can both
be seen as intended to strengthen or finally establish Bergen as the bishop’s seat in
western Norway. The king’s possible foundation of the Nonneseter convent may
also have strengthened Bergen’s role as an ecclesiastical centre on a general level. As
argued above, the substantial royal investments in ecclesiastical institutions can also
be seen as a way to show off social capacities and as a positive means to trigger and
encourage trade activities in the town. The presence of the gjaldker, who collected
royal income, in Bergen in ad 1159 (Helle 1982, 8; Holtsmark 1961, 692; Jónsson
1893–1901, 604) shows that the king now intended the townspeople to pay dues and
implies royal interest in the blooming commercial activities. The inclusion of the
southern town area into the urban landscape is also interpreted as a royal investment.
Analysis of land ownership in medieval Bergen shows that in the 13th century the
king still owned a large part of this area (Ersland 1989, 257ff; 1994, fig 12). Perhaps
this evidence suggests that the king was now planning to engage himself in activities
of a more ‘common’ character, which could not be conducted from Holmen.
In the northern and middle town areas plots were getting more densely occu-
pied and even less-attractive space was occupied. A plot in Bergen had apparently
become an asset for the townspeople. The large number of plots containing docu-
mented, well-established permanent settlement support this picture. The two possible
privately founded churches may (if they are correctly assigned to horizon 5) sug-
gest that Bergen was now so established that townspeople invested in activities
beyond those conducted from their individual plots. In the northern and middle town
areas the sources imply that trade was an important part of the townspeople’s
activities. This corresponds well with the kings’ aspirations (Hansen 2005, 236).
Did the artisans that worked sporadically on the plots come to Bergen as a direct
response to royal/top-down initiatives? The distribution of production waste and
finished products on different plots in the town area, and the relatively ‘affordable’
character of the products make it plausible that the artisans that we encounter in
the town area were not there primarily to serve the needs of the royal household.
Rather these artisans worked amongst, and had customers within, the social sphere
of the townspeople (cf Hansen 2015). At the current stage of research I find it most
likely that the presence of travelling artisans in Bergen during horizon 5, and perhaps
as early as horizon 4 as well, suggests that the town was now large enough for a visit,
but not yet large enough to provide a base for full-time resident specialists. Their
MAM 37: 14-hansen – Press

between plans and reality 195

presence should primarily be understood as an unintended consequence of the

existing and growing community, rather than as the result of careful planning from
the top down.
Likewise, the existence of new service-related trades should probably be seen as
an unintended consequence of the growing community. As visitors such as crafts-
people and merchants came in increasing numbers to Bergen, the traditional hospi-
tality that travellers could normally expect would no longer suffice. It is highly likely
that enterprising townspeople would see service-related trades as a new way of
making a living in Bergen (Hansen 2010). On the other hand, it cannot be discounted
that the kings encouraged or even initiated such enterprises in Bergen. Indeed,
Øystein Magnusson is known to have established a hostel at Dovre along the travel-
ling route across the mountains there, and the later king, Håkon Håkonarsson
(ad 1204–63), is also thought to have built hostels along the route between Oslo and
Nidaros (Steen 1929, 335 with references). At any rate, the existence of service-related
trades in Bergen shows that the town was beginning to develop a life of its own
and had become a thriving urban community. This development was probably due to
the synergies of the town’s actual role as a bishop’s seat, an ecclesiastical centre, a
prime location for exchanging royal and private goods, a hub for stockfish trade, and
a centre for royal administration.

conclusion
An actor-based approach has provided a varied picture of the changing
dynamics involved while Bergen emerged as a town. We have seen that the process of
urban development was slow and involved royal initiatives as well as investments by
the townspeople. In this interplay between actors from different levels of the social
hierarchy and their wider historical context, Bergen in time developed from a
materialized idea into a thriving urban community characterized by a diversity of
functions.

references
Andersen, P S, 1977 Samlingen av Norge og kristningen av landet 800–1130, Universitetsforlaget,
Bergen
Andrén, A, 1985 Den urbana scenen: städer och samhälle i det medeltida Danmark, CWK Gleerup,
Malmö
Andrén, A, 1989 ‘State and towns in the Middle Ages: the Scandinavian experience’, Theory and
Society 18, 585–609
Baug, I, 2002 Kvernsteinsbrota i Hyllestad: arkeologiske punktundersøkingar i steinbrotområdet i
Hyllestad i Sogn og Fjordane, Norsk bergverksmuseum, Kongsberg
Baug, I, 2013 Quarrying in western Norway: an archaeological study of production and distribu-
tion in the Viking period and Middle Ages, PhD thesis, University of Bergen, Bergen
Bjørgo, N, 1971 ‘Det eldste Bergen’, Sjøfartshistorisk Årbok 1970, 53–130
Brendalsmo, A J, 2001 Kirkebygg og kirkebyggere. Byggherrer i Trøndelag ca 1000–1600, Univer-
sitietet i Tromsø, Tromsø
MAM 37: 14-hansen – Press

196 maritime societies of the viking and medieval world

Carelli, P, 2001 En kapitalistisk anda: kulturella förändringar i 1100-talets Danmark, Almqvist &
Wiksell International, Stockholm
Christophersen, A, 1989 ‘Kjøpe, selge, bytte, gi. Vareutveksling og byoppkomst i Norge ca. 800–
1100: en model’, in A Andrén (ed), Medeltidens fødelse, Gyllenstiernska Krapperupsstiftelsen,
Lund, 109–145
Ersland, G A, 1989 Eit forsøk på rekonstruksjon av grunneigefordelinga i Bergen ved utgangen av
seinmellomalderen, MA thesis, University of Bergen, Bergen
Ersland, G A, 1994 Kven eigde byen? Dr Art. thesis, University of Bergen, Bergen
Giddens, A, 1984 The Constitution of Society: Outline of the Theory of Structuration, University
of California Press, Berkeley
Hansen, G, 2000 Bydannelse og forklaring af sociale fænomener. Individualisme, kollektivisme og
Giddens strukturationsteori, Medeltidsarkeologisk tidsskrift META vol 2000/4, 2–16
Hansen, G, 2005 Bergen c 800–c 1170: The Emergence of a Town, Fagbokforlaget, Bergen
Hansen, G, 2008 ‘Konger og byfolk i det eldste Bergen — byoppkomst i et aktørperspektiv’, in
H Andersson, G Hansen and I Øye (ed), De første 200 årene — nytt blikk på 27 skandinaviske
middelalderbyer, UBAS Universitetet i Bergens arkeologiske skrifter, Institutt for arkeologi,
historie, kulturvitenskap og religion, Bergen, 15–39
Hansen, G, 2010 ‘New pathways for 12th century women in Bergen, Norway?’, in L H Dommasnes,
T Hjørungdal, S Montón Subı́as, M Sánchez Romero and N L Wicker (ed), Situating Gender
in European Archaeologies, Archaeolingua, Budapest, 245–260
Hansen, G, 2015 ‘Itinerant craftspeople in 12th century Bergen, Norway — personal and social
identities’, in G Hansen, S Ashby and I Baug (ed), Everyday Products in the Middle Ages:
Crafts, Consumption and the Individual in Northern Europe c. ad 800–1600, Oxbow,
Oxford, 28–50
Helle, K, 1982 Kongssete og Kjøpstad: Fra opphavet til 1536, Universitetsforlaget, Bergen
Helle, K, 1992 ‘Tidlig byutvikling i Vestnorge’, in I Øye (ed), Våre første byer, Bryggens Museum,
Bergen, 7–30
Helle, K, 1995 Under kirke og kongemakt 1130–1350, H Aschehough & Co, Oslo
Helle, K and Nedkvitne, A, 1977 ‘Norge’, in G A Blom (ed), Urbaniseringsprocessen i Norden:
Middelaldersteder, Universitetsforlaget, Oslo, 189–272
Hjelle, K L, 1986 Paleobotanisk undersøkelse av marine sediment og avfallslag i Veisan — et
bidrag til bosetningshistorien i Bergen, Masters thesis, University of Bergen, Bergen
Hjelle, K L, 2000 Eksisterte det et tettsted i Bergen i Vikingtiden? Bosetningsutvikling basert på
botanisk materiale, Årbok for Bergen Museum 6, 58– 63
Holtsmark, A, 1961 Sverres saga. En tale mot biskopene, Aschehoug, Oslo
Holtsmark, A, 1970 Orkenøyingenes saga, H. Aschehaug og Co, Oslo
Jensen, S, 1990 ‘Handel med dagligvarer i vikingetiden’, in A B Gebauer and S Jensen (ed), Handel
og udveksling i Danmarks oldtid, Forlaget Hikuin, Højbjerg, 119–138
Jónsson, F (ed), 1893–1901 Heimskringla, Samfund til udgivelse af gammel nordisk litteratur,
S L Møllers Bogtrykerri, København
Keyser, R and Munch, P A (ed), 1846 Norges gamle Love indtil 1347, Bd I, Chr. Gröndahl,
Christiania
KLNM, 1956–78 Kulturhistorisk leksikon for nordisk middelalder, Rosenkilde og Bagger, Copen-
hagen
Krag, C, 1995 Vikingtid og rikssamling, Aschehougs norgeshistorie, H Aschehoug og Co, Oslo
Lidén, H E, 1993 ‘De eldste kirker i Bergen’, Bergen Historiske Forening Skrifter 1993, 71–82
Lidén, H E and Magerøy, E M, 1990 Norges Kirker, Bergen, Riksantikvaren, Oslo
Mitchell, J G, Askvik, H and Resi, H G, 1984 ‘Potassium-argon ages of schist honestones from
Viking Age sites at Kaupang (Norway), Aggersborg (Denmark), Hedeby (West Germany) and
MAM 37: 14-hansen – Press

between plans and reality 197

Wolin (Poland), and their archaeological implications’, Journal of Archaeological Science 11,
171–176
Myrvoll, S, 1986 ‘Skien og Telemark — naturressurser, produkter og kontakter i sen vikingtid og
tidlig middelalder’, Viking 49, 161–180
Ros, J, 2001 Sigtuna. Staden, kyrkorna och den kyrkliga organisationen, Institutionen för
arkeologi och antik historia, Uppsala universitetet, Uppsala
Skovgaard-Petersen, I, 1977 ‘Oldtid og vikingetid’, in A E Christensen, H P Clausen, S Ellehøj and
S Mørch (ed), Danmarks historie I, tiden indtil 1340, Gyldendal, Copenhagen, 15–209
Steen, S, 1929 Ferd og fest. Reiseliv i norsk sagatid og middelalder, Frydenlunds Bryggeri,
Aschehoug & Co, Oslo
Storm, G (ed), 1880 Monumenta Historica Norvegiæ, Norsk historisk kjeldeskrift-institutt, Styret
for Kjeldeskriftfondet, Kristiania
MAM 37: 15-sindbaek – Press

chapter 15
STEATITE VESSELS AND THE VIKING
DIASPORA: MIGRANTS, TRAVELLERS
AND CULTURAL CHANGE IN EARLY
MEDIEVAL BRITAIN AND IRELAND

By S ø ren Michael SindbÆk

Carved cooking vessels of steatite, produced in Norway or Shetland, are among

the archaeological signatures of Scandinavian settlement in Britain, Ireland and the
North Atlantic islands in the Viking period. Exchanged over long distances, they are
often highlighted as an example of Viking trade or as emblematic items of Scandi-
navian culture. With the conspicuous exception of Orkney and Shetland, however,
only small numbers of these vessels are known from the British Isles. Yet being few
and far between, these finds deserve particular attention. Tracing the margin of a
distinctive pattern of practice in time and space, they point to the course of a cultural
change. This chapter assembles published finds from Britain and Ireland, and
analyses their distribution in relation to patterns of travel, exchange and household
culture. It aims to trace a network of material-social interaction. Why was this par-
ticular type of utensil imported into some regions, or even adopted as a cultural inno-
vation, while in others it was rejected or swiftly abandoned in favour of other forms?
It is argued that steatite vessels quickly lost their cultural significance to Scandi-
navian settlers in most parts of the British Isles. In contrast to the situation in the
North Atlantic islands, the southern migrants readily adopted local alternatives. The
context, chronology and morphology of finds suggest that the few steatite vessels
noted in England, Ireland and parts of Scotland were left by long-distance travellers
rather than by long-term residents. Together with other aspects of cultural change,
the abandonment of steatite cooking vessels shows that the Scandinavian migrants
did not always identify with the culture of their homelands, and may even have been
keen to distinguish themselves from some aspect of their origins.

steatite vessels
Steatite, or soapstone, is a soft, finely textured rock, found over wide areas of the
north European mountain zones. It occurs abundantly in Norway, western Sweden,
Shetland and Greenland, with some sporadic outcrops in mainland Scotland and
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 199

Ireland (Resi 1979, 116). In regions where it is common, it was worked since pre-
history for purposes including casting moulds, net sinkers, spindle whorls and cook-
ing vessels. For the last use in particular, it gained a high prevalence in parts of
Scandinavia in the last centuries of the 1st millennium ad. Characteristically, the
word for cooking pot in Old Norse (and in various reflexes in modern Scandinavian
languages) is grýta derived from grjót — ‘stone’.
During the 9th century, apparently in pace with Viking expansion, round-
bottomed steatite cooking vessels acquired a distribution into regions which had no
natural supplies. In Denmark, where no steatite vessels are testified before ad 800, they
are found in almost every major excavated settlement from the following two
centuries. This is also the case in Iceland and the Faroe Islands, where, however,
access to the material may have become scarce after the initial landnám phase (Forster
2006, 64). In both these regions the vessels were acquired as an item of long-distance
import. While the settlers in Iceland were undoubtedly accustomed to their use before
the move, populations in Denmark actively adopted stone pots as a cultural innova-
tion. In Atlantic Scotland Scandinavian settlers also brought in Norwegian steatite
vessels, but soon began to exploit outcrops in Shetland to set up a local industry. In
this region steatite vessels were used both before and after the Viking period. Viking-
period vessels, however, can usually be told apart by their characteristic morphology:
circular bowls with rounded bottom, sometimes fitted with iron handles and/or
decorated with simple, incised lines below the rim, although a recent study applying
ICP-MS isotopic analysis to finds from Quoygrew demonstrates some Shetlandic
imitation of Norwegian vessel styles alongside actual imports (Batey et al 2012).
The production of steatite vessels was a part-time job, and not a particularly
specialized one. The makers were most probably farmers who, like later Norwegian
millstone carvers, went out seasonally to produce a supplementary income, or to fulfil
duties to a lord (Grenne et al 2008, 52). The bowls could be worked with readily
available technology and tools — most importantly the pick axe. Yet in terms of
distribution, this rural production fell little short of some of the most advanced
industries in early medieval Europe. Only a few Frankish pottery types, notably the
Badorf and later Pingsdorf wares from the Rhenish Vorgebirge, could be found
regularly more than 100km from their site of production (Janssen 1987; Sanke 2001).
Steatite bowls were regularly taken as far as 300km into Denmark, or even further
across the Atlantic to reach Iceland.
An interesting aspect of steatite exchange is the cultural context of demand and
consumption. In many regions into which steatite vessels were brought, their role as
cooking pots might just as well have been fulfilled by local pottery, whether produced
domestically or by specialist craftsmen. While patterns of contact or trade routes will
have influenced the choice of material, the persistent preference for stone pots in
some regions must be regarded as an expression of cultural practice and identity
(Barrett 2012, 278; Forster and Bond 2004). A mundane object as it was, a cooking
vessel was an integral part of food culture, and a highly exposed item in situations of
communal feasting. This made it a relevant vehicle of cultural expression.
The Scandinavian immigrants, whose arrival in the British Isles during the Viking
period is attested by written sources, place names, language and genetic evidence, have
MAM 37: 15-sindbaek – Press

200 maritime societies of the viking and medieval world

proven difficult to trace archaeologically (Kershaw 2013, 1). This lack of a solid point
of departure has impeded the study of the cultural accommodation of the settlers. As a
remedy, it has been suggested to ‘identify the use of material culture to proclaim a
particular identity, and to invent new ones’ (Richards 2000, 302f). Ideally this analysis
must involve a comparison of many groups of finds and materials. The complexity of
the data, however, demands a transparent and justified basis. This paper shall follow
a single set of links through Viking Age archaeology in Britain and Ireland. Insofar
as any single group of artefacts allows a comprehensible interpretation of cultural
patterns, steatite vessels should do so. Their use was well entrenched in western
Scandinavia by the time of the migrations, and archaeologically they have a high
degree of visibility. Their use as well as abandonment among the migrants can there-
fore be taken as significant pointers to cultural affiliation or change.
Steatite bowls occur through most parts of the Viking world. Finds in Scandi-
navia were first comprehensively surveyed by Heid G Resi (1979). Subsequent studies
have supplemented her catalogue for Denmark and the western Baltic Sea region
(Risbøl 1994; Sindbæk 2005; 2008). Individual studies have considered finds in Russia
and the Ukraine (Khvoshchinskaya 2007), Scotland (Graham-Campbell and Batey
1998), Shetland (Buttler 1989; 1991), and the North Atlantic islands (contributions in
Fitzhugh and Ward 2000). Recently, Amanda Forster has brought together the finds
from the North Atlantic region, including Shetland and Orkney (Forster 2004; 2006;
Forster and Turner 2009). Steatite vessels were also produced in the central Alpine
region. The characteristic lathe-turned products can be traced into the Po plain and
the Adriatic region (Santi et al 2005). The only known point of contact between this
industry and the northern one is, appropriately, the emporium of Dorestad at the
Rhine mouth, where sherds from both sources have been identified (Kars 1983, 86–87).
So far, however, the finds in Britain and Ireland beyond Atlantic Scotland have
attracted limited attention. Shetland and Orkney are the only parts of the British
Isles in which steatite vessels occur abundantly. The profuse occurrence in these
islands aligns with a North Atlantic pattern shared by Iceland and the Faroe Islands
(Figure 15.1). The finds in the Northern Isles have been studied elsewhere (see refer-
ences above), and shall not be considered in detail in the present paper. By com-
parison, the rest of Britain and Ireland clearly form the margin of a distribution zone.
Only 16 sites in this area are known to have yielded steatite vessel sherds from the
Viking period. Yet for the same reason, the locations and contexts in which these
finds occur raise a series of questions: does the context of the finds suggest an early
arrival with first-generation Scandinavian migrants, who were later to adopt local
alternatives? If so, why were these preferred? Or did steatite vessels continue to arrive
over an extended period of time? If so, were they merely brought by new arrivals and
travellers such as sailing merchants, or were they appreciated as prestige items by
communities with Scandinavian roots, or perhaps others?

the western isles

A number of Viking-period settlement sites in the Western Isles have been exca-
vated, and at least five have produced sherds of steatite vessels. Three are located
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 201

Figure 15.1 Finds of Viking-period steatite vessels in northern Europe. The map is based on
the works cited in the text supplemented by the author’s research (S M Sindbæk)

on the island of South Uist, in the farmsteads Cille Pheadair, Bornish and Drimore
Machair. The finds were associated with traces of what are believed to be large,
rectangular timber buildings from the 10th or early 11th centuries (MacLaren 1974;
Parker Pearson et al 2004; Sharples 2003; see Chapter 17). At Traigh Bostadh on the
island of Great Bernera, off Lewis, fragments of a steatite bowl with rivet holes were
also associated with a rectangular building, which partly covered an earlier stone-built
curvilinear house (Neighbour and Burgess 1996). Further finds are noted from the
St Kilda archipelago off the Outer Hebrides. They were found as washed-out deposits
in Village Bay on the main island Hirta (Graham-Campbell and Batey 1998, 77).
MAM 37: 15-sindbaek – Press

202 maritime societies of the viking and medieval world

On a map, these few finds from the Hebrides and beyond appear to align with
the relatively dense distribution noted in the Northern Isles (Figure 15.2). Yet a differ-
ence can be noted between the two regions. In the Hebrides, steatite is an occasional
component of the assemblages. In some contemporary sites (eg the Udal in North
Uist, which has also seen substantial excavations) it appears to be absent (Lane 2007,
15). This pattern contrasts with Shetland and Orkney, where excavated settlements
from the Viking and medieval periods yield substantial numbers of steatite fragments.
The difference is not due to an absence of Scandinavian cultural affiliations in the
Hebrides, as demonstrated by Scandinavian-type graves and hoards. Indeed, lin-
guistic evidence suggests that the northern and central parts of the islands were ‘over-
whelmed’ by Old Norse influence during the 9th century (Graham-Campbell and
Batey 1998, 74ff). Genetic evidence does imply that Scandinavian settlement in the
Western Isles was less intense than in Orkney and Shetland (Goodacre et al 2005).
The evidence for a continuation of local traditions is also stronger in the former, and

Figure 15.2 Finds of

Viking-period steatite
vessels in Britain and
Ireland. See list of finds
for numbers
(S M Sindbæk)
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 203

may point to different relations of power and status between residents and new-
comers (Barrett 2003, 98). However, at the most basic level, the salient point which
distinguishes Viking-period kitchenware of the Western Isles from that of the
Northern Isles is what becomes of the pre-existing tradition of pottery production.
While much of Scotland and parts of England were by and large aceramic
immediately prior to the Viking period, hand-made pottery was in use in the northern
and central parts of the Hebrides (Lane 2007, 13). A striking change occurred at the
onset of the Viking Age. Until the 9th century, the common forms of Hebridean
pottery were tall buckets and shouldered jars with flat bases. In the 9th–10th cen-
turies these forms were replaced by open bowls, often with sagging or rounded bases.
Alan Lane notes that this change coincides with the introduction of rectangular
buildings, and interprets both innovations as evidence of interaction between Scandi-
navian settlers and local populations. He suggests that the new pottery forms imitate
stone or wooden vessels and may indicate ‘new eating and cooking habits’ (Lane
2007, 14–15).
Even though steatite vessels were brought into the Western Isles, and do suggest
an association with settlers of Scandinavian origin, there is no sign that these settlers
had any obstinate preference for such vessels. Rather, they appear to have accepted
the pottery produced locally. Yet in order to be acceptable as an alternative to stone
vessels, pottery had to acquire a new morphology. While the new forms could be
simply imitative, as Lane suggests, it is important to note their relation to another
important feature introduced by the Scandinavians: the longhouse.
The transition into the Viking period in the Hebrides is marked by a change
from round, cellular buildings to rectangular structures, apparently Scandinavian-
type longhouses. Cooking in this style of building took place in a long, shallow
hearth, placed in the centre of the dwelling room. The round-bottomed steatite
vessels could be placed firmly and flexibly on such a hearth, either directly in the
embers or slightly raised over the fire on a tripod (or simply on three stones of
roughly similar size). By contrast, the flat-bottomed vessels used in the pre-Norse
period in the Hebrides would have been inconvenient for this purpose, and were
probably either placed on a grill or platform, or heated by means of pot-boilers.
The morphological changes in Hebridean pottery at the onset of the Viking
period can thus be seen as an accommodation to longhouse-style cooking and cui-
sine and need not reflect a symbolic imitation of steatite vessels. The scarcity of stea-
tite finds in the Western Isles suggests that pottery was mostly considered an
adequate alternative by the Scandinavian settlers. Steatite vessels saw some use, per-
haps due to advantages in particular contexts of use. They could, for example, be
made larger and more capacious than pottery ones. They do not, however, appear to
have been an essential element in the lifestyle of the Scandinavian immigrants.

mainland scotland
In contrast with the Western Isles, few Viking-period settlement sites are known
on the Scottish mainland. A substantial Scandinavian influence is indicated by graves,
hoards and place names, yet no settlement with unambiguous Scandinavian-type
MAM 37: 15-sindbaek – Press

204 maritime societies of the viking and medieval world

structures has yet been investigated in this region (Graham-Campbell and Batey 1998,
70; cf Crawford 2000). Moreover, only a few steatite vessels are known. A reason-
ably secure stray find can be quoted from Ospisdale, Caithness, where part of a
steatite vessel was found c1830 in a site from which a pair of oval brooches was also
recovered. It may either have come from a settlement or from a disturbed grave
(Graham-Campbell and Batey 1998, 68). An assemblage of some 50 steatite artefacts,
mostly vessel sherds, is recorded from excavations at Freswick Links, Caithness.
Colleen Batey, who has published this collection, regards the finds as belonging to the
Late Norse (medieval) period (Batey 1987, 180ff). Yet there are indications of Viking-
period activity on the site, and the steatite finds include sherds that are similar to
Viking-period types, and could belong to this phase (cf Forster 2004, 283–284).
Further south, two steatite sherds are recorded in settlements of a rather differ-
ent kind. One rim sherd of a smallish rounded vessel with a conical piercing from
an iron fitting emerged from excavations in the famous monastery of Whithorn. It
was found in what appears to be a dump of ash dated by context to ad c845–1050
(Nicholson 1997, 464). Excavations in another west Scottish monastic site, at
St Marnock’s church on the island of Inchmarnock, off the west coast of Bute, also
yielded a sherd of a steatite vessel. It was recovered during surface cleaning, and has
no chronological context (Lowe 2008).
The presence of what appear to be typically Scandinavian vessels in these two
monastic sites is intriguing. The steatite sherds could possibly belong to a phase of
secular and culturally mixed settlement following a sack of the sites, as speculated for
another Scottish monastery, Portmahomack on the northern coast of the Tarbat
Peninsula (Carver 2008, 147). Settlements with Scandinavian place names (ending in
-býr) are recorded in the vicinity of both Whithorn and Inchmarnock, probably indi-
cating some level of influence (Crawford 2000, fig 2). Yet neither site has strong
indications of Scandinavian affiliation. The sporadic occurrence of steatite vessel
sherds in large assemblages does not suggest a widespread use or adoption. It marks a
further step away from the occasional occurrence in the Hebrides. Whether seen in a
secular or ecclesiastic context, the rarity of the steatite finds makes it unlikely that
they represent a cultural aspect of the sites’ regular residents.
If the character of the sites gives any clue to the cultural context of the finds, it is
by their richness in non-local objects. Next to towns and emporia, monastic sites are
the most common locations of the Viking period to yield objects that have been
carried through long-distance exchange. Whithorn in particular has a memorable
record of imports: 6th-century African Red Slip ware and east Mediterranean
amphorae, 7th-century Frankish wheel-thrown jugs and pots, and for several centuries
one of the most extensive collections of imported glass anywhere in the British Isles
(Nicholson 1997). To label these sites simply as ‘monasteries’ may mislead our assess-
ment. Stephen Driscoll suggests that Whithorn in particular ‘remodelled itself as a
Viking trading centre’ in the 9th century (Driscoll 2002, 32). Perhaps we can see these
two sites in very general terms as places that were central in communication net-
works and would (for better or worse) be among the first locations to attract long-
distance visitors.
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 205

ireland
Ireland’s Viking-period and medieval archaeology is plentiful and has been
augmented by intensive fieldwork in recent years. A notable collection of steatite
vessels has been found in Dublin, the foremost port of long-distance sailing in the
Irish Sea region, and the most important centre of Scandinavian settlement in Ireland.
Most finds are unpublished, but they count at least half a dozen vessel sherds, some
coming from 11th-century levels at Wood Quay (Wallace 1981, 113; 1985, 136).
Steatite finds from Dublin also include a number of moulds for ingots, one of which
also has a Thor’s hammer matrix (Sheehan et al 2001, 107; cf Ó Rı́ordáin 1971, 73
and fig 21). Furthermore, soapstone spindle whorls were found in the ad 1830–40s
in the famous burial site Kilmainham-Islandbridge, near Dublin (illustrated in Bøe
1940, 49).
Only two other sites in Ireland are known to have yielded Viking-period steatite
vessels. Fragments of a shallow bowl were found in Crossreagh, Co Derry, a metal-
working site beside the Bann estuary (May and Batty 1948). The pieces were redis-
covered some years ago in the Ulster Museum by Cormac Bourke, who had the
identification as soapstone confirmed by the Geological Survey. In Beginish Island,
Co Kerry, on Ireland’s south-west coast, a number of fragments forming an almost
complete vessel were collected as surface finds in an eroding house site in 1959
(Sheehan et al 2001, 106). The fragments were found close to other buildings, which
had been excavated a few years previously, and in which other Norse artefact types
occurred. The bowl was circular, round-bottomed and with a flat rim, 34.8cm in
external diameter, and 17.5cm in height. Remains of an iron handle were attached by
means of two iron escutcheons. Both the internal and external sides were smooth-
ened, but traces of tool marks remained. Darkening on the outer surface as well as
carbonized accumulations on the inside confirm its use as a cooking vessel. The bowl
had seen long use as well as some appreciation, as appears from the fact that several
cracks had been repaired with iron wire. Unfortunately, the vessel was not recovered
from a closed, datable context. In general terms, the site was active from before the
9th century and into the 11th and possibly 12th century. The form of the vessel sup-
ports a date within the Viking period. Sheehan and colleagues (2001, 111) interpret
Beginish as a site of Norse settlers and a waystation on the sea route between
Limerick and Cork. The assemblage clearly indicates a Norse cultural context, rather
than an Irish household.
In assessing the few Irish finds, it must be borne in mind that identified archaeo-
logical evidence for Scandinavian settlement in Ireland is mainly confined to the five
towns of Dublin, Wexford, Waterford, Cork and Limerick. Among these only Dublin
has revealed substantial archaeological strata from before the mid-11th century
(Hurley 2010, 154). Excavations have taken place in more than 800 early medieval
sites in Ireland between 1970 and 2002, yet Scandinavian-type material has been
found in only a handful of investigations outside these five centres (O’Sullivan and
Harney 2008, 55, 110). No steatite is reported from the defended ship encampments
or longphorts investigated in recent years (O’Sullivan and Harney 2008, 112ff). The
implication must be that steatite vessels were not widely used by Ireland’s
MAM 37: 15-sindbaek – Press

206 maritime societies of the viking and medieval world

Scandinavian settlers, and that the occasional finds in Dublin are more likely to have
been used by visitors to the island’s main port, rather than by its regular residents.

england
Just four sites in England are presently known to have yielded Viking-period
finds of steatite vessels. All four were major urban centres in the Danelaw area of
north-east England. Four small sherds from hemispherical bowls were recovered
from excavations in Flaxengate, Lincoln (Mann 1984, 20–21). Two adjoining sherds
show traces of a decorative band formed by three shallow grooves, similar to decora-
tion known from finds in Scandinavia (cf Resi 1979, 45). Only these adjoining pieces
could be assigned a precise date: they appeared from levelling layers of the late 9th
and early 10th century, that is to the earliest generations of activity in the town
(Mann 1984, 55). A further vessel sherd, with a carved, rectangular ingot matrix
showing re-use as an ingot mould, was recovered from excavations in Saltergate,
Trench D (Bayley 2008, 20, fig 18). From Thetford in East Anglia three steatite frag-
ments are known from redeposited Viking-period contexts. They include the base of
a bowl, which has been re-carved as a dish. The carving cuts through soot on the
outside of the bowl, while a small hole for a rivet is evidence of repair (Rogerson and
Dallas 1984, 115).
In Norwich a rim sherd of steatite and another rim with an adjoining body sherd
have been recovered from excavations in Norwich Greyfriars. The latter vessel was
decorated along the rim with a band of three shallow grooves, similar to the example
found in Lincoln. The single sherd is found in a context dated to the 10th–11th
centuries, while the adjoining pieces are from redeposited strata (Mills and Ensom
2007). A further rim sherd is reported from more recent excavations in St Faiths Lane,
Norwich (T Hylton, pers comm). The fact that three out of four retrieved sherds
from Norwich are rim fragments should be noted, as it could mean that the less easily
recognizable body sherds have not been noticed in the excavation. The occurrence of
steatite in Viking Norwich may thus have been slightly more common than the
identified finds suggest.
The largest collection of steatite (or ‘talc-schist’) finds in England, however, is
from York. Twenty sherds from round vessels were recovered in the extensive exca-
vations at 16–22 Coppergate. Mainman and Rogers (2000, 2541–42) identify the
sherds as coming from thick-walled, open bowls. While the surfaces of some sherds
were carefully finished, others are quite rough, presumably due to wear. A number of
sherds are blackened on the outside as well as on the inside, where a residue adheres
to some pieces, testifying to their long use. While a few finds appear in redeposited
layers, the majority are from 10th- or early 11th-century contexts. Interestingly, none
were found in the substantial assemblage attributed to the late 9th and early 10th
century. Amanda Forster notes that the morphology of the vessels is ‘recognizably
Norwegian’, rather than representing the closer Shetland source (Forster 2006, 62).
Together with the absence of rough, unused vessels, this suggests that the pots were
not a general commodity imported for consumers in York, but arrived specifically
with Norwegian travellers.
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 207

A large number of Viking-period settlement sites across England have seen

archaeological excavation. This provides us with a chance to assess the four sites with
steatite vessels rather more directly than other regions. Figure 15.3 shows some 50
Viking-period/Late Saxon towns or rural settlements in England from which sub-
stantial fieldwork is reported. As an approximate measure of the scale of the assem-
blages, the sites are marked by signatures of different size, relating to the quantities
of Late Saxon pottery sherds recovered. This is not a fail-safe measure: the usage of
pottery varies across regions and sites with kilns yield disproportionately large
numbers. Yet for most sites it provides a reasonable comparison. It is clear that all
finds of steatite vessels occur in some of the largest assemblages recorded, in which
rare types must have a better chance of turning up than in sites with fewer finds.
However, the finds also cluster in the regions and places in England for which Scandi-
navian settlement is most strongly attested by written sources and linguistic evidence.
Though there are sites with comparable volumes of finds in the south and west — in
Northampton, Stafford or Steyning, for example — steatite vessels have not been
recovered there.
The fact that steatite sherds are attested in just the four towns of York, Lincoln,
Norwich and Thetford reinforces their association with Scandinavians. Lincoln
began to flourish as an urban centre after the Danish occupation of Mercia in ad 877
(Vince 2001). Indeed, the site and street name ‘Flaxengate’ is of Scandinavian origin
(Fellows-Jensen 2004, 357). Thetford similarly appears to have originated as a winter-
ing camp for the Viking ‘Great Army’ in ad 869–870 (Richards 2004, 105). Norwich,
which only assumed prominence in the 10th century, developed within the English
kingdom, following Edward the Elder’s conquest of East Anglia in ad 917. Yet it is
the centre of a region with particularly rich linguistic evidence of Scandinavian settle-
ment (Hadley 2006, 176). As a major river port on a tidal coast, it was also a natural
hub for North Sea traffic. York, of course, is yet another place with strong Scandi-
navian affiliations. Written sources identify the town as the chosen seat of Viking
kings and their armies, and a preferred destination for merchants, who ‘come from
all parts, but above all from the Danish people’ (the Life of St Oswald, quoted after
Rollason 2004, 322).
Linguistic evidence points to an even stronger coincidence: street names with
Scandinavian elements are found in a number of north English towns, but especially
in Lincoln, Norwich, Nottingham, Leicester, Stamford and York. York has by far the
greatest number of such names, and is also the town from which the largest number
of Scandinavian personal names is attested in 10th-/11th-century sources (Fellows-
Jensen 2004, 358–359). It is striking that the greatest number of steatite sherds is also
found in York, followed by Norwich and Lincoln. Most of Thetford was abandoned
in the later Middle Ages, hence no early street names are preserved. Conversely,
archaeological evidence is limited for Stamford, Leicester and Nottingham. Given
these reservations, the artefactual and linguistic evidence show remarkably similar
patterns.
Beyond the towns, a number of rural settlements have also seen excavation, but
no steatite vessels are known from these. Rural sites generally yield far fewer finds
than towns, but their combined evidence contributes some weight to the observation.
MAM 37: 15-sindbaek – Press

208 maritime societies of the viking and medieval world

The presence of Scandinavian immigrants has been proposed in several sites, on

account of finds or building styles. Yet nothing unusual can be noted in patterns
of vessel use. Sites in the north and east, like Ribblehead, Bryant’s Gill, Simy Folds,
Cottam or Wharram Percy (South Manor), have sparse or non-existent ceramic
assemblages, and may have seen cooking practised in vessels of organic materials by
means of pot boilers. Those in the south and east, like Goltho or North Elmham,
were supplied with wheel-thrown ceramics from urban workshops. Both situations
are consistent with broad, regional patterns witnessed across England, and give no
suggestion of a distinctively Scandinavian kitchen culture (see Richards 2004, 64ff
and Hadley 2006, 104ff with references).
Nonetheless, the arrival of Scandinavians in these regions did coincide with an
abrupt change in the use of cooking vessels, albeit of a very different nature than was
noticed in the Western Isles. The urban centres, which were established in the late
9th century in the Danelaw, saw a rapid establishment of professional pottery work-
shops. The potters were not Scandinavians, but almost certainly of Continental
origin. While their products aligned in some respects with Middle Saxon traditions in
England, they made no allusions to Scandinavian forms (Kilmurry 1980; Vince 1993,
152). The vessels were almost universally globular cooking pots with narrow mouths
and everted rims, echoing types produced in the Rhineland. There are no convincing
examples of open, round-bottomed bowls that might be related to steatite vessels, or
to pottery forms used in Scandinavia.
These newly introduced wheel-thrown pottery forms constitute the bulk of
cooking pots found in Viking-period settlements in northern and eastern England,
regardless of the fact that some inhabitants of these sites are likely to have been
Scandinavians or of Scandinavian ancestry. There is nothing to suggest that stone pots
were particularly valued among these settlers. Indeed, the population of the Danelaw
was quicker in accepting the new, wheel-thrown wares than the inhabitants of
southern England, who had not been subject to a Viking invasion (Brown 2003, 25).
Perhaps paradoxically, the arrival of Scandinavians in Anglo-Saxon England led
in many respects to the establishment of a Frankish pattern of society, with urban
centres, great manorial estates, and indeed the adoption of Frankish-type tableware
produced by craft specialists. This development may have more to do with political
rupture and new structures of power than with the cultural traditions of the immi-
grants. It should also be remembered that the vanguard of the Scandinavian settlers
— the warriors of the Great Army — must have counted many who had spent years
fighting in Frankish realms, and who would have viewed their home culture in the
light of other models.
The absence of steatite vessels in rural sites puts the few sherds found in towns
into perspective. In York and Norwich vessels continued to arrive in the mid- or late
10th century. Yet they are unlikely to have been in regular demand among the towns’
Scandinavian residents; more substantial numbers of finds might have been expected
in that case. Characteristically, finds of Norwegian hones are more numerous and
more widely distributed than steatite vessels, showing that items could indeed be
supplied from Scandinavia if they were in demand. The most likely conclusion is
therefore the same as we have found for other areas — that steatite was used by
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 209

short-term visitors, foreign travellers and merchants, who visited the trading towns
from regions where these vessels were a standard item of domestic culture.

migrants, travellers and cultural change

The Scandinavian migrants, who settled in the British Isles in the 9th and 10th
centuries, were familiar with the use of steatite vessels as items of household culture
from their homelands. The vessels were often acquired from a considerable distance,
and must have been objects of some value and appreciation. Yet this study has found
no evidence that these vessels continued to be used among the settlers in western
and mainland Scotland, England and Ireland, neither as regular daily life objects, nor
as prestige goods. Outside the Northern Isles of Scotland, Scandinavians readily
adopted local alternatives. Some vessels may have been brought by early settlers, but
it is only in towns and other hubs of long-distance communication that we find a
continued, sporadic occurrence throughout the 10th and 11th centuries. This pattern
is more consistent with an interpretation as personal possessions of visitors from
steatite-using regions.
Though settlers arrived from various corners of Scandinavia, the distribution of
steatite vessels does not align with the often assumed predominance of ‘Norwegians’
in the north and ‘Danes’ in the south (eg Kershaw 2013, 227). The supposedly Norwe-
gian settlers in the Irish Sea area cannot be seen to have practised a significantly more
widespread use of steatite than the Danes in England. Populations in Denmark had
keenly adopted steatite vessels by the time of the Viking settlement in England, and
they could have claimed these as symbols of Scandinavian origin and identity with as
much right as the Norwegians. The evidence suggests that neither did.
This recognition necessitates a re-examination of the situation in other regions
(Figure 15.1). The use of steatite vessels in the islands of the North Atlantic certainly
bespeaks a Scandinavian cultural tradition. Yet steatite was arguably a practical and
economic source of cooking vessels in these regions. In Iceland, scarcity of suitable
clay limited the options. In the Northern Isles, Shetland provided a local source of
soapstone. A stronger case for a symbolic significance can be argued for Denmark,
where pottery was produced, yet steatite vessels were also a major import, quite pos-
sibly a prestige item (cf survey in Sindbæk 2008). Yet their abundance also reflects
pre-existing trade with Norway, in which it has been argued that stone vessels served
as a convenient ‘economic ballast’ (Crumlin-Pedersen 1984, 32). They may represent
an added opportunity for Norwegian suppliers of iron and hones, rather than a
specially requested commodity.
In the Slavic settlement areas along the southern Baltic coast steatite sherds are
found only in major ports, where they probably indicate visitors from Scandinavia —
a pattern echoing that noted in most of Britain and Ireland. The Slavs had different
traditions of cooking, feasting and housing, and produced excellent pottery. A similar
pattern is seen in Russia, where individual finds of vessels are noted in Staraja
Ladoga, Gnezdovo and Kiev, all key centres of Scandinavian contacts, similar in this
respect to the Danelaw towns. In central Sweden, perhaps surprisingly, steatite
vessels are also found only in major centres of exchange. To some extent this may
MAM 37: 15-sindbaek – Press

210 maritime societies of the viking and medieval world

be explained by the greater distance by sea from the sources. Yet access to active
steatite quarries in western Sweden was available via great lake systems, and if
these vessels had been objects of particular symbolic value they should have merited
the journey.
Steatite vessels, then, had a close association with Scandinavians insofar as they
were adopted widely among these, and not by other groups in early medieval northern
Europe; yet Scandinavians do not appear to have put a consistent symbolical
emphasis on this association. Scandinavian migrants in most regions of Britain and
Ireland maintained a strong linguistic identity, and sometimes referred to their origin
by use of traditional metal ornaments or art styles (eg Sidebottom 2000; Thomas
2000; Kershaw 2013). Steatite vessels, however, were not embraced as an identity
marker.
The unhesitating abandonment of steatite cooking pots in all regions of Scandi-
navian settlement in Britain and Ireland, apart from Orkney, Shetland, and possibly
northern Scotland, marks a break with tradition in Scandinavia. It points to an
ambiguous relation among the settlers to their ancestral culture. Perhaps they even
perceived steatite vessels as symbols of former conditions which had moved them to
seek a new home country. The transformations certainly point to willingness for cul-
tural change among the migrants, and perhaps to a reason why these have proven so
difficult to trace by means of archaeology.

list of finds
The present study is based on published site reports and catalogues, supple-
mented by information kindly provided by colleagues (see acknowledgements). In
addition to steatite proper, mention is made of vessels of talc or chlorite schist, a rock
with similar qualities, which was used interchangeably with steatite in some regions.
Finds are recorded for England, Ireland, mainland Scotland and the Western Isles. No
finds of steatite vessels have been encountered in Wales or in the Isle of Man. Some
objects are only recently published, while others have been hiding in finds lists, appar-
ently without their proper significance being recognized. Further examples certainly
wait to be discovered.
The research for this article has included systematic searches in a number of
periodicals and databases, including The Society for Medieval Archaeology’s annual
survey of fieldwork and discoveries in Medieval Britain and Ireland, the Royal
Commission on the Ancient and Historical Monuments of Scotland’s database
Canmore (http://canmore.rcahms.gov.uk), the journal Discovery and Excavation in
Scotland, as well as finds reported to the Portable Antiquities Scheme, as screened in
The Viking and Anglo-Saxon Landscape and Economy (VASLE) Project database.
All are made available electronically by the Archaeology Data Service (ADS), see
http://ads.ahds.ac.uk/.
Western Isles
1. Bostadh, Island of Great Bernera, Outer Hebrides. Steatite bowl fragments with
rivet holes recovered from rescue excavation in settlement site 1996. The sherds were
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 211

found in a midden associated with a rectangular building presumably from the Norse
period. The building and the midden partly covered earlier stone-built roundhouses
(Neighbour and Burgess 1996).
2. Village Bay, St Kilda, Outer Hebrides. Stray finds of vessel sherds were found as
washed-out deposits in excavations in Village Bay on the main island Hirta (Graham-
Campbell and Batey 1998, 77).
3. Drimore Machair, South Uist, Outer Hebrides. Sixteen vessel sherds recovered in
excavation at settlement site 1956. The finds were associated with a Norse-type hall
building (Graham-Campbell and Batey 1998, 175–177; MacLaren 1974).
4. Bornais, South Uist, Outer Hebrides. Some dozen fragments of steatite vessels re-
covered from excavations 1994 onwards. Some fragments were excavated in 2003 in
the floor layers of a large house dating to the 10th century (Sharples 2003; Chapter 17).
5. Cille Pheadair, South Uist, Outer Hebrides. Fragments of steatite vessels found in
pits from the earliest phase (early 11th century) of a farmstead excavated 1996–98.
They were found in pits associated with traces believed to mark a timber longhouse.
Further steatite artefacts, including spindle whorls, occur in later phases in associ-
ation with stone and turf walled longhouses (Parker Pearson et al 2004).
A block of worked steatite was recovered from excavation 1912–13 at a settlement
site in Garry Iochdrach, North Uist, Outer Hebrides (Graham-Campbell and Batey
1998, 80). A steatite spindle whorl is reported from a burial site in Cruach Mhór,
Islay, Inner Hebrides (Graham-Campbell and Batey 1998, 89).
Mainland Scotland
6. Freswick Links, Caithness, Highland. Vessel sherds, including a rim sherd with
conical piercing, recovered from long periods of excavations in settlement site. Some
50 steatite artefacts are recorded — except for seven whorls and two weights all
vessels sherds from rounded vessels. The assemblage is partly Late Norse, but likely
to include Viking-period finds (Batey 1987, 180ff).
7. Ospisdale, Caithness, Highland. Part of a steatite vessel found c1830 in site from
which a pair of oval brooches was also recovered (Graham-Campbell and Batey
1998, 68).
8. St Marnock’s church, Inchmarnock, Argyll and Bute. One sherd from a steatite
vessel appeared from surface cleaning during excavations in 2003 in an early medieval
monastic site (Lowe 2008).
9. Whithorn, Dumfries and Galloway. One rim sherd of a smallish rounded vessel
with one conical piercing recovered from excavations 1984–91. Found in a patch of
ash dated by context to ad c845–1050 (Nicholson 1997, 464).
Further possibly Viking-period steatite objects are listed in the Royal Commission on
the Ancient and Historical Monuments of Scotland’s database Canmore. Excluding a
number of presumably Iron Age cups and lamps, the following should be noted.
Dores Parish, Highland — spindle whorl, 1½ inches in diameter (Canmore
ID 13306); Dumbarton Rock, West Dunbartonshire — ribbed steatite bead 1F inches
MAM 37: 15-sindbaek – Press

212 maritime societies of the viking and medieval world

in diameter (Canmore ID 42404); Swordale, Highland — ‘vessel’ found 1880, possibly

a lamp now in the Dunrobin Museum (Canmore ID 14039); Keiss, Caithness, High-
land — spindle whorl from a mound burial (Canmore ID 126356); Reay, Caithness,
Highland — spindle whorl from a burial (Canmore ID 7252). A large flat-bottomed
bowl, 50cm in diameter, rim up to 4cm wide, overall height of 15cm found in Tanera
More, Highland, is morphologically so dissimilar from other Viking-period examples
that it is hardly associated with them (Hubbard and Harden 1989).
Ireland
10. Crossreagh, Co Derry (N): Fragments of a shallow soapstone bowl found in a
metalworking site beside the Bann estuary. Now in the Ulster Museum (acc no
A26505) (May and Batty 1948; C. Bourke, pers comm).
11. Dublin, Co Dublin. At least half a dozen vessel sherds and a number of ingot
moulds recovered from various excavations within the Viking-period town (Sheehan
et al 2001, 107; cf also Ó Rı́ordáin 1971, 73, fig 21; Wallace 1985).
12. Beginish Island, Co Kerry. A number of fragments forming an almost complete,
circular, round-bottomed vessel found 1959 on eroded surface near previously exca-
vated 9th- to 11th-century settlement with Norse material (Sheehan et al 2001, 106).
A further three early medieval sites in Ireland have produced steatite objects. In
the famous burial site Kilmainham-Islandbridge, Co Dublin, three soapstone spindle
whorls were found in the 1830–40s (Bøe 1940). A steatite (or chlorite schist?) spindle
whorl was discovered as a stray find in Tallavbaun, Co Mayo, an early medieval
church site (Corlett 2001, 50). Several spindle whorls of chlorite schist as well as frag-
ments of whorls, some apparently broken during the course of manufacture,
recovered from excavations in Inishkea North, Co Mayo, an early medieval church
site, mainly in midden deposits (Henry 1945; 1952).
Isle of Man
No finds of steatite vessels are recorded by Manx National Heritage (Allison
Fox and Andrew Johnson, pers comm); however, two possibly Viking-period steatite
objects are held in the Manx Museum. A polished stone ring or spindle whorl broken
in two pieces was found near Keeill Vael, Balladoole, Malew (acc no 1954-0298). Part
of a steatite or slate bracelet was excavated from a roundhouse farmstead at
Ballanorris, Arbory (acc no 1975-0115/015).
Wales
No finds are recorded (Mark Redknap, pers comm).
England
13. York, North Yorkshire. Seven sherds from round vessels of steatite and 13 similar
sherds identified as talc schist (including rims from at least three different vessels)
were recovered from excavations 1976–81 at 16–22 Coppergate. No complete forms
can be reconstructed from the fragments. All determinable sherds, however, are from
large, hemispherical vessels, mostly 19–20mm thick, and in one case with a preserved
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 213

iron suspension loop. Five of the steatite sherds from York, and 12 of those of talc
schist, can be assigned to contexts dated to the 10th and early 11th century. The
earliest finds appear in layers dated to c930–c975. Ten ingot moulds of steatite or talc
schist are found in the same excavation (Mainman and Rogers 2000).
14. Lincoln (Flaxengate), Lincolnshire. Four small vessel sherds recovered during
excavations 1972–76. The sherds are too small for the form of the vessel to be deter-
mined more exactly. Two are from undated, redeposited strata and two are from
levelling layers dated to the late 9th or early 10th century (Mann 1984, 20–21).
A vessel sherd re-used as an ingot mould was recovered from Saltergate Trench D
(Bayley 2008, 20, fig 18).
15. Thetford, East Anglia. Two smallish sherds, as well as the base of a bowl, which
had been re-carved to be used as a dish, c18cm in diameter, were recovered during
excavations 1948–59. The three fragments were found in various pits and fillings,
which do not provide a specific date or context. The geological attribution is attested
by mineralogical analysis (Rogerson and Dallas 1984).
16. Norwich, Norfolk. One rim sherd of a steatite vessel, c30cm in diameter, was
found in excavations in 1990–95 in a context dated to the 10th–11th centuries. A rim
with adjoining body sherd from a talc schist vessel of similar size, decorated along
the rim with three shallow grooves, was recovered from redeposited strata in the
same excavation (Mills and Ensom 2007). A further rim sherd is reported from more
recent excavation in St Faiths Lane. This piece is from a smallish hemispherical
vessel, c22cm in diameter, with an upright, flat-topped rim. Attached to the sherd are
remains of iron fittings, fastened with a rivet or nail through a hole in the vessel (T
Hylton, pers comm).
An irregular steatite (?) spindle whorl from Eccleston, Cheshire, found together with
supposedly Viking-period detector finds is recorded by the Portable Antiquities
Scheme (PAS find nr LVPL2071), and listed in the Viking and Anglo-Saxon Land-
scape and Economy (VASLE) Project database.

comparative sites for england

The following published sites are included for comparison in the map
(Figure 15.3) (for full references, see Sindbæk 2012): Botolphs, Sussex (Gardiner
1990); Bryant’s Gill, Cumbria (Dickinson 1985); Cadbury Castle, Somerset (Alcock
1995); Caister-on-Sea, Norfolk (Darling 1993); Canterbury, Kent (Frere 1983; 1987;
Macpherson-Grant 1993); Catholme, Staffordshire (Losco-Bradley and Kinsley 2002);
Cheddar, Somerset (Rahtz 1979); Chester, Cheshire (Ward 1994); Colchester, Essex
(Crummy 1981); Cottam, Yorkshire (Richards 2001); Derby, Derbyshire (Bain 2006);
Exeter, Devon (Allan 1984); Faccombe Netherton, Hampshire (Fairbrother 1990);
Gloucester, Gloucestershire (Heighway et al 1979); Goltho, Lincolnshire (Beresford
1987); Hall Farm, Baston, Lincolnshire (Taylor 2003); Hereford, Herefordshire
(Shoesmith 1985; Thomas and Boucher 2002); Ipswich, Suffolk (Wade 1988); Lang-
hale, Norfolk (Wade 1976); Leicester, Leicestershire (Courtney 1998); Lincoln
MAM 37: 15-sindbaek – Press

214 maritime societies of the viking and medieval world

Figure 15.3 Finds of Viking-period steatite vessels in England compared to recorded

excavations in towns and rural settlements. Legend: ‘H’ steatite vessels; ‘˚’ kiln sites. Size of
symbols: less than 100 (smallest), more than 100, 1000, and 10,000 (largest) Late Saxon pottery
sherds in published assemblage (S M Sindbæk)

(Flaxengate), Lincolnshire (Gilmour 1988; Mann 1982; Perring 1981; Young et al

2006); Little Paxton, Cambridgeshire (Addyman 1969); London, Borough of London
(Vince 1990); Mawgan Porth, Cornwall (Bruce-Mitford 1997); Michelmerch, Hamp-
shire (Addyman et al 1972); Middle Harling, Norfolk (Rogerson 1995); Newark-on-
Trent, Nottinghamshire (Kinsley 1993); North Elmham, Norfolk (Wade-Martins
1980); Northampton, Northamptonshire (Williams 1979); Norwich, Norfolk (Emery
2007); Nottingham, Nottinghamshire (Hall 1989); Oxford, Oxfordshire (Dodd 2003);
Portchester Castle, Hampshire (Cunliffe 1977); Raunds (Audouy and Chapman
2009); Ribblehead, Yorkshire (King 1978); Sedgeford, Norfolk (Cabot et al 2004);
Shrewsbury, Shropshire (Baker 2003; 2008; Carver 1978); Simy Folds, Yorkshire
(Coggins et al 1983); Springfield Lyons, Essex (Tylor and Major 2005); St Neots,
Cambridgeshire (Addyman 1972); Stafford, Staffordshire (Carver 1981; 1983; forth-
coming); Stamford, Lincolnshire (Mahony et al 1982); Steyning, Sussex (Gardiner
1990; 1993; Gardiner and Greatorex 1997); Thetford, Norfolk (Rogerson and Dallas
1984); Torksey, Lincolnshire (Barley 1964; 1981); Waltham Abbey, Essex (Huggins
1976); Wharram, Yorkshire (Hurst 1983); Worcester, Worcestershire (Dalwood and
Edwards 2004; Dalwood et al 1992); York, Yorkshire (Mainman and Rogers 2000);
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 215

Flixborough, Lincolnshire (Loveluck et al 2007–2009); Tamworth, Staffordshire

(Gould 1967; 1968).

note
1 I am grateful to the following for valuable hints Graham-Campbell, Tora Hylton, Andrew John-
and discussions, and for guidance in unfamiliar son, Ray Moore, Hilary Paterson, Mark Redknap
country: Steven Ashby, James Barrett, Cormac and John Sheehan.
Bourke, Amanda Forster, Allison Fox, James

references
Barrett, J H, 2003 ‘Culture contact in Viking Age Scotland’, in J H Barrett (ed), Contact, Con-
tinuity, and Collapse: The Norse Colonization of the North Atlantic, Brepols, Turnhout,
73–111
Barrett, J H, 2012 ‘Being an islander’, in J H Barrett (ed), Being an Islander: Production and
Identity at Quoygrew, Orkney, ad 900–1600, McDonald Institute for Archaeological
Research, Cambridge, 275–292
Batey, C E, 1987 Freswick Links, Caithness: A Re-appraisal of the Late Norse Site in its Context,
British Archaeological Reports, British Series 179, vols 1–2, Archaeopress, Oxford
Batey, C E, Forster, A K, Jones, R E, Gaunt, G, Breckenridge, F J, Bunbury J and Barrett, J H, 2012
‘Local availability and long-range trade: the worked stone assemblage’, in J H Barrett (ed),
Being an Islander: Production and Identity at Quoygrew, Orkney, ad 900–1600, McDonald
Institute for Archaeological Research, Cambridge, 207–228
Bayley, J, 2008 Lincoln: Evidence for Metalworking on Flaxengate and other Sites in the City,
English Heritage, Portsmouth
Bøe, J, 1940 Norse Antiquities in Ireland, Viking Antiquities in Great Britain and Ireland 3,
H Aschehoug & Co, Oslo
Brown, D H, 2003 ‘Bound by tradition: a study of pottery in Anglo-Saxon England’, in D Griffith,
A Reynolds and S Semple (ed), Boundaries in Early Medieval Britain, Oxford University
School of Archaeology, Oxford, 21–27
Buttler, S J, 1989 ‘Steatite in Norse Shetland’, Hikuin 15, 193–206
Buttler, S J, 1991 ‘Steatite in the Norse North Atlantic’, Acta Archaeologica 61, 228–232
Carver, M, 2008 Portmahomack: Monastery of the Picts, Edinburgh University Press, Edinburgh
Corlett, C, 2001 Antiquities of West Mayo, Wordwell Press, Bray
Crawford, B E, 2000 ‘The Scandinavian contribution to the development of the kingdom of
Scotland’, Acta Archaeologica 71, 123–134
Crumlin-Pedersen, O, 1984 ‘Ships, navigation and routes in the reports of Ohthere and Wulfstan’,
in N Lund (ed), Two Voyagers at the Court of King Alfred, William Sessions Limited, York,
30–42
Driscoll, S T, 2002 Alba: The Gaelic Kingdom of Scotland ad 800–1124, Birlinn, Edinburgh
Fellows-Jensen, G, 2004 ‘The Anglo-Scandinavian street-names of York’, in R Hall (ed), Aspects of
Anglo-Scandinavian York, Council for British Archaeology, York, 357–371
Fitzhugh, W W and Ward, E I (ed), 2000 Vikings: The North Atlantic Saga, Smithsonian and
National Museum of Natural History, Washington
Forster, A K, 2004 Shetland and the trade of steatite goods in the North Atlantic region during the
Viking and early medieval period, Unpublished PhD thesis, University of Bradford, Bradford
Forster, A K, 2006 ‘Steatite, resource control and the Orkney earldom’, in O Owen (ed), The
World of Orkneyinga Saga: ‘The Broad-cloth Viking Trip’, Orcadian, Kirkwall, 55–74
MAM 37: 15-sindbaek – Press

216 maritime societies of the viking and medieval world

Forster, A K and Bond, J M, 2004 ‘North Atlantic networks: preliminary research into the trade of
steatite in the Viking and Norse periods’, in R A Housley and G Coles (ed), Atlantic Connec-
tions and Adaptations: Economies, Environments and Subsistence in Lands Bordering the
North Atlantic, Oxbow Books, Oxford, 218–229
Forster, A K and Turner, J (ed), 2009 Kleber: The Archaeology of Shetland’s Steatite Industry from
the Neolithic to the Present Day, Shetland Amenity Trust, Lerwick
Goodacre, S, Helgason, A, Nicholson, J, Southam, L, Ferguson, L, Hickey, E, Vega, E, Stefáns-
son, K, Ward, R and Sykes, B, 2005 ‘Genetic evidence for a family-based Scandinavian settle-
ment of Shetland and Orkney during the Viking periods’, Heredity 95, 129–135
Graham-Campbell, J and Batey, C E, 1998 Vikings in Scotland: An Archaeological Survey, Edin-
burgh University Press, Edinburgh
Grenne, T, Heldal, T, Meyer, G B and Bloxam, E G, 2008 ‘From Hyllestad to Selbu: Norwegian
millstone quarrying through 1300 years’, NGU Special Publications 11, 47–66
Hadley, D M, 2006 The Vikings in England: Settlement, Society and Culture, Manchester Univer-
sity Press, Manchester
Henry, F, 1945 ‘Remains of the Early Christian period on Inishkea North, Co. Mayo’, Journal of
the Royal Society of Antiquaries of Ireland 75, 127–155
Henry, F, 1952 ‘A wooden hut on Inishkea North, Co. Mayo (Site 3, House A)’, Journal of the
Royal Society of Antiquaries of Ireland 82, 163–178
Hubbard, S and Harden, G, 1989 ‘Tanera More (Lochbroom parish), steatite bowl’, Discovery and
Excavation in Scotland 1989, 32
Hurley, M F, 2010 ‘Viking elements in Irish towns: Cork and Waterford’, in J Sheehan and D Ó
Corráin (ed), The Viking Age: Ireland and the West, Four Courts Press Dublin, 154–164
Janssen, W, 1987 Die Importkeramik von Haithabu, Wachholtz, Neumünster
Kars, H, 1983 ‘Early-medieval Dorestad, an archaeo-petrological study part VIII: summary of the
petrographical results’, Berichten van de Rijksdienst voor het Oudheidkundig Bodemonder-
zoek 33, 83–94
Kershaw, J, 2013 Viking Identities: Scandinavian Jewellery in England, Oxford University Press,
Oxford
Khvoshchinskaya, N V, 2007 ‘Kamennye litejnye formochki v kollektsii Ryurikova Gorodishcha i
steatitovye nakhodki na Rusi [Small steatite moulds from the collection from Ryurik
Gorodishche and other steatite finds in Rus]’, in E N Nosov and A E Musin (ed), The Origins
of the Russian State: To the 30th Anniversary of the Archaeological Studies of Novgorod
Ryurik Gorodishche and the Novgorod Oblast Archaeological Expedition, Dmitrij Bulanin,
St Petersburg, 199–202
Kilmurry, K, 1980 The Pottery Industry of Stamford, Lincolnshire, c. ad 850–1250: Its Manufac-
ture, Trade, and Relationship with Continental Wares, with a Classification and Chronology,
British Archaeological Reports, British Series 84, Archaeopress, Oxford
Lane, A, 2007 Ceramic and Cultural Change in the Hebrides ad 500–1300, University of Cardiff,
Cardiff
Lowe, C, 2008 Inchmarnock: An Early Historic Island Monastery and its Archaeological Land-
scape, Society of Antiquaries of Scotland, Edinburgh
MacLaren, A, 1974 ‘A Norse house on Drimore Machair, South Uist’, Glasgow Archaeological
Journal 3, 9–18
Mainman, A J and Rogers, N S H, 2000 Craft, Industry and Everyday Life: Finds from Anglo-
Scandinavian York, Council for British Archaeology, York
Mann, J E, 1984 Early Medieval Finds from Flaxengate, Council for British Archaeology for the
Lincoln Archaeological Trust, London
MAM 37: 15-sindbaek – Press

steatite vessels and the viking diaspora 217

May, A M and Batty, J, 1948 ‘The Sandhill cultures of the River Bann estuary, Co. Londonderry’,
Journal of the Royal Society of Antiquaries of Ireland 78, 130–156
Mills, J M and Ensom, P C, 2007 ‘Stone bowls/dishes’, in P A Emery (ed), Norwich Greyfriars: Pre-
Conquest Town and Medieval Friary, Historic Environment, Dereham, 155–156
Neighbour, T and Burgess, C, 1996 ‘Traigh Bostadh, (Uig Parish)’, Discovery and Excavation in
Scotland 1996, 113–114
Nicholson, A, 1997 ‘The stone artefacts’, in P Hill (ed), Whithorn and St Ninian: The Excavation
of a Monastic Town, 1984–91, Sutton Publishing for The Whithorn Trust, Stroud, 447–464
Ó Rı́ordáin, B, 1971 ‘Excavations at High Street and Winetavern Street, Dublin’, Medieval Archae-
ology 15, 73–85
O’Sullivan, A and Harney, L, 2008 Early Medieval Archaeology Project: Investigating the Char-
acter of Early Medieval Archaeological Excavations, 1970–2002, UCD School of Archae-
ology, Dublin
Parker Pearson, M, Smith, H, Mulville, J and Brennand, M, 2004 ‘Cille Pheadair: the life and times
of a Norse-period farmstead c. 1000–1300’, in J Hines, A Lane and M Redknap (ed), Land,
Sea and Home: Proceedings of a Conference on Viking-period Settlement, at Cardiff, July
2001, Maney/Society for Medieval Archaeology, Leeds, 235–254
Resi, H G, 1979 Die Specksteinfunde aus Haithabu, Wachholtz, Neumünster
Richards, J D, 2000 ‘Identifying Anglo-Scandinavian settlements’, in D M Hadley and J D Richards
(ed), Cultures in Contact: Scandinavian Settlement in England in the Ninth and Tenth
Centuries, Brepols, Turnhout, 295–309
Richards, J, 2004 Viking Age England, Tempus, Stroud
Risbøl, O, 1994 ‘Socialøkonomiske aspekter ved vikingetidens klæberstenshandel i Sydskandi-
navien’, Lag 5, 115–169
Rogerson, A and Dallas, C, 1984 Excavations in Thetford, 1948–59 and 1973–80, Norfolk
Archaeological Unit, Gressenhall
Rollason, D, 2004 ‘Anglo-Scandinavian York: the evidence of historical sources’, in R Hall (ed),
Aspects of Anglo-Scandinavian York, Council for British Archaeology, York, 305–324
Sanke, M, 2001 ‘Gelbe Irdenware’, in H Lüdtke and K Schietzel (ed), Handbuch zur mittelalter-
lichen Keramik in Nordeuropa, Band 1: Text, Wachholtz, Neumünster, 271–428
Santi, P, Antonelli, F and Renzulli, A, 2005 ‘Provenance of medieval pietra ollare artefacts found in
archaeological sites of central-eastern Italy: insights into the alpine soapstone trade’,
Archaeometry 47, 253–264
Sharples, N, 2003 ‘Bornish (South Uist parish)’, Discovery and Excavation in Scotland 2003, 132
Sheehan, J, Stummann Hansen, S and Ó Corráin, D, 2001 ‘A Viking Age maritime haven: a
reassessment of the island settlement at Beginish, Co. Kerry’, The Journal of Irish Archae-
ology 10, 93–119
Sidebottom, P C, 2000 ‘Viking Age stone monuments and social identity in Derbyshire’, in
D M Hadley and J D Richards (ed), Cultures in Contact: Scandinavian Settlement in England
in the Ninth and Tenth Centuries, Brepols, Turnhout, 213–236
Sindbæk, S M, 2005 Ruter og Rutinisering. Vikingetidens Fjernhandel i Nordeuropa, Multivers,
Copenhagen
Sindbæk, S M, 2008 ‘The lands of Denemearce: cultural differences and social networks of the
Viking age in south Scandinavia’, Viking and Medieval Scandinavia 4, 169–208
Sindbæk, S M 2012 ‘Viking disruptions or growing integration? Contextualising communication
networks in the 10th century North Sea’, in S Kleingärtner and G Zeilinger (ed) Raumbildung
durch Netzwerke. der Ostseeraum zwischen Wikingerzeit und Spätmittelalter aus archäo-
logischer und geschichtswissenschaftlicher Persperktive, Dr Rudolf Habelt GMBH, Bonn,
19–38
MAM 37: 15-sindbaek – Press

218 maritime societies of the viking and medieval world

Thomas, G, 2000 ‘Anglo-Scandinavian metalwork from the Danelaw: exploring social and cultural
interaction’, in D M Hadley and J D Richards (ed), Cultures in Contact: Scandinavian
Settlement in England in the Ninth and Tenth Centuries, Brepols, Turnhout, 237–255
Vince, A, 1993 ‘Forms, functions and manufacturing techniques of late 9th- and 10th-century
wheelthrown pottery in England and their origins’, in D Piton (ed), La Céramique du Vème au
Xème siècle dans l’Europe du Nord-Ouest, Centre de recherches archéologiques et de
diffusion culturelle, Saint-Josse-sur-Mer, 151–164
Vince, A, 2001 ‘Lincoln in the Viking Age’, in J Graham-Campbell, R Hall, J Jesch and
D N Parssons (ed), Vikings and the Danelaw: Select Papers from the Proceedings of the
Thirteenth Viking Congress, Oxford University Press, Oxford, 157–179
Wallace, P F, 1981 ‘Dublin’s waterfront at Wood Quay: 900–1317’, in G Milne and B Hobley (ed),
Waterfront Archaeology in Britain and Northern Europe, Council for British Archaeology,
London, 109–118
Wallace, P F, 1985 ‘The archaeology of Viking Dublin’, in H Clarke and A Simms (ed), The
Comparative History of Urban Origins in Non-Roman Europe, British Archaeological
Reports, International Series 255, Archaeopress, Oxford, 103–146
MAM 37: 16-grifths – Press

chapter 16
STATUS AND IDENTITY IN NORSE
SETTLEMENTS: A CASE STUDY
FROM ORKNEY

By David Griffiths

The predominant theme in discussions of Norse Scotland and other parts of the
North Atlantic is the date of arrival, and the subsequent cultural impact, of Scandi-
navian settlers. Yet identifying early (pre-10th-century) Norse settlement has proven
particularly tricky and elusive in archaeological terms. Even where there are arguably
elements of continuity from pre-Norse times, excavations on settlement sites have
yielded much more extensive and coherent data on occupation, economy and land-
scape formation for the later 10th and 11th centuries onwards. There are acknowl-
edged dangers in seeing cultural and technological innovation exclusively in terms of
migration-led introduction. Many, if not most, of the major structural changes
wrought to patterns of settlement and resource exploitation occurred, not in the early
Viking period, but around ad 1000 or even later, as shown by the work of James
Barrett and colleagues on fishing and the marine economy (eg Barrett et al 2000;
Barrett 2012). Thus many of the changes towards a stable political geography in
Orkney, an archipelago off northern Scotland, may be seen as coincident with the
development of a semi-autonomous Norse earldom in the Middle Ages, rather than
as true ‘Viking’ innovations of the 9th or early 10th centuries. Increased economic
diversification and social complexity associated with the emergence of large central
farms within each parish, all components of a consolidated Norse presence, are major
themes to be addressed in their own right.
Much remains to be done to disentangle the social and settlement hierarchy. In
particular, there seems yet to be little methodological common ground when it comes
to gauging perceptions of past status. Work on place names, and historical structures
of territorial lordship, has yet to be fully integrated with studies of climate, soils and
environmental resources. The role of material culture in shaping the development of
a mature, landed Norse society is arguably also as yet underexplored. This paper
seeks not to establish a new interdisciplinary platform for ongoing research on these
questions (which would be far too ambitious an objective), but to offer some
thoughts on the period of consolidation of Norse presence between the mid-10th and
early 13th centuries ad, based on recent findings in a landscape research project in
MAM 37: 16-grifths – Press

220 maritime societies of the viking and medieval world

Orkney. As summarized below, the project investigates a complex of multi-focal or

neighbouring mounded settlement sites, with large stone buildings, engaged in
intensive terrestrial and marine exploitation, which has left its mark on the surround-
ing landscape.

the birsay-skaill landscape project

Since 2003, Oxford University, with the support of Historic Scotland and
Orkney Islands Council, has been undertaking landscape archaeological research on
the west-facing coastal areas of Orkney’s north-west Mainland: the Birsay-Skaill
Landscape Archaeology Project.1 Since 2004, this has involved a partnership with
Orkney College Geophysics Unit (OCGU). An essential underpinning of the project is
to contextualize and enhance existing archaeological knowledge from excavations
and chance discoveries with new techniques of extensive landscape survey and
sample excavation. The areas selected for study are those with the greatest settlement
and economic attractions in the past, yet also those most vulnerable to coastal
erosion, namely the three bay hinterlands of Birsay, Marwick and Skaill (Figure 16.1).
In each case we know most about their archaeological potential through destructive

Figure 16.1 Map of

Bay of Skaill and (inset)
Orkney (# University of
Oxford)
MAM 37: 16-grifths – Press

a case study from orkney 221

exposures in the shore erosion zone, some of which were subjected to systematic
excavation, albeit small in area extent. Each of these areas is a low-lying opening in
the otherwise precipitous red sandstone Atlantic frontage of the Orkney Mainland’s
west coast. At Birsay and the Bay of Skaill in particular, the shore is backed by links
(extensive deposits of calcareous wind-blown sand, now mostly grassed over to form
machair), a type of landscape overburden which, except at erosive margins and in
deflation exposures, has hitherto been regarded as making the extensive detection and
mapping of underlying archaeological deposits very difficult. This chapter will focus
on one of these areas, the Bay of Skaill, a focus of major survey and excavation by the
project which has yielded much evidence regarding late Viking Age and medieval
society.
The Bay of Skaill (Figure 16.1), in Sandwick Parish 9km south of Birsay Bay, is a
deep erosive opening in the West Mainland Atlantic façade, with a history of dune
formation stretching throughout the Holocene (De la Vega Leinert et al 2000). Its
circular shape reflects its probable origins as a freshwater loch opened to the sea, and
it is best known as the site of Skara Brae, the Neolithic settlement (often termed a
‘village’) now open to the public and at the centre of a portion of the Heart of Neo-
lithic Orkney World Heritage Site. As ‘Brae’ (Scots: ‘hill’) indicates, until cleared and
made accessible to visitors in the 1920s and 1930s, it was located within an upstand-
ing mound of sand and archaeological settlement debris. This was part-destroyed by
the encroaching sea in 1850, leading to the discovery of the Neolithic settlement.
Another sandy mound, 100m to the west of Skara Brae, is still partially intact to its
full height, but badly eroded on its northern sea-facing side with masonry and
middens showing in exposed section. A Viking grave was found in its upper layers in
1888. The mound has received the sporadic attention of archaeologists and soil
scientists in the last few decades (Griffiths 2006; Morris 1985). The World Heritage
Area has recently been subjected to a full geophysical survey by Orkney College on
behalf of Historic Scotland.
Most of the work of the Birsay-Skaill Landscape Project has so far been concen-
trated on the north side of the Bay of Skaill, where fewer indications of archaeo-
logical potential were previously known. A badly eroding broch (a monumental Iron
Age domestic building, predating the Viking Age by several centuries) known as
Verron stands on the cliff edge at the outermost north-western point of the bay, and
further inland in a former sand quarry in ‘Sand Fiold’, the extensive sand deposits
lining the higher ground north-east of the bay, an apparently isolated prehistoric
rock-cut cist grave was discovered and excavated in 1989 (Dalland 1999).
Beginning in 2003, a swathe of land measuring 1000m by 400m on the northern
edge of the extensive sandy links surrounding the centre of the bay has been sub-
jected to geophysical, topographic and remote-sensing survey and mapping, supple-
mented by coastal survey and documentary research. These investigations have
identified and mapped an extensive cluster of sandy mounds of archaeological char-
acter, which number at least four (and if the morphology of two of these is taken as
suggesting a coalescence of more than one individual mound, their number may be up
to seven). The mounds are located in an arc around the bay hinterland on the edge of
the sandy links, fronting a system of freshwater streams connecting dune slacks
MAM 37: 16-grifths – Press

222 maritime societies of the viking and medieval world

within the links to the sea. The most prominent mound, although not the most exten-
sive in area, is the only one to have a name — the ‘Castle of Snusgar’. It is attested in
records going back at least to the Old Statistical Account (OSA) of 1795, which
recorded some stone masonry exposed on the flank of the mound (Sinclair 1791–99,
458), and the first Ordnance Survey maps.2 On the earliest known map of the bay,
Walden’s Plan of the Links of Skaill (1772) (reproduced in Lysaght 1974), Snusgar is
not annotated by name, but an apparently upstanding structure on a large mound is
depicted at approximately its location. This is also the reputed site of the Skaill
Hoard, the largest Viking silver hoard yet found in Scotland, which was composed
of at least 115 pieces of bullion and 21 coins, the date of which suggests deposition
between ad 950 and 980 (Graham-Campbell 1995, 108–127; Griffiths 2013). The
initial unearthing of silver items in early March 1858 occurred when a boy engaged in
kelping3 delved into a rabbit-hole with an iron hook and brought out a piece of silver
wire; this was followed by more discoveries over the following days as a result of
extensive sand sifting by numerous other people. It was over two weeks later that a
Kirkwall antiquarian, George Petrie, heard about the discovery, and began purchas-
ing and collecting silver items from their finders, subsequent to which they entered
the collections of the National Museum of Scotland.4 The precise location of the
hoard was not accurately recorded at the time, however, and sufficient doubt persists
that it is not possible to determine the location except within an area (measuring
roughly 300m north–south) between the ‘Castle of Snusgar’ and the parish kirk of
St Peter, which is the most precise definition given by Petrie himself at the time.
The results of geophysical and topographic survey on the mound cluster to the
north of the bay, whilst showing promising evidence of concentrations of settlement
marked by extensive magnetic contrast (normally derived from in situ burning associ-
ated with domestic and industrial activity), did not in themselves provide a date or a
cultural attribution for the deposits identified. In order to explore the nature of the
geophysical anomalies further, selective excavation was undertaken on top and on
the flanks of Snusgar in 2004–06. The excavation trenches on top of Snusgar revealed
a badly eroded and robbed-out, but nonetheless impressive, series of masonry
features including an east–west wall of 1.3m width with external facings and a rubble
core. Associated with these features were extensive midden layers, rich in Norse-
period finds, sandwiched between lenses of wind-blown sand. There had been an
extensive building complex here; its reduced and partial state of preservation is
probably partly due to stone robbing for the construction of field dykes in the 18th
and 19th centuries. The stratigraphy of Snusgar was investigated in deeper section in
another trench on the south-eastern flank of the mound in 2005–06. This confirmed
that the mound was constructed of layers of domestic refuse and peat ash, with inter-
mittent episodes of clean wind-blown sand. It is clear that the struggle to maintain
stability and fertility against blanketing influxes of sand was an unremitting one for
the inhabitants. The layers were generally flat in profile, and soft surfaces vulnerable
to erosion had been deliberately stabilized by the admixture of stones and organic
material. The latter has provided a significant return of archaeobotanical, archaeo-
zoological and micro-artefactual data, and residues such as hammerscale from metal-
working, through systematic soil sampling and flotation. The finds indicated that,
MAM 37: 16-grifths – Press

a case study from orkney 223

apart from some later industrial activity on top probably related to kelping, the areas
of Snusgar so far investigated reflect predominantly Norse-period construction
activity. Radiocarbon dates taken from charred grain and short-lived wood species
date the principal build-up of deposits to between ad 900 and 1050,5 a picture con-
firmed by a series of Optically Stimulated Luminescence (OSL) dates taken in 2006 on
midden deposits and the wind-blown sand layers.6 The lowest buried land surface
sampled on the north-east flank of the mound produced a cormorant bone, which
was dated as ad 489–774 at 95.4% probability.7 Hints of an earlier prehistoric core
to the mound, which were picked up by Ground Penetrating Radar (GPR) in 2005,
have yet to be tested or confirmed. It must be remembered that these large sandy
settlement-influenced formations are not single-phase or even single-process accre-
tions. One or more earlier concentrations of archaeology may lie within them, and
indeed may have assisted accretion by acting as a trap for wind-blown sand.
Another less prominent but more extensive mound inland 200m to the east of
Snusgar, now termed the ‘East Mound’, was also subjected to geophysical survey in
2004. A test trench here in 2005 was extended and deepened to a full-scale area
excavation in 2006–10. Lacking the apparently historic name and antiquarian associ-
ations of the neighbouring mound, this was initially a less-promising target. Magnet-
ometry and GPR gave only diffuse hints of archaeological structures. It became clear
in trial trenching why this was the case — the archaeology was buried under a thick
layer of wind-blown sand which has since been shown to be up to 1.5m deep over
some areas of the site. An initial discovery of coherent and (compared to Snusgar)
much less badly damaged and disturbed stone walling and midden material in 2005
led to the progressive widening and deepening of this investigation. Successive small
rectilinear stone buildings on the crest of the mound, surrounded by midden material,
were investigated in 2006–10. These turned out to be ancillary structures within a
closely grouped cluster based on a significant residence. Out with their south-west
corner, a small stone-walled yard area produced extensive ferrous metalworking
residues, which are undergoing further analysis. In 2010, the building into which the
flagged entrance led was revealed as a substantially preserved stone east–west aligned
longhouse measuring 26.3m in length and 4.9m in width (Figure 16.2). The 2010
exposure revealed an internal division between domestic and byre areas. The shape
of the exterior walls shows evidence for the classic Scandinavian ‘bow-sided’ mor-
phology. The ‘domestic’ area was identified as such by the presence of orthostatic
stone side benching facing a central hearth area. This is differentiated from the ‘byre’
area by a half-width stub wall. A flagstone-paved passageway was exposed, running
longitudinally through the centre of the ‘byre’ area of the major building, before turn-
ing abruptly south, passing through a wide and handsomely built flagged entrance
and threshold, crossing a paved yard (with a network of drains underneath the
paving), and connecting with a stone staircase leading up six steps to the upper build-
ing sequence already investigated. The passageway was kerbed by larger stones with
square sockets cut into them, evidently for supporting part of a now-vanished
wooden roof superstructure.
So thick and dense was the blanketing layer of wind-blown sand which had con-
cealed these buildings since their abandonment, that preservation is unusually good.
MAM 37: 16-grifths – Press

224 maritime societies of the viking and medieval world

Figure 16.2 ‘East Mound’ excavation, August 2010, overhead view from west. The main
longhouse building is oriented east–west. The domestic area (with wide side benches) is in the
foreground; the central hall (with narrow side benches) and byre (with a central stone
passageway) are towards the eastern end; the entrance yard, metalworking area and ancillary
structures are to the south (on the right of the excavation trench in photograph)
(# University of Oxford)

In the ‘byre’ area of the major building, as the sand was peeled back during excava-
tion, we revealed the hoof-prints of animals in the dark trampled floor surface. Some
wall masonry had collapsed under the weight of the sand, undermined in part by
rabbit burrows. Nevertheless, so deeply buried were these remains that they escaped
the extensive later stone robbing witnessed on the roughly contemporary structures
in the neighbouring mound. Radiocarbon determinations taken predominantly from
charred grain and short-lived wood date this complex of buildings to between ad 980
and 1230, suggesting it flourished somewhat later than its neighbour on Snusgar,
although there is an overlap in the late 10th/early 11th centuries.

artefact and economy at the bay of skaill

The structures outlined above are rich in finds.8 These come both from ‘aban-
donment’ layers associated with the final silting and presumably de-roofing of the
structures, and also from active occupation layers including those sealed by later
superimpositions. Surrounding the building complex are extensive midden deposits
MAM 37: 16-grifths – Press

a case study from orkney 225

which were used in conjunction with spreads of stone to stabilize exterior ground
surfaces. The finds, which almost all fall within the material culture repertoire char-
acteristic of the Viking Age and medieval period, provide a parallel chronology to
the radiocarbon dates. Amongst the earliest independently datable objects are three
beads (Hoffmann, in Griffiths et al 2009). A biconical blue glass bead (Callmer type
A176) is paralleled in Scandinavian contexts dating to ad 790–860. Even earlier is a
white opaque glass bead with crossing blue swags. Scandinavian white beads of the
period ad 800–1000 usually have red or brown swags, whereas an earlier type of
white glass bead which does have blue swags, known from Anglo-Saxon England but
commoner in the Rhineland and northern France (Guido Group 3iiia, or alternatively
Koch 34 blue), dates to the period ad 580–650. An incomplete flat concentric reddish
amber bead from an occupation layer within the domestic area of the ‘East Mound’
major building is paralleled by an example from a child’s grave in a small Viking
cemetery at Cnip, Lewis (Dunwell et al 1995, 726, fig 725, 721). It is of likely Baltic
provenance and probably dates to the 9th or 10th century.
Bracketing the date range of finds at the later end of the life of the longhouse
complex is a series of sherds of amber-brown-glazed redware pottery, some of which
has decorative orange appliqué pads. Derek Hall and the late Alan Vince have both
examined the pottery to facilitate comparison with existing European data, and
plasma spectrometry analysis (ICP) has indicated that they are a whiteware from
Normandy, a type series which is normally dated to the 12th century at the earliest.
Apart from these decorated glazed sherds, very little other pottery has been found at
either Snusgar or East Mound; four grey ware sherds show ICP readings consistent
with an Ulster origin, and a clay spindle whorl is possibly a residual Iron Age find.
The predominant material used for vessels and containers here was steatite, of which
over 50 examples have been found to date. The majority of the steatite objects have
been identified by Amanda Forster as of Shetland provenance (Forster, in Griffiths
et al 2006), although one probable Norwegian hemispherical bowl sherd was found.
Many are charred with residues; the majority come from thick sub-rectangular
vessels, whereas some have been adapted for secondary use such as line sinkers.
Small bone pins and needles, and decorated antler-comb fragments, fall into type
groups of the 10th to 12th centuries associated with Viking Age and medieval sites
elsewhere in Orkney and the western Viking world. The combs are exclusively of
iron-riveted rather than copper-riveted types, strengthening their date attribution to
the earlier phase of the Viking comb tradition (see Chapter 18). The majority were
broken and discarded and some are so fragmentary that they were retrieved in flota-
tion, although one large decorated antler comb found in 2007 was complete (measur-
ing 200635mm) and apparently carefully laid in a soft fill layer underneath a flagged
floor in the ‘East Mound’ complex of buildings. Other bone finds include a pierced
hollow cylindrical needle case (with needles still inside), a pierced polished strip, an
unfinished bone spindle whorl made from a hip ball-joint, and two small rectilinear
worked whalebone objects, one of which may be a gaming piece.
Metal finds have predominantly been small iron objects such as ties and nails,
many of which arguably formed part of the wooden superstructure. Non-ferrous metal
artefacts are rarer, but include several small copper-alloy clasps, strips and wires.
MAM 37: 16-grifths – Press

226 maritime societies of the viking and medieval world

There is a walled smithing area with four bowl hearths, ferrous residues and charcoal
which strongly suggest that metalworking was a significant part of the Norse economy
here, and five fine small whetstones (two pierced) contribute to the picture. Several
more whetstone fragments also exist. One of the complete whetstones, a tiny green/
brown banded and pierced (or ‘pendant’) example measuring only 456868mm in
length (Figure 16.3), is closely paralleled at the major Viking Age settlement complex
at Borg, Lofoten, Norway (Johansen et al 2003, 153) and Hedeby (Resi 1990). A
complete copper-alloy ringed pin (Figure 16.3), 113mm in length, of Fanning’s stirrup-
ringed, crutch-headed type was found in 2006, which by analogy with excavated
examples from stratified contexts in Dublin, may be dated to the 11th or early 12th
century (Fanning 1994, 41–46). Very few ringed pins of this type have been found in
Scotland, with just one other example from Jarlshof, Shetland (Hamilton 1956, 127);
the closely related crutch-headed stick pin type has also been found at Jarlshof and the
Udal, North Uist (Graham-Campbell 1974).
A fragment of a rotary quernstone with hand-hold of diameter 45.6cm found
within the East Mound major building in 2008 is an indicator of the importance
of cereal production in the surrounding landscape. Environmental archaeological
sampling has involved both hand collection for larger items, and flotation through
5mm, 1mm, 500mm, 300mm and 100mm meshes. Archaeobotanical evidence from
trenches on both Snusgar and East Mound (Alldritt, in Griffiths et al 2005; 2006;
2007; 2008; 2009) points towards the improvement of soil by the admixture of sea-
weed and burnt peat, and the predominance of oat and barley cultivation. Dryland
weeds such as corn spurrey and chickweed suggest that the surrounding sandy arable
land was in use for cultivation crops, perhaps not predominantly for direct human

Figure 16.3 Copper-alloy ringed pin (top) and green/brown banded ‘pendant’
whetstone (# University of Oxford, drawn by AOC Scotland)
MAM 37: 16-grifths – Press

a case study from orkney 227

consumption, but for animal fodder. Flax was also grown and retted nearby, prob-
ably in the extensive dune slack system within the Links of Skaill, from which wet-
land weed flora such as sedge, club-rushes and crowberry are also represented. The
archaeozoological material favours cattle, sheep/goat and pig (with in many cases
evidence of butchery), together with lower incidences of red deer, cat, dog, horse and
seal. The bird bone is predominantly from marine species, such as the gannet, and
the presence of a large number of fractured gannet humeri in a particular midden
layer (context 1504, radiocarbon dated to ad 940–1050 at 95.4% probability) has
led Ingrid Mainland to suggest that systematic seabird exploitation was occurring.
Exploitation of the marine environment, to be expected at such a coastal location, is
further underlined by the general occurrence in midden layers, but also in some dis-
crete specialized deposits, of marine molluscan shell dominated by limpet and peri-
winkle (Mainland, in Griffiths et al 2005; 2006; 2007; 2008; 2009). Fish remains occur
in most midden and occupation deposits but are not disproportionately abundant.
They are dominated by gadids (particularly cod, with larger size most numerous but
smaller sizes also represented; and saithe, pollack, ling, hake and torsk), together
with smaller amounts of herring, garfish, halibut, mackerel, eel, dogfish, snake
blenny, ray and salmon or trout. Layers dated to earlier in the excavated sequence,
particularly on Snusgar where their chronology begins earlier in the 10th century, are
less dominated by gadids than those of the 11th and 12th centuries (Nicholson, in
Griffiths et al 2006; 2007; 2008; 2009). This picture is consonant with studies on data
from elsewhere (eg Barrett 2005) which chart the rise of cod fishing as a North
Atlantic commodity in the 11th century. No fish-processing type middens such as
those at Freswick, Robert’s Haven, St Boniface and Quoygrew have been exposed
here, although Nicholson notes that cod head bones are common in the assemblage,
suggesting the possible preparation of a storable product such as stockfish, albeit on a
less ambitious scale than at these other sites.

consumption and materiality

The broad range of archaeobotanical, archaeozoological and artefactual evi-
dence found at the Bay of Skaill sites, but with no ‘specialized’ group, mark them
out predominantly as ‘consumer’ rather than single-purpose ‘producer’ settlements,
although there was production here in metalworking, infield cultivation and animal
husbandry, and some evidence of systematic exploitation of marine resources. The
scale of these suggest they were probably for domestic consumption and local
redistribution rather than export. The evidence compares well in broad terms with
that from Beachview (Morris 1996), although proportionately more extra-mural
midden layers have been investigated here. The strong representation of butchered
cow and pig bones suggests that some of the inhabitants at least enjoyed a meat-rich
diet in the 10th, 11th and 12th centuries, and perhaps hints at a feasting culture.
The artefacts from the clusters of Norse structures exposed here are broadly
comparable in scale and type to assemblages from other well-known Viking Age and
medieval settlements in Atlantic Scotland, such as the Brough of Birsay (Curle 1982),
Skaill, Deerness (Buteux 1997), Pool, Sanday (Hunter 2007), Quoygrew, Westray
MAM 37: 16-grifths – Press

228 maritime societies of the viking and medieval world

(Barrett 2012), Jarlshof, Shetland (Hamilton 1956) and the South Uist machair sites of
Bornais (Sharples 2005) and Cille Pheadair (Parker Pearson et al 2004a). Unlike many
of these sites, the Bay of Skaill settlements described here do not (or have not yet been
demonstrated to), overlie or re-use significant remains of the immediate pre-Norse/
Pictish period. Hence continued or re-use of objects found in situ by the Norse
settlers, such as those found by Ritchie (1977) at Buckquoy, does not appear to be a
major factor here. Objects which suggest a pre-Viking date, such as the biconical and
white swag beads, may equally well be curated objects or ‘heirlooms’ imported to the
site during the Viking period. The majority of objects found are of relatively mun-
dane raw materials which for the most part were probably sourced by local trade
within the area of the Norse earldom of Orkney (including Shetland and Caithness),
such as bone, antler and steatite, although many of the bone and antler objects are
well made and decorated with simple but recognizable motifs such as symmetrical
cross-hatching, herringbone and linked ring-and-dot patterns.
Archaeologists are becoming increasingly sensitized to the active role of material
culture as ‘agency’ in the shaping of human status and identity (eg Edgeworth 2006;
Hodder 1993; Jones 1997; Kopytoff 2006). Objects may be seen as having ‘lives’ in
that their materials, geographic origins, past and current patterns of possession,
ownership, value, use and deposition are embodied within them; they act as a tactile
and visual mnemonic, the acknowledging and understanding of which is equivalent
to a biography of their unwritten history of human/personal associations. In some
contexts, such as ‘pagan’ graves, weapons seem to have a particularly dynamic place
within an inherited and mythologized Viking past. Imported ‘exotica’ or finer items
intended for display, such as those represented here, may also be regarded as having
enhanced importance in promoting notions of inherited and aspired identity. Objects
such as the amber bead, the steatite hemispherical bowl sherd and the green/brown
banded whetstone carry Norwegian cultural associations, yet in this respect seem to
be in something of a numerical minority. The decorated pottery, amongst the latest of
the material represented, is a form of fine tableware and appears to suggest an
element of display associated with eating and drinking. Its presence as an apparently
unusual and isolated example in Norse Orkney is perhaps best seen as evidence of
indirect trade links with the north-west European continent, perhaps via Norway or
England. Other objects betray an appetite for a more eclectic and hybrid material
culture, redolent of the new links established by Norse settlers in Britain and Ireland.
Perhaps most striking are a group of objects which point towards the Irish Sea
zone as a source of cultural reference. The attribution of the crutch-headed ringed pin
to a Dublin provenance is near certain, as the city accounts for by far the predomi-
nant number of ringed pins found, coupled with the only archaeologically demon-
strable evidence for their manufacture. The grey-ware sherds referred to above also
have a likely Irish or Irish Sea provenance. Several of the combs have Irish parallels
(S. Ashby, in Griffiths forthcoming). More spectacular as a signifier of Irish Sea and
more particularly Manx links are some of the more elaborate silver ornaments within
the Skaill Hoard. This includes ten complete or semi-complete penannular brooches
together with other fragments. Of these, the majority — seven — have ‘ball-type’
terminals, with three more having stamp-decorated polygonal terminals and one with
MAM 37: 16-grifths – Press

a case study from orkney 229

flat terminals. The ball terminals of three are ingeniously decorated on their spherical
field with double-bordered ribbon beasts with oval eyes, backward-turned heads and
spiral hips, which Graham-Campbell attributed to the Mammen style of 10th-century
art (Graham-Campbell 1995, 108–127). One has a human figure struggling to escape
the embrace of a serpent, and the other ball-brooches have plain or brambled
terminals. One, which only joined the rest of the hoard in 1981, has an incised quatre-
foil design within concentric circles on the ball terminal, an almost identical piece to
one found in the 1894 Ballaquayle Hoard from the Isle of Man, the deposition of
which is dated to ad c970, making it virtually contemporary with the Skaill Hoard.
On the basis of the art styles and motifs represented on the Skaill brooches (which are
paralleled not only in silver, but on stone crosses in the Isle of Man), Graham-
Campbell made an assertive case that some, if not all, of the Skaill Hoard was of
Manx provenance (Graham-Campbell 1983, 70–71). The ring money, twisted arm
rings, kufic and Anglo-Saxon coins in the Skaill Hoard could also have been
assembled in the Irish Sea zone.
The Skaill Hoard is so vast and rich in content that it is difficult to argue with the
(apparently universally accepted) notion that this is an example of a ‘chiefly treasury’
(a description used by Barrett et al 2000, 4). Its juxtaposition with newly discovered
Viking Age and medieval settlement clusters, the nearest of which having now been
shown to have come into existence by ad 950–970, is surely no coincidence. Its date
coincides with the period of Christian conversion and the move towards establish-
ing permanent, settled lordship in the Orcadian landscape. The mounds themselves,
and the substantial stone structures within them, also arguably represent an ‘active’
statement of wealth and power in the landscape. Mounded settlement and complex
masonry are interlinked as expressions of permanence and visibility. The evidence
from the Snusgar midden layers suggests that these were not simply cast out from a
domestic core as a series of random dumps, but were carefully structured and laid
with stones, apparently deliberately raising up the height of occupation surfaces in
successive ‘managed’ events. These were further inflated by layers of wind-blown
sand. ‘Farm Mounds’, a type of settlement expression well known in Norse and medi-
eval Orkney and Norway (Bertelsen and Lamb 1993; Harrison 2013), have mostly
been discussed in functional terms as a natural outcome of settlement accumulation
in one location over several generations. Their phenomenological presence in the
landscape, as deliberately construed territorial and settlement markers, has been less
coherently addressed. The evidence from Snusgar and East Mound at the Bay of
Skaill suggests the accumulation of these mounds was not a long, unplanned incre-
mental process but a more rapid and deliberate series of constructional episodes.

status in the landscape

Snusgar and its neighbouring mounds present a dominant phalanx of past occu-
pation traces on the north side of the Bay of Skaill. Rather more extensive than the
smaller mound cluster on the south side of the bay within which Skara Brae was
discovered, they nevertheless share with these a position partly sheltered from the
open sea by the presence of higher ground to the rear, and stand in immediate
MAM 37: 16-grifths – Press

230 maritime societies of the viking and medieval world

proximity to a supply of fresh water. At the centre of the bay, facing the open sea, is
low-lying and boggy ground; a ‘tumulus’ in this area opened in or before 1772 by
Sir Joseph Banks contained three stone cists and a bead necklace, possibly of jet
(Lysaght 1974). So far characterized by ancient funerary rather than settlement
remains, this poorer land in the middle seems to have divided the bay landscape into
separate north and south foci. A further hint of early historic importance on the north
side of the Bay is the (now redundant) parish kirk of St Peter, which is located 300m
north-west of Snusgar. The present building dates from 1836, and its predecessor was
built in 1670. However, the Petrine dedication may imply a much earlier origin (Lamb
1993, 262; but cf Gibbon 2006). Churches with this dedication occur elsewhere in
Orkney in close proximity to powerful farms, including those which bear the Norse
place name Skaill (ON Skáli) which (although sometimes having humble associations
elsewhere) denotes the presence of a significant farmstead in an Orcadian context.9
Thomson (1995a, 55–57) stated:
These . . . occupy high-status coastal sites in arable districts. . . . these features are
consistent with a special type of high-status hall appearing on already-settled prime
sites. ‘Skaills’ are often found in close association with a church or a district
chapel, suggesting that they flourished at a time when church organisation was
being consolidated, perhaps in the eleventh or twelfth centuries.
The ‘Castle of Snusgar’ is a name of uncertain antiquity. Norse castles are a
well-known phenomenon in Orkney. Three are mentioned as kastali in Orkneyinga
saga (Cubbie Roo’s Castle, Wyre; Damsay, on the Bay of Firth; and Kjarrekstaðir,
probably Cairston near Stromness). These were studied, and in the case of Cairston
excavated, by J Storer Clouston (Clouston 1929; 1931; Grieve 1999). Other significant
and probably semi-fortified Late Norse aristocratic residences include Tuquoy on
Westray and ‘The Wirk’ at Westness on Rousay. All of these sites have significant
proximate relationships with churches.
Could ‘East Mound’ or Snusgar, not mentioned in Orkneyinga saga, have
originally been Skáli sites? The likelihood of an association of the [Bay of] Skáli name
with the settlement remains exposed on Snusgar or East Mound is not entirely un-
problematic. On the south side of the bay stands Skaill House, a substantial baronial
residence and seat of the Laird of Breckness. The earliest phase of the building dates
to 1628, and it was apparently a wholly new structure at that time. However, two
runic inscriptions on stone slabs, neither of which were found in their original con-
text, were discovered during construction work nearby. Moreover, a medieval ceme-
tery bordering a field known as the ‘chapel field’ next to Skaill House was excavated
in 1996 when new drainage was installed (James 1999). Twelve adult and 15 infant
burials were discovered, buried under up to 1.3m of wind-blown sand, and radio-
carbon dated to between the 11th and 14th centuries. This, perhaps more than the
status of the present house itself, hints at an early historic focus here. Nevertheless,
this in itself does not reduce the claim of Snusgar or its neighbours, as there may have
been more than one such centre around the bay.
Further evidence suggesting the north side of the bay was once at least as
important as the south side is to be found in the history of township development and
MAM 37: 16-grifths – Press

a case study from orkney 231

medieval tax (‘skat’) rentals. The earliest rentals to survive intact are the Sinclair
assessments of 1492 and 1500, compiled just after the transfer of Orkney and
Shetland from Denmark–Norway to the Scottish Crown in 1468 (Peterkin 1820).
Pennylands, which approximated to one farm, formed the basic unit of land assess-
ment, 18 of which counted as an Ounceland (‘Urisland’). Townships were sub-parish
units, fundamental to the distribution of agricultural land types between households,
which occurred with varying multiples of pennyland values. It is, however, less clear
how far back these land apportionments may be taken chronologically. The view
popular amongst an earlier generation of historians, that as a whole they represent an
unchanged inheritance from the Viking period, has been refuted in more recent years.
Thomson (2001, 218) prefers the 12th century as the period when this system crystal-
lized across Orkney. Their documented form is clearly later medieval in character,
but there are hints in the values of different townships of older land divisions being
maintained within this system. These concern those townships which reproduce
within their values clear whole or half multiples of the 18-pennyland ounceland
(Clouston 1923; Marwick 1952).
Most Orkney townships are far too small even to include one whole ounceland.
The Bay of Skaill, however, is characterized by larger values. Snusgar lies within the
township of Scarwell, which was rated in 1492/1500 as 45 pennylands or 2.5 ounce-
lands. The neighbouring township to the north-west, Northdyke, is also a 45 penny-
land/2.5 ounceland township. These may in fact represent equal divisions of an
original five-ounceland township,10 which would certainly have been the dominant
land unit bordering the bay. Southerquoy, which includes Skaill House, was rated at
24 pennylands, a less obvious multiple of the basic unit of 18. To equate multiple
values of 18 with relict Norse estates will invite scepticism. Nevertheless, not only
were Scarwell and Northdyke together equal in value to five ouncelands, but, if com-
bined, both St Peter’s Kirk and the Snusgar mound complex stand at their geographic
centre. The relatively high value of Scarwell in particular seems curious today, as,
unlike Northdyke, a large part of the south of the township bordering the bay is now
very poor-quality agricultural land. A thin veneer of grassed topsoil, which only pro-
vides rough grazing, covers deep deposits of wind-blown sand. Much of this sand
inundation is, however, most likely to be later medieval in date, and possibly associ-
ated with the end of the ‘Medieval Warm Period’ (also known as the Medieval
Climate Anomaly) and the transition into the ‘Little Ice Age’ in the North Atlantic
(Lamb 1982, 171ff; Dawson et al 2011). The skat value of Scarwell township may
therefore suggest that a richer and more productive landscape around the Bay of
Skaill, which was apparently agriculturally viable in the Viking Age and a few cen-
turies thereafter, was being covered up and lost to wind-blown sand by the 15th cen-
tury at the latest.11 Moreover, an OSL date taken from the wind-blown sand overlying
the buildings in East Mound was determined at 620t125 bp, centres on the 14th cen-
tury, confirming that by then the sand covered the abandoned settlement remains.
The abandonment of the buildings on Snusgar and East Mound points towards
a process of settlement shift around the Bay of Skaill. The principal farmsteads of
the medieval and post-medieval periods are located away from the shore frontage,
and are associated with localized spreads of augmented ‘plaggen’ soils that indicate
MAM 37: 16-grifths – Press

232 maritime societies of the viking and medieval world

the locations of infields (cf Simpson 1997). Quoyloo contains the important farm of
Stove and its satellites Midstove and Netherstove, which overlook the links from the
edge of the arable land to the north. These, arguably, are the successor settlements to
Snusgar and its neighbour. Similar patterns of retreat from the shore are evident in
Birsay and Marwick Bays, also on the west coast of Orkney’s Mainland (eg Simpson
1997; Thomson 1995b). Aristocratic settlement seems also to have developed inland
in north-west Mainland, with the important farm of Housebay by Sabiston Loch
(Crawford 2006) perhaps taking on enhanced significance as the focus of dues and
obligations. Unlike situations further east in Orkney, such as Skaill in Deerness,
late medieval successor settlements did not occur in close proximity to the Viking Age
focus at the Bay of Skaill. This may be due to the particularly intractable nature of the
wind-blown sand regime on the west Mainland, facing the full force of the North
Atlantic. It is apparently this reason why the historic importance of the coastal zone
was lost to later generations. Only the seemingly isolated parish kirk and the cluster
of grassy mounds concealing deeply buried abandoned settlements were left to denote
its former significance. In this respect it has parallels elsewhere in Atlantic Scotland,
notably in the shift away from the machair on to the blacklands in the Uists of the
Outer Hebrides (Parker Pearson et al 2004b, 161–168; see Chapter 17).

power and identity

The Bay of Skaill has been subjected to the most sustained inquiry within the remit of
the Birsay-Skaill Landscape Archaeology Project. At least two major mounded settle-
ment foci prospered near each other in the period ad 950–1200, with an overlap of
perhaps up to 100 years in the later 10th and 11th centuries. Stone buildings, middens,
metalworking and evidence of landscape and marine exploitation tell the story of the
establishment and development of a mature, landed, and very probably Christian,
Norse presence. The building forms are similar to those observed elsewhere in the
northern world, but the material culture of the settlements is far from dominated by
Scandinavian traits. Outside local intra-earldom trading dependencies, contacts with
the southern North Sea, and with Dublin and the Irish Sea, are perhaps the most
persistently visible amongst the finds. The activities in Ireland of the 11th-century
Orkney earls are well known, including most famously Sigurd the Stout who is said
to have died carrying a Raven banner at the Battle of Clontarf in 1014. Hiberno-
Norse influences may have acted in Orkney itself to distance the self-perception of the
settled population from the Viking ancestral homeland in Norway in the 11th and
12th centuries, and in so doing perhaps helped to establish a genuinely distinctive
Orcadian identity.

notes
1 Project website: http://www.conted.ox.ac.uk/ which yet bears the name of Castle of Snusgar’.
research/projects/birsay-skaill. ‘Snu’ or ‘Snoo’ refer elsewhere in Orkney to
2 The ‘Castle’ appellative is an interesting hint of nose-like headlands and promontories (eg the
former status (see below), but lacks the antiquity point of Snooshan, Birsay), possibly coupled
of a saga or medieval reference. The OSA refers with ON Garðr (enclosure) (see also Fraser
to ‘the ruins or remains of a large building, 1923).
MAM 37: 16-grifths – Press

a case study from orkney 233

3 Seaweed gathering and reduction by burning in 9 For extensive and informative discussions of
stone-lined pits to ash, for export to industrial Skaill names in relation to the topography of
centres in Scotland and England for iodine Norse settlements, see Lamb (1997) and Thom-
extraction, was common amongst the poor in son (2012).
coastal communities of Orkney from the 18th 10 I am grateful to Barbara Crawford for this sug-
century to the 1930s (Thomson 1983). gestion — she also pointed out that both Scar-
4 Two more items in private ownership were re- well and Northdyke are late names of secondary
united with the rest of the hoard as recently as character, the latter being of Scots rather than
1981. Scandinavian origin. The name of the putative
5 Accelerated Mass Spectrometry carried out by earlier 5-ounceland unit is not known, but ‘Sand-
the Scottish Universities Environmental Research wick’ itself is a possibility (see also Marwick
Centre (SUERC). 1952, 147).
6 Optically Stimulated Luminescence sampling 11 There are intriguing (if chronologically hazy)
and dating and carried out by the Research hints of wind-blown sand covering up settle-
Laboratory for Archaeology, Oxford University ments around the Bay of Skaill in Orcadian folk-
(RLAHA). lore, in particular a tale called the ‘Fin King’
7 1710t35 bp calibrated with Marine 13 using a collected by the folklorist George Marwick and
DR value of x47t52. recently retold by Tom Muir (1998).
8 Images of the finds may be viewed at http://www.
conted.ox.ac.uk/research/projects/birsay-skaill.

references
Barrett, J H, 2005 ‘Economic intensification in Viking Age and medieval Orkney, Scotland:
excavations at Quoygrew’, in A Mortensen and S V Arge (ed), Viking and Norse in the North
Atlantic: Select Papers from the Proceedings of the Fourteenth Viking Congress, Tórshavn,
19–30 July 2001, Føroya Fróðskaparfelag, Tórshavn, 264–283
Barrett, J H (ed), 2012 Being an Islander: Production and Identity at Quoygrew, Orkney,
ad 900–1600, McDonald Institute for Archaeological Research, Cambridge
Barrett, J H, Beukens, R, Simpson, I, Ashmore, P, Poaps, S and Huntley, J, 2000 ‘What was the
Viking Age and when did it happen? A view from Orkney’, Norwegian Archaeological
Review 33, 1–39
Bertelsen, R and Lamb, R G, 1993 ‘Settlement mounds in the North Atlantic’, in C E Batey,
C D Morris and J Jesch (ed), The Viking Age in Caithness, Orkney and the North Atlantic:
Select Papers from the Proceedings of the Eleventh Viking Congress, Edinburgh University
Press, Edinburgh, 544–554
Buteux, S (ed), 1997 Settlements at Skaill, Deerness, Orkney: Excavations by Peter Gelling of the
Prehistoric, Pictish, Viking and Later Periods 1963–81, British Archaeological Reports, British
Series 260, Archaeopress, Oxford
Clouston, J S, 1923 ‘The Orkney Pennylands’, Scottish Historical Review 20, 19–27
Clouston, J S, 1929 ‘Three Norse strongholds in Orkney’, Proceedings of the Orkney Antiquarian
Society 7, 57–74
Clouston, J S, 1931 Early Norse Castles, The Orcadian, Kirkwall
Crawford, B E, 2006 ‘Kongemakt og jarlemakt, stedsnavn som bevis? Betydningen av Housebay,
Harray og staðir navn på Orkenøyenes West Mainland’, Viking 69, 195–214
Curle, C L, 1982 Pictish and Norse Finds from the Brough of Birsay 1934–1974, Society of
Antiquaries of Scotland, Edinburgh
Dalland, M, 1999 ‘Sand Fiold, the excavation of an exceptional cist in Orkney’, Proceedings of the
Prehistoric Society 65, 373–413
Dawson, A, Dawson, S and Jordan, J, 2011 ‘North Atlantic climate change and Late Holocene
windstorm activity in the Outer Hebrides, UK’, in D Griffiths and P Ashmore (ed), Aeolian
MAM 37: 16-grifths – Press

234 maritime societies of the viking and medieval world

Archaeology: The Archaeology of Sand Landscapes in Scotland, Scottish Archaeological

Internet Reports 48, Society of Antiquaries of Scotland, Edinburgh, 25–36
De la Vega Leinert, A C, Keen, D H, Jones, R L, Wells, J M and Smith, D E, 2000 ‘Mid-Holocene
environmental changes in the Bay of Skaill, Mainland Orkney, Scotland: an integrated geo-
morphological, sedimentological and stratigraphical study’, Journal of Quaternary Science
15, 509–528
Dunwell, A J, Cowie T G, Bruce, M F, Neighbour, T and Rees, A R, 1995 ‘A Viking Age cemetery
at Cnip, Uig, Isle of Lewis’, Proceedings of the Society of Antiquaries of Scotland 125,
719–752
Edgeworth, M (ed), 2006 Ethnographies of Archaeological Practice, Altamira, Lanham (MD)
Fanning, T, 1994 Viking Age Ringed Pins from Dublin, Medieval Dublin Series, Royal Irish
Academy, Dublin
Fraser, J, 1923 ‘Antiquities of Sandwick Parish’, Proceedings of the Orkney Antiquarian Society
2, 24
Gibbon, S J, 2006 The origins and early development of the parochial system in the Orkney
earldom, Unpublished PhD thesis, University of the Highlands and Islands, Inverness
Grieve (Gibbon), S J, 1999 Norse Castles in Orkney, Unpublished MPhil thesis, University of
Glasgow, Glasgow
Graham-Campbell, J A, 1974 ‘A preliminary note on certain small finds of Viking-age date from
the Udal excavations, North Uist’, Scottish Archaeological Forum 6, 17–22
Graham-Campbell, J A, 1983 ‘The Viking-Age silver hoards of the Isle of Man’, in C Fell, P Foote,
J Graham-Campbell and R Thomson (ed), The Viking Age in the Isle of Man: Select Papers
from the Ninth Viking Congress, Isle of Man, 4–14 July 1981, Viking Society for Northern
Research, London, 53–80
Graham-Campbell, J A, 1995 The Viking-Age Gold and Silver Hoards of Scotland (ad 850–1100),
National Museums of Scotland, Edinburgh
Griffiths, D, 2006 ‘Birsay and Skaill, Orkney, landscape survey 2003–4’, in R E Jones and L Sharpe
(ed), Going Over Old Ground: Perspectives on Archaeological, Geophysical and Geochemical
Survey in Scotland, British Archaeological Reports, British Series 416, Archaeopress, Oxford,
213–224
Griffiths, D, 2013 ‘The Context of the 1858 Skaill Hoard’, in A Reynolds and L Webster (ed), Early
Medieval Art and Archaeology in the Northern World: Studies in Honour of James Graham-
Campbell, Brill, Leiden, 501–525
Griffiths, D (ed), forthcoming Beside the Ocean: Coastal Landscapes at the Bay of Skaill, Marwick,
and Birsay Bay, Orkney, Archaeological Investigations 2003–12, Oxbow, Oxford
Griffiths, D, Harrison, J and Hanshaw, S, 2005 Birsay-Skaill Landscape Archaeology Project, Data
Structure Report Phases I–III, 2003–5, OUDCE, Oxford
Griffiths, D, Harrison, J and Hanshaw, S, 2006 Birsay-Skaill Landscape Archaeology Project, Data
Structure Report Phases IV, 2005, OUDCE, Oxford
Griffiths, D, Harrison, J and Hanshaw, S, 2007 Birsay-Skaill Landscape Archaeology Project, Data
Structure Report Phases V–VI, 2006, OUDCE, Oxford
Griffiths, D, Harrison, J and Hanshaw, S, 2008 Birsay-Skaill Landscape Archaeology Project, Data
Structure Report Phases VII–IX, 2007, OUDCE, Oxford
Griffiths, D, Harrison, J and Hanshaw, S, 2009 Birsay-Skaill Landscape Archaeology Project, Data
Structure Report Phase X, 2008, OUDCE, Oxford
Hamilton, J R C, 1956 Excavations at Jarlshof, Shetland, HMSO, Edinburgh
Harrison, J, 2013 ‘Building mounds: longhouses, coastal mounds and cultural connections:
Norway and the Northern Isles, c ad 800–1200’, Medieval Archaeology 57, 35–60
MAM 37: 16-grifths – Press

a case study from orkney 235

Hodder, I, 1993 ‘The narrative and rhetoric of material culture sequences’, World Archaeology 25,
268–281
Hunter, J R, with Bond, J M and Smith, A N, 2007 Excavations at Pool, Sanday: A Multi-period
Settlement from Neolithic to Late Norse Times, The Orcadian in association with Historic
Scotland, Kirkwall
James, H F, 1990 ‘Excavations of a medieval cemetery at Skaill House, and a cist in the Bay of
Skaill, Sandwick, Orkney’, Proceedings of the Society of Antiquaries of Scotland 129, 753–777
Johansen, O S, Kristiansen, K and Munch, G S, 2003 ‘Soapstone artefacts and whetstones’, in
G S Munch, O S Johansen and E Roesdahl (ed), Borg in Lofoten: A Chieftain’s Farm in North
Norway, Tapir Academic Press, Trondheim, 141–158
Jones, S, 1997 The Archaeology of Ethnicity, Routledge, London
Kopytoff, I, 2006 ‘The cultural biography of things: commoditization as a process’, in A Appadurai
(ed), The Social Life of Things: Commodities in Cultural Perspective, Cambridge University
Press, Cambridge, 64–91
Lamb, H R, 1982 Climate, History and the Modern World, Methuen, London
Lamb, R, 1993 ‘Carolingian Orkney and its transformation’, in C E Batey, C D Morris and J Jesch
(ed), The Viking Age in Caithness, Orkney and the North Atlantic, Edinburgh University
Press, Edinburgh, 260–271
Lamb, R, 1997 ‘Historical background to the Norse settlement’, in S Buteux (ed), Settlements at
Skaill, Deerness, Orkney: Excavations by Peter Gelling of the Prehistoric, Pictish, Viking and
Later Periods 1963–81, British Archaeological Reports, British Series 260, Archaeopress,
Oxford, 13–16
Lysaght, A M, 1974 ‘Note on a grave excavated by Joseph Banks and George Low at Skaill in
1772’, Proceedings of the Society of Antiquaries of Scotland 104, 285–289
Marwick, H, 1952 Orkney Farm Names, W R Mackintosh, Kirkwall
Morris, C D, 1985 ‘Skaill, Sandwick, Orkney: preliminary investigations of a mound-site near
Skara Brae’, Glasgow Archaeological Journal 12, 82–92
Morris, C D, 1996 The Birsay Bay Project, Volume 2: Sites in Birsay Village and on the Brough of
Birsay, Orkney, University of Durham, Durham
Muir, T, 1998 The Mermaid Bride and Other Orkney Folk Tales, Orcadian Press, Kirkwall
Parker Pearson, M, Smith, H, Mulville, J and Brennand, M, 2004a ‘Cille Pheadair: the life and
times of a Norse period farmstead, c. 1000–1300’, in J Hines, A Lane and M Redknap (ed),
Land, Sea and Home: Settlement in the Viking Period, Maney/Society for Medieval Archae-
ology, Leeds, 235–254
Parker Pearson, M, Sharples, N and Symonds, J, 2004b South Uist: Archaeology and History of a
Hebridean Island, Tempus, Stroud
Peterkin, A, 1820 Rentals of the Ancient Earldom and Bishoprick of Orkney, with Some Other
Documents, Collected by A. Peterkin, Edinburgh
Resi, H G, 1990 Die Wetz- und Schliefsteine aus Haithabu, Wachholtz, Neumünster
Ritchie, A, 1977 ‘Excavation of Pictish and Viking-age farmsteads at Buckquoy, Orkney’, Proceed-
ings of the Society of Antiquaries of Scotland 108, 174–227
Sharples, N, 2005 Excavations at Mound 3, Bornais, South Uist, Oxbow Books, Oxford
Simpson, I A, 1997 ‘Relict properties of anthropogenic deep top soils as indicators of infield
management in Marwick, West Mainland, Orkney’, Journal of Archaeological Science 24,
365–380
Sinclair, J (ed), 1791–99 The Statistical Account of Scotland, Drawn Up from the Communications
of the Ministers of the Different Parishes, Volume 16 (Orkney), Edinburgh
Thomson, W P L, 1983 Kelp Making in Orkney, Orkney Press, Stromness
MAM 37: 16-grifths – Press

236 maritime societies of the viking and medieval world

Thomson, W P L, 1995a ‘Orkney farm names: a re-assessment of their chronology’, in B E Crawford

(ed), Scandinavian Settlement in Northern Britain, Leicester University Press, Leicester, 42–63
Thomson, W P L, 1995b ‘The landscape of medieval Birsay’, in B E Crawford (ed), Northern Isles
Connections: Studies Presented to Per Sveaas Andersen, Orkney Press, Kirkwall, 47–75
Thomson, W P L, 2001 The New History of Orkney, Mercat Press, Edinburgh
Thomson, W, 2012 ‘Orkney Skaill-names’, Northern Scotland 3, 1–15
MAM 37: 17-sharples – Press

chapter 17
THE VIKING OCCUPATION OF THE
HEBRIDES: EVIDENCE FROM THE
EXCAVATIONS AT BORNAIS,
SOUTH UIST

By Niall M Sharples, Claire Ingrem,

Peter Marshall, Jacqui Mulville,
Adrienne Powell and Kelly Reed

introduction
The date and nature of the Viking settlement of the Atlantic islands of Scotland
has been a subject of considerable debate for many decades. The process has been
characterized as a relatively peaceful event involving the assimilation of a substantial
indigenous population (Crawford 1987; Owen 2004; Ritchie 1974) or alternatively
as genocide, comprising the slaughter and expulsion of the indigenous population
during a prolonged period of raiding and the later occupation of an essentially
deserted landscape by an immigrant population (eg Jennings and Kruse 2005; Smith
2001). These views vary in popularity from decade to decade and often reflect the
context of the writer rather than any significant change in the nature of the evidence
(Barrett 2004). However, sometimes the perspective of individuals has been known to
change quite dramatically due to their personal engagement with the evidence (com-
pare Crawford 1974 and Crawford 1981).
This paper presents some new information that hopefully contributes towards an
understanding of the process that was the Norse colonization of the Outer Hebrides.
The evidence derives from a long-term fieldwork project in the island of South Uist in
the Outer Hebrides (Parker Pearson et al 2004; Sharples et al 2004). The project was
designed to examine, using archaeological evidence, the long-term settlement of the
island from its earliest occupation in the Mesolithic, through to the Clearances in
the middle of the 19th century. As part of this project two settlements, Bornais and
Cille Pheadair, were excavated to provide detailed information on the period from the
7th century ad to the 15th century ad. These excavations were remarkably successful
in providing a wealth of information on a period which had previously not been well
documented. Furthermore, as a result of the prolonged period of excavation at
MAM 37: 17-sharples – Press

238 maritime societies of the viking and medieval world

Bornais, it was possible to locate and excavate settlement remains that can be securely
dated to the 10th century and which may begin as early as the 9th century. A suf-
ficiently large quantity of material was recovered from these deposits to discuss with
confidence the economic and architectural basis of the early settlement.
The settlement at Bornais is exceptionally large and no attempt will be made to
summarize the full extent of activity either spatially or chronologically (see Sharples

Figure 17.1 The location of the settlement mounds on the Bornais machair showing the
excavated areas (N Sharples and Ian Dennis)
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 239

2005). The settlement can be divided into five separate mounds numbered 1, 2, 2A, 2B
and 3 (Figure 17.1). Mounds 2 and 2A have been substantially excavated and have
the most important evidence for the Norse colonization. Excavation of mound 1 was
largely confined to a settlement of 5th- to 6th-century date, but some early Norse
material was identified. There has also been some superficial excavation of mound 3
but this only exposed late medieval deposits (Sharples 2005).

mound 2
I will begin my discussion of the archaeological evidence from Bornais with
mound 2 as this seems to have been the centre of the Norse settlement and it has the
best evidence for the Pictish1 (pre-Viking Age) to Norse (Viking Age and early medi-
eval) transition. The archaeological evidence from this mound is dominated by the
presence of three Norse houses (Figure 17.2). The earliest definable house is dug into
truncated floors and badly preserved structural remains that contain distinctive
combs and pins of pre-Norse type (Figure 17.4, 1, 5, 6). The first two houses are bow-
walled structures of Scandinavian style, whereas the final house has straight sides and
sharp corners which indicate the development of a distinctive Hebridean form of
architecture. House 2 was a bow-sided hall just under 20m long and up to 5m wide in
the centre. It was entered from the south at the east end and was a semi-subterranean
structure with a substantial stone internal revetment wall. There was a long central

Figure 17.2 A plan of the main structural elements that define the three large houses on
mound 2 (N Sharples and Ian Dennis)
MAM 37: 17-sharples – Press

240 maritime societies of the viking and medieval world

hearth area and the structure has produced an enormous quantity of artefacts of the
highest quality (Sharples 2004).
House 1 was approximately 23.1m long, but the maximum width is difficult to
estimate. It was semi-subterranean and has evidence for a substantial timber structure
which was not observed in house 2. Detailed knowledge of this house is limited to
two areas at the east and west ends of the structure. The area in between was covered
by the later house. Various features were excavated under house 2 that help with the
interpretation of house 1, but time did not allow for the detailed recording and
excavation of this area.
The east end of house 1 comprises an area of paving on which the entrance to
house 2 was built (Figure 17.3, D). The west wall (2633) of the house 2 entrance sat on
top of the paving, but the east wall was bedded into and appeared to define the edge
of this paving and is interpreted as the gable end of house 1. The extent of the paving
suggests the end wall was perforated by an entrance, and a wall (2223) flanking
the south side of this entrance was identified along the edge of the excavated area.

Figure 17.3 The remains of house 1, mound 2: (A) the floor deposits at the west end;
(B) features indicating structural elements; (C) pits and scoops; (D) the paved floor that
defined the east end of the house (N Sharples and Ian Dennis)
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 241

The north wall of the house was unmarked at the east end, but over half-way along
the proposed line of this wall two stones were noted protruding from underneath
house 2. These stones almost certainly indicate the position of the north wall of house
1 and were deliberately left in place by the builders of house 2. The survival and
re-use of odd remnants of earlier structures is a recurrent feature of Norse house
sequences in the Hebrides.
The west end of house 1 was thoroughly excavated and revealed some important
structural elements. A compact charcoal-rich floor was identified and just inside the
area available for excavation was the end of the central hearth (Figure 17.3, A). The
removal of this floor revealed a complex palimpsest of features comprising a series of
post and stake holes (Figure 17.3, B) that cut a series of pits which ranged from
shallow scoops to deep straight-sided examples (Figure 17.3, C). The most important
post holes are the two lines that flank the floor of the house and which appear to be
major structural elements which supported the roof of the house. These posts do not
extend the full length of the house, but stop approximately 2m from the end, and this
coincides with the appearance of at least two lines of smaller post holes which cut
across the interior of the house. The area beyond this partition was where most of the
shallow pits were found but there were some pits close to the hearth that included a
deep almost vertical-sided pit (1591).
Most of the pits had largely sterile fills and it is worth noting that these would
have been almost impossible to see if the weather conditions in 2003 had not been
ideal for their recognition. The vertical-sided pit (1591) contained placed animal bone
deposits, a steatite bead (Figure 17.4, 12), made from a re-used vessel sherd, and an
iron ringed pin (Figure 17.4, 14). A concentration of steatite vessel fragments and
pottery was found amongst the floor deposits of the south-west corner of this build-
ing and the floor also produced a small lead cross (Figure 17.4, 15). An important col-
lection of iron objects was found in the overlying deposits which appear to represent
the decay of the structure and this includes a possible candle-stick holder (Figure 17.4,
9) and a coin of Eadgar (Besly, in Sharples 2003) which suggests a date late in the 10th
or early in the 11th century for the abandonment of this structure (allowing the
possibility of some decades between the date of its production and deposition).
The chronology of this mound is securely based on a series of 35 radiocarbon
dates. The dates are derived from four principal structural phases. The first five are
associated with pre-Viking Age occupation. There are five dates from house 1, seven
from house 2 and six from house 3. There are also three dates from features that
precede the building of house 1, two dates that precede the use of house 2, three dates
from secondary activity inside house 2, two dates from pits that precede the use of
house 3 and one date from secondary activity inside house 3. The stratigraphic rela-
tionships of the dates provide an ideal opportunity to undertake Bayesian modelling2
of the results (Figure 17.5) and estimates have been calculated for the dates of a
number of events in the structural sequence:
1. The start of pre-Viking (Late Iron Age) activity is estimated to lie between
ad 630–775 (95% probability; Boundary start_LIA) and probably ad 675–735
(68% probability). The end of this Late Iron Age activity is estimated to have
MAM 37: 17-sharples – Press

242 maritime societies of the viking and medieval world

Figure 17.4 A selection of finds from the early contexts: (1) a short double-sided comb from
the Pictish layers on mound 2; (2) a pair of side plates from the cultivation soil at the base of
mound 2A; (3) and (4) pig fibula pins from early Norse deposits; (5) and (6) hipped pins of Pictish
type from mound 2; (7) crook-headed stick pin of copper alloy from mound 1; (8) the ogham
inscribed plaque from mound 1; (9) an iron candle stick from the abandonment of house 1 mound
2; (10) a large whalebone implement from mound 1; (11) an inscribed phalange from mound 1;
(12) steatite bead from early deposits on mound 2; (13) a copper alloy ring from a ring-headed pin
from mound 2A; (14) an iron ring from a ring-headed pin from mound 2; (15) a lead cross from
the floor of house 1 mound 2 (N Sharples and Ian Dennis)
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 243

Figure 17.5 Probability distributions for the radiocarbon dates from Bornais mound 2. Each
distribution represents the relative probability that an event occurred at some particular time. For
each of the radiocarbon measurements two distributions have been plotted, one in outline, which
is the result of simple radiocarbon calibration, and a solid one, which is based on the
chronological model used. The other distributions correspond to aspects of the model. For
example, the distribution ‘Boundary end_LIA’ is the estimated date for the end of (pre-Viking
Age) Late Iron Age activity. The large square brackets down the left hand side along with the
OxCal keywords define the overall model exactly (P Marshall)
MAM 37: 17-sharples – Press

244 maritime societies of the viking and medieval world

taken place in ad 690–840 (95% probability; Boundary end_LIA) and probably

ad 695–750 (68% probability).
2. The earliest activity associated with house 1, the digging of a series of pits, started
in ad 790–965 (95% probability; Boundary start_house_1) and probably ad 850–
945 (68% probability). The best estimate for the start of the use of house 1 as a
structure, that is the floor being constructed, is ad 880–990 (95% probability;
house 1_use_start) and probably ad 920–980 (68% probability).
3. The end of use of house 1 and the start of activity associated with the construction
of house 2 is estimated to have taken place in ad 1025–1105 (95% probability;
Boundary end_house_1/start_house_2) and probably ad 1030–75 (68% prob-
ability).
4. The end of use of house 2 took place in ad 1040–1150 (95% probability; Boundary
end_house_2/start_infilling) and probably ad 1090–1135 (68% probability).
5. House 3 was constructed in ad 1135–1265 (95% probability; Boundary end_
infilling_start_house_3) and probably ad 1185–1255 (56% probability). House 3
went out of use in ad 1280–1365 (95% probability; Boundary end_house_3) and
probably ad 1285–1320 (68% probability).
6. The end of activity on mound 2 following the use of later subsidiary structures
within house 3 is estimated to have occurred in ad 1290–1340 (37% probability;
Boundary end_mound_2) or ad 1360–1420 (31% probability). There is a 70%
probability that activity on mound 2 does not extend into the 15th century ad.
7. It is also possible to estimate the length of time of some of these events and these
estimates are depicted in Figure 17.6. The gap between the pre-Viking Late Iron
Age activity and the earliest Norse activity is between 10–225 years (95% prob-
ability; Difference LIA-Norse) and probably 70–190 years (68% probability).
House 1 was in use for between 80–285 years (95% probability; Span house_1) and
probably one to two centuries (105–220 years; 68% probability). House 2 is esti-
mated to have been in use for between 1–95 years (95% probability; Span house_2)
and probably 1–65 years (68% probability) and house 3 for between 25–210 years
(95% probability; Span house_3) and probably 40–150 years (68% probability).

mound 1
The excavation of mound 1 focused on an abandoned and systematically dis-
mantled wheelhouse (a type of monumental dwelling characteristic of the pre-Viking

Figure 17.6 Probability distributions of the number of years during which various activities
occurred at Bornais mound 2 (note that some of the tails of these distributions have been
truncated to enable detailed examination of the highest areas of probability). The estimates are
derived from the model shown in Figure 17.5 (P Marshall)
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 245

Iron Age in parts of Atlantic Scotland) that dates from the 5th to 6th century ad.
However, the initial trial trench revealed a subterranean rectangular Norse house
(CD), infilled with midden, and overlying the Late Iron Age deposits was a scatter of
features (CE), unfortunately badly disturbed by rabbits, wind erosion and cattle
trampling. The house dates from the 12th–13th centuries ad, but contrary to expec-
tations the scatter of features produced dates (Marshall, in Sharples 2012) estimated
to have started in ad 720–990 (95% probability) and probably ad 760–890 (68%
probability). They are clearly Norse and are associated with steatite vessel fragments
and fish bone which are consistent with this date.
The features include several pits and two hearths. The best-preserved eastern
hearth was associated with a very large quantity of clinker that indicates sustained
high temperature, fusing the surrounding wind-blown sand. There is no clear indica-
tion as to why these hearths had such high temperatures and we do not find anything
like these quantities of clinker on the domestic hearths in the Norse houses (Young, in
Sharples 2005). One of the pits was clearly associated with a hearth and was filled with
this clinker. The other pits had less-distinctive fills, but one produced a very large
whalebone tool which is best interpreted as a flax scutcher (Figure 17.4, 10) and an
elaborately decorated phalange (Figure 17.4, 11) which may be Late Iron Age in date.
The excavated area had been eroding badly prior to the excavation and several
important finds were recovered from the surface. These included a small bone plaque
with an ogham inscription (Figure 17.4, 8) which Kathrine Forsyth (2007) suggests
may date to the Viking period, a Norwegian coin of Olaf Kyre (Williams and
Sharples 2003) and a crook-headed pin of copper alloy (Figure 17.4, 7).

mound 2a
The excavation of mound 2A was designed to complement the excavations of
mound 2. The structural evidence suggested mound 2A was not the site of a sub-
stantial house and would therefore enable comparisons to be made between high-
and low-status areas of the settlement. The main area of excavation lay to the north
of the principal domestic structures and it was therefore possible to expose a rela-
tively large area of early deposits and recover a considerable quantity of material
from the early Norse occupation.
The earliest deposits identified comprised natural wind-blown sand scarred by
ard marks (Figure 17.7, A). Most of these ard marks were oriented east to west, but a
later series of marks, only identifiable in a few areas, was oriented north to south and
probably indicate a separate later phase of cultivation. In the upper layers of the
plough soil a couple of simple hearths were located (Figure 17.7, B). At the level in
which they were excavated there was no sign of any structures associated with these
hearths, but it is possible that a turf and timber structure existed that was simply not
identifiable. The elongated nature of the hearth deposits which seem to lie in roughly
rectangular troughs would be appropriate for small rectangular structures similar to
those identified in the early occupation of the Udal (Selkirk 1996).
The plough soil associated with the early phase of cultivation and these hearths/
houses produced large quantities of animal bone and artefacts, and flotation of soil
MAM 37: 17-sharples – Press

246 maritime societies of the viking and medieval world

Figure 17.7 The early activity on mound 2A: (A) cultivation marks visible on the surface
of wind-blown sand; (B) hearths within the cultivation soil overlying the cultivation marks
(N Sharples and Ian Dennis)

samples has provided a substantial carbonized plant assemblage. The artefacts

include a copper alloy ring probably from a ring-headed pin (Figure 17.4, 13),
tweezers, whetstones, pig fibula pins with perforated heads (Figure 17.4, 3, 4), and a
pair of side plates (Figure 17.4, 2) from a composite comb. The comb has good
parallels in phase 4B, at Coppergate, York (McGregor et al 1999, fig 884) and was
dated to the 10th century.
This activity was succeeded by a period when there appears to have been
relatively little activity in the area excavated and a thick grey sand built up. This was
succeeded by a sequence of houses and ancillary buildings. Most of the houses lay in
the southern half of the mound and were not explored in detail. In the excavated
area, however, the structural sequence begins with a pair of furnaces linked by an arc
of wall. These were succeeded by two small ancillary buildings and finally a larger
building which may have been a house. Around the periphery of the mound was a
sequence of midden layers that were contemporary with the use of these ancillary
structures. The thickest sequence was located on the east side of the mound.
Twenty-five radiocarbon dates have been obtained on samples from mound 2A.
Again the stratigraphic sequence enables a Bayesian model for the chronology of this
mound (Figure 17.8) and estimates have been calculated for the dates of a number of
events in the structural sequence:
1. The earliest activity is dated to between ad 800–970 (95% probability; Boundary
start_ploughsoil) and probably ad 870–950 (68% probability), and is therefore
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 247

Figure 17.8 Probability distributions for the radiocarbon dates from Bornais mound 2A. Each
distribution represents the relative probability that an event occurred at some particular time. For
each of the radiocarbon measurements two distributions have been plotted, one in outline, which
is the result of simple radiocarbon calibration, and a solid one, which is based on the
chronological model used. The other distributions correspond to aspects of the model. For
example, the distribution ‘Boundary start kilns’ is the estimated date for construction of the kilns.
The large square brackets down the left hand side along with the OxCal keywords define the
overall model exactly (P Marshall)
MAM 37: 17-sharples – Press

248 maritime societies of the viking and medieval world

likely to be Norse. The majority of the material in the plough soil was probably
deposited in the 10th century ad.
2. The grey sand layers started to accumulate in ad 980–1040 (95% probability;
start_grey_sand) and probably ad 1000–20 (68% probability) and was over in
ad 1020–1120 (95% probability; Boundary end_grey_sand) and probably ad 1030–
50 (68% probability), thus representing a very short hiatus in activity.
3. The kilns and associated deposits are estimated to have first been used in ad 1040–
1150 (95% probability; Boundary start_kilns) and probably ad 1050–1130 (68%
probability). The end of use of the kilns and associated deposits took place in
ad 1070–1210 (95% probability; Boundary end_kilns) and probably ad 1110–90
(68% probability).
4. The first of the ancillary structures is estimated to have been constructed in
ad 1140–1260 (95% probability; Boundary start_ancillary) and probably ad 1160–
1230 (68% probability) and the last went out of use in ad 1290–1440 (95% prob-
ability; Boundary end_ancillary) and probably ad 1300–70 (68% probability).
5. The midden started to accumulate in ad 1150–1250 (95% probability; Boundary
start_midden) and probably ad 1170–1220 (68% probability) and finished in
ad 1190–1290 (95% probability; Boundary end_midden) and probably ad 1210–
60 (68% probability).
6. The probability that activity on mound 2A finished before ad 1500 is 99%.
7. The modelling suggests that the material deposited in the plough soil on mound 2A
and the accumulation of the grey sand are contemporary with the use of house 1
on mound 2. It is probable that house 2 on mound 2 was contemporary with the
kilns, though house 2 started a little earlier and the kilns continued in use after
house 2 had gone out of use. The start and use of the ancillary buildings on mound
2A is probably contemporary with the start and use of house 3 on mound 2. How-
ever, the ancillary buildings continued in use after house 3 had gone out of use.

cille pheadair
The excavations at Cille Pheadair, South Uist (Parker Pearson et al 2004)
revealed a single small settlement mound that comprises a sequence of structures
and middens very similar to that exposed on the smaller of the mounds at Bornais.
Modelling of the radiocarbon dates from this settlement indicates activity from the
late 9th and 10th centuries ad, contemporary with the earliest activity from Bornais,
but these are residual dates and are not associated with the earliest structural
evidence which appears to begin in the middle of the 11th century ad. Occupation is
estimated to have ended in the late 12th or early 13th century ad (ad 1170–1220;
68% probability) and probably a century or more before the abandonment of the
settlement at Bornais.

sequence and chronology

The chronological evidence indicates that the (pre-Viking Age) Late Iron Age
occupation of Bornais came to an end in the first half of the 8th century ad
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 249

(ad 695–750 (68% probability)). There then appears to have been a hiatus in activity
until the mid-9th to mid-10th centuries ad, when house 1 on mound 2 was con-
structed (ad 850–945; 68% probability) and a plough soil developed on mound 2A
(ad 870–950; 68% probability). The plough soil on mound 2A was associated with
hearths that may represent more insubstantial structural activity dispersed around
the main house on mound 2; similar early hearths were also found on mound 1.
Clearly, these observations must be treated with caution as we have only exca-
vated a small proportion of the site. The dates from the pre-Viking Age settlement
came from an area that had been truncated by the construction of the Viking house.
The construction of this house may have removed evidence for a final 9th-century
phases of the Late Iron Age settlement. There may also be additional large houses,
constructed in the 10th century, in unexcavated parts of the settlement.
Nevertheless, the evidence does suggest that the recorded Viking incursions of
the late 9th century did have an impact on the settlements of the Uist machair and
may even have led to their abandonment. There is a possible hiatus until the mid-9th
to mid-10th century, when a large Viking longhouse was constructed. Contemporary
with its construction were a number of smaller insubstantial structures. The latter
structures may indicate the settlement was the focus for large temporary gatherings of
people, or that it was associated with a large, but subsidiary, population of depen-
dants. The latter interpretation would fit the evidence from the later periods when the
large houses on mound 2 were surrounded by subsidiary farmhouses.
The only other area of the North Atlantic where a very large suite of scientific
dates has been acquired from early Norse settlement contexts is from Myvatnssveit in
Iceland (McGovern et al 2007; Vésteinsson and McGovern 2012). Here a substantial
landscape project has resulted in the dating of several settlements and produces a
chronological pattern very similar to that identified on South Uist. For example, the
earliest settlement dates from Hofstaðir and Hrı́sheimar indicate construction of
these buildings at the end of the 9th or the beginning of the 10th centuries ad.
Myvatnssveit is in the north of Iceland, well inland from the coast, and the estab-
lishment of substantial settlements in this region, at this time, indicates the colon-
ization of Iceland occurred very rapidly (Vésteinsson and McGovern 2012).

animal bones
Excavations of mounds 2 and 2A have produced a very large assemblage of
bone. Analysis of the large mammal bones is now complete, and analysis of the fish
bones is in progress, but we have yet to complete the analysis of the bird and small
mammal bones. The quantity of bone recovered from the different phases is
substantial and even when subdivided into successive chronological phases enough
bones are recovered to provide meaningful results. The only period from which it was
not possible to recover a large assemblage was the Pictish component of mound 2 as
only a small sample of the relevant deposits was excavated. The early Norse phase in
contrast is represented by an assemblage of over 3500 bones from the plough soil on
mound 2A and an assemblage of roughly 500 bones from house 1 on mound 2.3 The
11th century is represented by an assemblage of roughly 1400 bones from house 2
MAM 37: 17-sharples – Press

250 maritime societies of the viking and medieval world

and the 13th to 14th century is represented on both mounds, but particularly by the
middens on mound 2A.
The chronological distribution of the main domestic species is summarized in
Figure 17.9. This shows the relative proportions of pig, cattle and sheep bones. The

Figure 17.9 The variation in the principal domestic animal species on

mound 2 and 2A (N Sharples, J Mulville and A Powell)
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 251

most striking feature of the distribution is the presence of a significant quantity of pig
in the early Norse phases. In the plough soil on mound 2A, pig comprise over 20% of
the assemblage and in the floor of house 1 they comprise 15% of the assemblage. This
contrasts markedly with the assemblage from the Pictish deposits on mound 2, where
pig were less than 10%, and also with the Late Iron Age assemblage from mound 1.
On both mounds the proportion of pigs declines steadily to between 6% and 8% of
the assemblage in the 13th and 14th centuries, a figure comparable to the pre-Norse
occupation. The decline coincides with an increase in the significance of sheep which
form 58–59% of the assemblage from the final phase of activity. Cattle were the most
important animal in the Pictish layers on mound 2, but seem to have much less
importance in the early Norse period. In the 11th century their significance increased,
but declined again in the final phases of occupation.
The large numbers of bones recovered from the site has enabled an analysis of
the age profile of the cattle and sheep assemblage from three different phases
(Figure 17.10).4 The sheep mortality pattern is very similar across all three periods
though there is a noticeable peak in sheep of wear stage D (12–24 months) in the
10th century. In contrast, the cattle mortality patterns show significant differences for
the three periods. In the later two periods there were a large number of deaths in the
first year. After 18 months the number of deaths drops dramatically before there is a
slight rise in adult deaths. This pattern can be interpreted as likely to represent an
assemblage of animals kept for milk production. The early Norse assemblage in
contrast shows no evidence for large numbers of early deaths and there is a significant
assemblage of young adults that would suggest an interest in meat production.
Recent work on animal bones from the Faroe Islands (Church et al 2005) and
Iceland (McGovern et al 2000, 2001, 2007) have indicated the importance of pigs in
the early settlements. This species may have been an environmental adaptation to
settlement conditions, being ‘particularly well suited to feeding on the root systems of
artic birch and willow’ (Vésteinsson et al 2002). Although relevant to Iceland, this
interpretation does not suit the Faroe Islands which lacked a forested environment
prior to Norse colonization and the Hebrides were also completely deforested at this
time. Indeed the presence of pigs on the fragile machair sites of the west coast of
South Uist is very dangerous as their digging proclivities could break up the protec-
tive turf cover and encourage wind-blown deflation. The introduction of pigs is more
likely to indicate economic and social pressures. Pigs are an easy way to produce
meat. They require little in the way of complex management and can be fed on a
variety of waste materials. They have routinely been used as a resource for feasting
throughout prehistory and are well suited to a society where communal and con-
spicuous consumption is an important social strategy for achieving status.

fish
Large quantities of fish bone have now been recovered and identified from the
pre-Viking Age (Late Iron Age) and Norse deposits on mound 1 (1020 identified
bones) and mound 2 (10,317 identified bones), and from the 13th- to 14th-century
farm on mound 3 (3073 identified bones).5 Fish bones are present in much lower
MAM 37: 17-sharples – Press

252 maritime societies of the viking and medieval world

Figure 17.10 The age distribution of the cattle and sheep

(N Sharples, J Mulville and A Powell)

frequencies in the pre-Norse deposits and herring are notably absent. The only
deposits rich in fish bones from 5th-century middens on mound 1 were dominated by
salmonid and saithe, indicating an inshore fishing strategy very different to that of the
Norse settlement.
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 253

The Norse settlement produced much larger densities of fish bone and the
assemblage is dominated by herring which were probably caught by net fishing from
boats. The importance of fishing and herring is identifiable in the earliest Norse
deposits (house 1 floors) and represents a dramatic change from the pre-Norse econ-
omy. The densities of fish remains present in a litre of soil vary considerably from
context to context, but on mound 2 higher densities are much more common in later
contexts (house 3 floors).
The assemblages of herring bones from all the house floor layers are dominated
by vertebrae, especially those from the abdominal region, suggesting that the
majority of herring arrived in a decapitated form. The smaller assemblage of saithe
bones in contrast includes both vertebrae and cranial bones and suggests these fish
arrived whole. The presence of processed fish indicates the settlement at Bornais is a
consumption rather than a processing site, but the very large numbers of herring
present do suggest that a surplus was available that could be traded.
Herring is an oily fish and long-distance transportation as a commodity histo-
rically required the fish to be pickled in brine or smoked (Cutting 1956). This con-
trasts with the white fish that dominate the assemblages from Norse settlements in
the Northern Isles of Scotland which can be preserved by salting and/or air drying.
The fact that the herring from Bornais were decapitated may indicate that the fish
were air-dried in a manner similar to the white fish of the Northern Isles, as similar
practices have been recorded in Iceland (Cutting 1956, 53). However, it is still difficult
to understand how these fish would be preserved for long-distance trade.

crop remains
The machair plain of the island of South Uist is well suited to the production of
cereals, and even today the machair is ploughed and barley and oats are cultivated,
though primarily for animal rather than human consumption. Historically the Uists
were among the most productive islands, comparable only to Islay and Tiree on the
western seaboard of Scotland – a landscape otherwise dominated by rock and bog.
The historical record is supported by our programme of extensive flotation. We have
so far processed 9580 litres of sediment for carbonized remains. Barley and oats were
present in almost all the samples from Norse contexts and are much more prevalent
than in the (pre-Viking) Late Iron Age samples.
Figure 17.11 illustrates the density of barley, oat, flax and rye grains recovered
from a litre of sediment and how this changed through time on mounds 2 and 2A.
The barley densities rise significantly between the Late Iron Age and Viking phases
on both mounds, particularly on mound 2A. The densities then drop to a level that is
still higher than the pre-Viking Age levels. Oat densities are negligible in the Late Iron
Age on both mounds, but become high in the Viking period particularly on mound
2A. The mound 2A densities are always higher than those on mound 2 and this
probably reflects a difference between the role of the middens and ancillary structures
sampled on mound 2A and the house floors sampled on mound 2. The very high
densities in the 11th century relate to a thick layer of burnt oats, which must represent
the destruction of a harvest.
MAM 37: 17-sharples – Press

254 maritime societies of the viking and medieval world

Figure 17.11 The density of barley oats, rye and flax remains through time
(N Sharples and K Reed)

The densities of flax and rye are much lower than the densities for barley and
oats (the y-axis for these graphs on Figure 17.11 is at a different scale). Flax was
present before the Viking Age, but levels were low and remained low in the early
Norse phases. Rye in contrast appears to have been introduced to the region as a crop
during the Norse colonization. Both crops became increasingly important in the later
Norse periods. The house floors of mound 2 produced higher densities of flax, but rye
was more important on mound 2A.
These patterns will require further analysis, however, they suggest the Norse
colonization was characterized by an intensification of arable agricultural. This
initially was focused on the established crops of barley and oats, though new crops
(rye and flax) were also present. There then appears to have been a gradual decline in
the importance of barley and oats that coincided with increasingly diverse use of
crops such as rye and flax.

conclusions
The evidence presented here is just a small part of the material recovered from
the settlement at Bornais. Nevertheless, it is directly relevant to how we perceive
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 255

the Viking colonization of the islands. In 1999 Sharples co-authored an article in

Norwegian Archaeological Review that argued that ‘it was logical to assume that the
bulk of the population survived the coming of the Vikings and adapted their way of
life to the new social and political regime’ (Sharples and Parker Pearson 1999, 58).
The argument was based on:
1. a continuity of settlement and what was perceived as the complexity of farming
practice on the fragile machair soils
2. the need for a knowledgeable farming labour force
3. the distinctive architectural features of the houses, notably ‘a subterranean con-
struction and single-skinned’ revetment wall
4. the creation of a ceramic industry which though different from pre-Viking Age
(Late Iron Age) pottery, seems to originate in the rich ceramic traditions of the
region and was an innovation for a Viking community.
Since this paper was published, Jennings and Kruse (2005) have thoroughly
reviewed the evidence and come to quite different conclusions. They believe that the
re-occupation of long-lived settlements on the machair plain is only to be expected
from incoming Norse colonizers and that subterranean house construction was both
present in the Norse building tradition and represents a sensible response to building
in the exposed machair environments. They go on to argue that the pottery evidence
represents a significant break with existing technology and the introduction of potters
from Northern Ireland (Jennings and Kruse 2005, 289). The presence of a dependent
(enslaved?) Irish Gaelic population may also explain the early presence of Gaelic
place names in Lewis (Cox 2002) and the mixed DNA results which are a charac-
teristic of the Western Isles (Goodacre et al 2005).
The new economic evidence from Bornais certainly supports these arguments.
There does seem to be a chronological disruption in the settlement sequence and it
seems we underestimated the capacities of Norse farmers and the ability of the
machair landscape to be more extensively exploited. The animal bones and crop
remains suggest the agricultural exploitation of the landscape intensified considerably
during the 10th century. The introduction of pigs and new cereal types was ultimately
not sustainable, but in the short term these innovations enabled the development of
settlements that had exchange relationships that extended across the North Atlantic
(Sharples and Smith 2009) and which may reflect the export of an agricultural surplus.
It was not until the climatic downturn of the late Middle Ages that the long-term
consequences of the over-exploitation of the machair became apparent. The machair
plain became unstable and was abandoned in the 15th century (cf Chapter 16), and the
islands entered a long period of isolation and social unrest.
Whilst we accept the basic premise of the argument presented by Jennings and
Kruse, there are elements of their interpretation that are debatable. The islands of the
Outer Hebrides are large enough to have maintained a population during the period
of raiding. Their comparison with Noirmoutier, a small island in the mouth of the
Loire, may be relevant to the situation on Iona, but not the larger islands of Lewis
and the Uists. The subterranean stone-lined houses are distinctive and represent a
significant change in normal Norse practice that may indicate the influence of a
MAM 37: 17-sharples – Press

256 maritime societies of the viking and medieval world

remnant local population. Similarly, the origins of the ceramics require further study
and the relationship with Northern Ireland is not as secure as Jennings and Kruse
argue (Lane 2010). It remains possible that a small, but nevertheless significant,
population of the indigenous inhabitants of the Hebrides survived to influence the
Norse settlers.6

notes
1 I will refer to the immediately pre-Norse deposits include the Late Iron Age material. The three
on mound 2 as Pictish and this should be taken periods consist of the bones from: (a) BB-, house 1
as a shorthand label for the activity belonging to foundation pits, occupation and abandonment on
the 7th or 8th centuries ad. It is distinguished mound 2 and GA- contexts from the plough soil
from the Late Iron Age occupation at mound 1, and grey sand on mound 2A; (b) BC- and BD,
which dates to the 5th to 6th centuries ad. The house 2 and post-abandonment structures on
applicability of this cultural term (based on histo- mound 2 and GB the kilns and their associated
rical kingdoms of eastern Scotland) to the Outer middens on mound 2A; and (c) BE- and DF,
Hebrides remains open to debate, but its use is house 3 and house 3 abandonment deposits on
consistent with archaeological evidence such as mound 2 and GC-, GD-, GE- GF- which consist
the distribution of square cairn burials. of all the ancillary buildings and their associated
2 The ranges quoted in italics are posterior density middens on mound 2A
estimates derived from mathematical modelling 5 Work on the large assemblage from mound 2A is
of archaeological problems (Buck et al 1996; in progress, but it would appear to confirm the
Bayliss et al 2007). The technique used is a form general patterns identified in the assemblage from
of Markov Chain Monte Carlo sampling, and has mound 2.
been applied using the program OxCal v4.0.5 6 This article could not have been written without
(http://c14.arch.ox.ac.uk/) using the calibration the contribution of a large number of people,
curve of Reimer et al (2004). Details of the algo- particularly the residents of South Uist who pro-
rithms employed by this program are available vided invaluable support for the excavations. The
from the on-line manual or in Bronk Ramsey excavation and post-excavation processes have
(1995; 1998; 2001). been generously supported by Historic Scotland
3 These values differ from the totals in Figure 17.7 and the University of Cardiff. We would like to
as the latter are restricted to the three main thank all the students and specialists who have
domesticates. worked on material from the excavations and in
4 The figures from both mound 2 and 2A have been particular: E Besly, S Colledge, A Forster, K For-
amalgamated and the phasing simplified to obtain syth, F Morris, C Riley, A Rowe, H Smith,
sufficient quantities of material that can be aged. R Smith and F Taylor.
Unfortunately the quantities were too low to

references
Barrett, J H, 2004 ‘Beyond war or peace: the study of culture contact in Viking-age Scotland’, in
J Hines, M Redknap and A Lane (ed), Land, Sea and Home, Maney/Society for Medieval
Archaeology, Leeds, 207–218
Bayliss, A, Bronk Ramsey, C, van der Plicht, J and Whittle, A, 2007 ‘Bradshaw and Bayes: towards
a timetable for the Neolithic’, Cambridge Archaeological Journal 17, 1–28
Bronk Ramsey, C, 1995 ‘Radiocarbon calibration and analysis of stratigraphy: the OxCal pro-
gram’, Radiocarbon 37, 425–430
Bronk Ramsey, C, 1998 ‘Probability and dating’, Radiocarbon 40, 461–474
Bronk Ramsey, C, 2001 ‘Development of the radiocarbon program OxCal’, Radiocarbon 43,
355–363
Buck, C E, Cavanagh, W G and Litton, C D, 1996 Bayesian Approach to Interpreting Archaeo-
logical Data, Wiley, Chichester
Church, M, Aarge, S, Brewington, S, McGovern, T H, Woollett, J M, Perdikaris, S, Lawson, I T,
Cook, G T, Amundsen, C, Harrison, R, Krivogorskya, Y and Dunbar, E, 2005 ‘Puffins, pigs,
MAM 37: 17-sharples – Press

the viking occupation of the hebrides 257

cod and barley: palaeoeconomy at Undir Junkarinsfløtti, Sandoy, Faroe Islands’, Environ-
mental Archaeology 10, 179–197
Cox, R A V, 2002 The Gaelic Place-names of Carloway, Isle of Lewis: Their Structure and
Significance, School of Celtic Studies, Dublin
Crawford, B E, 1987 Scandinavian Scotland, Leicester University Press, Leicester
Crawford, I, 1974 ‘Norseman and Gael’, Scottish Archaeological Forum 6, 1–15
Crawford, I A, 1981 ‘War or peace — Viking colonisation in the Northern or Western Isles of
Scotland reviewed’, in H Bekker-Nielson, P Foote and O Olsen (ed), Proceedings of the Eighth
Viking Congress, Aarhus, 1977, Odense University Press, Odense, 259–269
Cutting, C L, 1956 Fish Saving, Philosophical Library, New York
Forsyth, K, 2007 ‘An ogham-inscribed plaque from Bornais, South Uist’, in B B Smith, S Taylor and
G Williams (ed), West Over the Sea: Studies in Scandinavian Sea-borne Expansion and
Settlement before 1300, Brill, Leiden, 461–477
Goodacre, S, Helgason, A, Nicholson, J, Southam, L, Ferguson, L, Hickey, E, Vega, E, Stefáns-
son, K, Ward, R and Sykes, B, 2005 ‘Genetic evidence for a family-based Scandinavian settle-
ment of Shetland and Orkney during the Viking periods’, Heredity 95, 129–135
Jennings, A and Kruse, A, 2005 ‘An ethnic enigma — Norse, Pict and Gael in the Western Isles’, in
A Mortensen and S V Arge (ed), Viking and Norse in the North Atlantic: Select Papers from
the Proceedings of the Fourteenth Viking Congress, Tórshavn, 19–30 July 2001, Føroya
Fróðskaparfelag, Tórshavn, 284–296
Lane, A, 2010 ‘Viking-age and Norse pottery in the Hebrides’, in J Sheehan and D O Corrain (ed),
The Viking Age: Ireland and the West, Four Courts, Dublin, 204–216
McGovern, T H and Ogilvie, A E J, 2000 ‘Sagas and science: climate and human impacts in the
North Atlantic’, in W H Fitzhugh and E I Ward (ed), Vikings: The North Atlantic Saga,
Smithsonian Institution Press, Washington, 385–393
McGovern, T H, Perdikaris, S and Tinsley, C, 2001 ‘The economy of landnám: the evidence of
zooarchaeology’, in A Wawn and Þ Sigurðardóttir (ed), Approaches to Vinland, Sigurður
Nordal Instutite, Reykjavik, 154–165
McGovern, T H, Vésteinsson, O, Fridriksson, A, Church, M, Lawson, I, Simpson, I A, Einarsson,
A, Dugmore, A, Cook, G, Perdikaris, S, Edwards, K J, Thomson, A M, Adderley, W P,
Newton, A, Lucas, G, Edvardsson, R, Aldred, O and Dunbar, E, 2007 ‘Landscapes of settle-
ment in northern Iceland: historical ecology of human impact and climate fluctuation on the
millennial scale’, American Anthropologist 109, 27–51
MacGregor, A, Mainman, A J and Rogers, N S H, 1999 Craft, Industry and Everyday Life: Bone,
Antler, Ivory and Horn from Anglo-Scandinavian and Medieval York, Council for British
Archaeology/York Archaeological Trust, York
Owen, O, 2004 ‘Where are the Vikings? The missing decades of the Viking Age in the Northern
Isles’, in J Adams and K Holman (ed), Scandinavia and Europe 800–1350: Contact, Conflict
and Co-existence, Brepols, Turnhout
Parker Pearson, M, Sharples, N M and Symonds, J 2004 South Uist: Archaeology and History of a
Hebridean Island, Tempus, Stroud
Parker Pearson, M, Smith, H, Mulville, J and Brennand, M, 2004 ‘Cille Pheadair: the life and times
of a Norse-period farmstead c. 1000–1300’, in J Hines, M Redknap and A Lane (ed), Land,
Sea and Home, Maney/Society for Medieval Archaeology, Leeds, 236–254
Reimer, P J, Baillie, M G L, Bard, E, Bayliss, A, Beck, J W, Bertrand, C J H, Blackwell, P G, Buck,
C E, Burr, G S, Cutler, K B, Damon, P E, Edwards, R L, Fairbanks, R G, Friedrich, M,
Guilderson, T P, Hogg, A G, Hughen, K A, Kromer, B, McCormac, G, Manning, S, Bronk
Ramsey, C, Reimer, R W, Remmele, S, Southon, J R, Stuiver, M, Talamo, S, Taylor, F W, van
MAM 37: 17-sharples – Press

258 maritime societies of the viking and medieval world

der Plicht, J and Weyhenmeyer, C E, 2004 ‘IntCal04 Terrestrial radiocarbon age calibration,
0–26 cal kyr BP’, Radiocarbon 46, 1029–58
Ritchie, A, 1974 ‘Pict and Norseman in northern Scotland’, Scottish Archaeological Forum 6,
23–36
Selkirk, A, 1996 ‘The Udal’, Current Archaeology 147, 84–94
Sharples, N M, 2003 The Iron Age and Norse Settlement at Bornish, South Uist: An Interim Report
on the 2003 Excavations, Cardiff Studies in Archaeology Specialist Report 26, SHARE,
Cardiff University, Cardiff
Sharples, N M, 2004 ‘A find of Ringerike art from Bornais in the Outer Hebrides’, in J Hines,
M Redknap and A Lane (ed), Land, Sea and Home, Maney/Society for Medieval Archaeology,
Leeds, 255–272
Sharples, N M, 2005 A Norse Farmstead in the Hebrides: Excavations at Bornais Volume 1,
Oxbow Books, Oxford
Sharples, N M, 2012 A Late Iron Age Farmstead in the Hebrides: Excavations at Bornais Volume 2,
Oxbow Books, Oxford
Sharples, N M and Parker Pearson, M, 1999 ‘Norse settlement on the Outer Hebrides’, Norwegian
Archaeological Review 32, 41–62
Sharples, N M and Smith, R, 2009 ‘Norse settlement in the Western Isles’, in A Woolf (ed), Scan-
dinavian Scotland — Twenty Years After, University of St Andrews, St Andrews, 103–130
Sharples, N M, Parker Pearson, M and Symonds, J, 2004 ‘The archaeological landscape of South
Uist’, in R A Housley and G Coles (ed), Atlantic Connections and Adaptations: Economics,
Environments and Subsistence in the North Atlantic, Oxbow Books, Oxford, 28–47
Smith, B, 2001 ‘The Picts and the Martyrs or did the Vikings kill the native population of Orkney
and Shetland?’, Northern Studies 36, 7–32
Vésteinsson, O, and McGovern, T H, 2012 ‘The Peopling of Iceland’, Norwegian Archaeological
Review 45, 206–218
Vésteinsson, O, McGovern, T H and Keiller, C, 2002 ‘Enduring impacts: social and environmental
aspects of Viking Age settlement in Iceland and Greenland’, Archaeologia Islandica 2, 98–136
Williams, G and Sharples, N M, 2003 ‘Et nytt myntfunn fra Olav Kyrre fra Hebridene’, Nordisk
Numismatisk Unions Medlemsblad 3–4, 55–56
MAM 37: 18-ashby – Press

chapter 18
DISENTANGLING TRADE: COMBS IN
THE NORTH AND IRISH SEAS IN THE
LONG VIKING AGE

By Steven P Ashby

introduction
For Britain and Ireland in the centuries either side of ad 1000, bone and antler
hair combs are among the most frequently recovered artefacts. They were clearly
socially meaningful components of the canon of portable material culture (eg Callmer
1998, 479), and offer great potential for the high-resolution investigation of pattern-
ing in time and space (see Ashby 2006a). They therefore represent an appropriate
medium for the study of contact and exchange in Britain between ad 900 and 1400.
This paper relates to a wider research project on bone and antler hair combs
from Viking Age and medieval northern Britain (Ashby 2006a). This project con-
sidered over 2300 combs, fragments and related material from Viking Age and medi-
eval northern England, Scotland and Scandinavia. The present paper focuses on
combs from the Northern and Western Isles of Scotland, though as is fitting for the
theme of the volume they will be considered in the context of Great Britain, Ireland,
Scandinavia and northern Europe at large. Chronologically, the study is framed
within the latter centuries of what we may term the ‘long Viking Age’ of Atlantic
Scotland, a period between the 10th and 15th centuries, during which maritime con-
tact and trade were highly significant. The aim of this paper is to show how combs
may be used to inform our understanding of social and economic interaction between
the lands bordering the North and Irish Seas.

the nature of comb production

Following Kristina Ambrosiani’s (1981) seminal work, it is commonly assumed
that European Viking Age combs were uniform in design and ornament, and manu-
factured by a large, pan-continental network of itinerant craftsmen. This idea is
intuitively attractive, and it is not difficult to see why it has been so widely adopted
(eg Hansen 2005, 184; Nicholson 1997, 495; see also Christophersen 1980 for a
parallel argument). However, Ambrosiani’s model is helpful only in the broadest
MAM 37: 18-ashby – Press

260 maritime societies of the viking and medieval world

sense, and its application within Britain and Ireland is problematic, given a number of
ambiguities. In particular, Ambrosiani includes no detail regarding the scale on which
her proposed itinerancy may have acted, and is equally silent on the means by which
Britain and Ireland were tied into the Scandinavian and continental network, if at all.
Her model may work for the understanding of connections between Birka, Hedeby
and Ribe, but it should not be imported uncritically to the study of Britain and
Ireland.
Moreover, one might question the logic of such a scenario; travelling over long
distances (whether by land, sea or river) makes little ergonomic sense and, if such a
mode of commerce were to be profitable, it imposes an exceedingly high value on an
individual comb. Equally, the idea of comb makers travelling to market without the
requisite materials, in the hope that they could make the necessary acquisitions from
locals, customers or fellow merchants (as Ambrosiani suggests) seems inherently un-
likely. Such a model introduces an unnecessary element of complexity, not to say risk,
into the profession. Furthermore, any scenario that sees comb makers travelling to
market with nothing but their toolkits presupposes that, on arrival at market, they
would be happy to spend time first acquiring materials, and then manufacturing the
combs themselves. Given that the manufacture of a single composite comb — from
production of blanks, through assembly, to finishing and ornament — can take up to
a day to complete (see Galloway and Newcomer 1981 for a discussion of the pro-
duction process), this does not seem to represent an efficient use of time. Finally, in
addition to these economic, ergonomic and logistical objections to Ambrosiani’s
thesis, it may be suggested that the model also presupposes an unsubstantiated level
of mercantile freedom. We should take care before we invoke the existence of such
unfettered commerce in the 10th and 11th centuries, particularly when comb making
in previous centuries seems to have been tied to the estates of elite magnates and kings
(cf Cnotliwy 1973, 320).
Furthermore, recent research (Ashby 2006a; 2012) has cast doubt on Ambrosi-
ani’s perception of a uniform European corpus of combs, and has demonstrated the
existence of local fashions and regional manufacturing traditions, such that any
model of comb manufacture that sees craftspeople travelling on inter-regional scales
is now problematic. Regional variability is probably the key (see Ashby 2012), but the
situation must also have been chronologically variable; while Ambrosiani’s study
focused specifically on 9th- and 10th-century Scandinavia, her ‘itinerancy’ model has
become the default position of any study of comb making in the centuries either side
of the end of the 1st millennium (eg Carlsson 2002a, 2; Flodin 1989; Nicholson 1997,
495; though see Rogers 1993, 1958; Mainman and Rogers 1999, 1919–22; note the
lack of recent detailed treatments of the issue).

combs in scotland during the long viking age

Notwithstanding such ambiguity, comb making in Viking Age and medieval
England has been well studied (see Ashby 2006b; 2012; MacGregor 1985; MacGregor
et al 1999; Riddler 1990; 1991; 1998), at least relative to the Scottish context, where a
paucity of evidence for local manufacture has meant that the issue of comb
MAM 37: 18-ashby – Press

disentangling trade 261

production remains largely unexplored. In spite of these difficulties, some progress

may be made through consideration of the combs themselves, paying particular
attention to patterning in time and space (Ashby 2009; cf Clarke and Heald 2002). As
one of the few forms of portable material culture well represented in England,
Scotland, Ireland and Scandinavia, combs are particularly well suited as markers
for the shifting currents of north-western European society in the late 1st and early
2nd millennia. In the following, we will see how spatial patterning in comb style
illuminates the nature of inter-regional relationships between the 10th and 14th cen-
turies. The resultant model invokes the existence of two contemporaneous comb
networks, which, though overlapping to some extent, were largely buffered by the
western Baltic Sea. This is interesting, as the Baltic is characteristically viewed as a
medium of contact rather than one of separation, and leads to considerations regard-
ing the reasons for the existence of two comb ‘worlds’.
Before we begin to consider the distributions of comb types and manufacturing
traditions, it is necessary to briefly review the diversity of forms common in the
period of interest. In order to facilitate this, a new typology was created by the
author. This was developed using a combination of site-based stratigraphic analysis
and statistical techniques such as correspondence analysis (see Ashby 2006a for a
detailed treatment of the means by which the typology was devised, and Ashby 2007;
2011 for overviews). This typology was designed with the aim of being applicable
across Scandinavian Britain, Ireland and northern Europe. For the purposes of
economy and parsimony, the focus herein is restricted to those types immediately
relevant to the later and ‘long’ Viking Age (see Figure 18.1).1 First, however, we need
some context, and we must therefore consider the combs of early Viking Age Atlantic
Scotland.
When the first Scandinavians began to settle in Atlantic Scotland, the combs in
popular use were types 11, 12 and 1c (not discussed herein; see Ashby 2009), which
are indicative of contact with both Anglo-Saxon England and Ireland. It was into this
complex milieu that the first Scandinavian settlers appeared in the 9th century,
bringing with them their distinctive ‘Norse’ combs (type 5), and one may only spec-
ulate upon the various ways in which this diverse repertoire of comb forms was used
in mediating culture contact.2 Contrary to the traditional assumptions of formal
homogeneity and a common stylistic evolution through the Viking Age, this situation
was reached by a process that was contingent upon local developments as much as
broader chronological trends. Indeed, there are tangible differences between the
comb repertoires of early historic Scotland and contemporary Anglo-Saxon England
(cf Ashby 2009; 2012). Thus, if the combs are to be believed, the social context of
9th-century Atlantic Scotland should not be characterized in terms of ‘Norseman’
and ‘native’, but was rather a matter of the entrance of new players onto a stage on
which complex local and regional identities were already being negotiated.
Such negotiation was to continue through successive generations, and from the
10th century onward the development of Atlantic Scotland’s comb corpus begins to
hint at new patterning in terms of contacts, communities and identities. Indeed, one
might argue that the combs of later Viking Age Europe as a whole are characterized
by a greater degree of diversity than previously seen (cf Ashby 2009). Moreover, this
MAM 37: 18-ashby – Press

262 maritime societies of the viking and medieval world

Figure 18.1 Main comb types referred to

in the text: type 6 from Birka; types 7, 8a,
and 8b from York; type 8c from Durham;
type 9 from Trondheim; type 13 from
Freswick Links, Caithness, type 14b from
York (Richard Jackson, Hayley Saul and Pat
Walsh, after Ashby and Batey 2012, fig 13.3)
MAM 37: 18-ashby – Press

disentangling trade 263

variety is not randomly distributed, but displays conspicuous patterning when viewed
at the appropriate frame of reference. Thus, in order to put the Scottish material into
context, it is important that we consider both the local trajectory of development (see
above), and broader, long-range trends.
From the 10th century, northern England, Ireland and the southern Baltic see
the appearance of types 6, 7 and 8 (Ashby 2011). Type 6 combs are short (less than
c150mm) in length, with a straight profile, and connecting plates that have a deep
plano-convex section. They have previously been referred to as B combs (Ambrosiani
1981, 62), though Dunlevy (1988, 364–365) includes them within her broader class
‘F2’. They may be decorated with ring-and-dot, interlace, or simple incised-line
ornament. However, this diverse repertoire of motifs is expressed within a restricted
grammar, and the range of decorative schemes is limited (see Ambrosiani 1981,
figs 28–33; Ashby 2006a, tab 8.24).
Type 7 combs have connecting plates of a similarly deep plano-convex section,
but are much longer (generally being over 170mm from end to end), and have a
characteristically naturalistic, irregular profile. This may be broadly concavo-convex
or straight, as it seems to reflect the morphology of the antler from which the blanks
were cut. Type 7 combs are decorated with incised-line ornament or ring-and-dot
motifs; central fields containing saltires or debased interlace are the most common
theme.
Type 8 combs may be compared to both types 6 and 7 in morphological terms,
but differ fundamentally in connecting-plate form. There are three subclasses: type 8a
(derived from type 6) has connecting plates of triangular section, and may be
decorated with simple incised line ornament, frequently in the form of chevrons.
Type 8b (derived from type 7) has connecting plates of trapezoidal section, com-
monly ornamented with groups of incised lines arranged to form a chequerboard
pattern. Finally, type 8c (possibly derived from a variant of type 9; see below) has
connecting plates of semi-circular section and straight profile, and lacks ornament
altogether. The three subtypes date to between the 10th and 13th centuries, with
type 8a probably being the earliest, followed by 8b and then 8c, though there is some
overlap. The three subtypes are clearly connected, showing similarities of morphol-
ogy and decorative treatment, and often occurring at the same sites.3
In contrast, from the late 10th century in Scandinavia we see the arrival of a new
form of comb, type 9, and from the end of the 12th century, its double-sided suc-
cessor, type 13. These two forms are separated from all previous types by their
rejection of complex incised ornament in favour of openwork designs, decoratively
arranged rivets and metal plate. These divergent traditions find their confluence
in late- and post-Viking Age Atlantic Scotland, and it is the geography of this
phenomenon that is of interest herein (see Figure 18.2).

distribution: type 6 combs

Type 6 combs were clearly the successors of type 5, developing — as Ambrosiani
(1981, 32–40) suggests — in the Baltic area in the earliest 10th century. The type was
clearly known in central Sweden, but, though a few examples are known from the
MAM 37: 18-ashby – Press

264 maritime societies of the viking and medieval world

Figure 18.2 Key sites and localities referred to in the text: (1) Orkney and Shetland,
including Brough of Birsay, Buckquoy, Skaill in Deerness, Quoygrew, Pool and Jarlshof;
(2) Western Isles, including Bornish, Cille Pheadair, the Udal and Cnip; (3) Dublin;
(4) York; (5) Ribe; (6) Hedeby and Schleswig; (7) Trondheim; (8) Bergen; (9) Kaupang and
Oslo; (10) Birka and Sigtuna; (11) Lund (S P Ashby)

earliest phases of towns such as Trondheim (Flodin 1989, ill 9), their precise chron-
ology and distribution in northern and western Scandinavia is less easy to assess.4
Nonetheless, type 6 combs are much less common in Scotland (Table 18.1) than
they are in southern Scandinavia, and are absent from the 10th-century comb col-
lections of a number of key sites, including Beachview (Batey 1996, 142–144) and the
Brough Road, Birsay (Batey 1989). In fact, aside from two examples found at Skaill,
Deerness, and one (complete with a case) from a burial in the Bay of Skaill (west

Table 18.1 Distribution of comb types in Atlantic Scotland. Numbers in bold indicate
examples in which 50% or more of the total length is preserved; those in parentheses are
fragments representing less than 50% of total length

6 7 8a 8b 8c 9 13 Total

Northern Scotland 3 (4) 2 1 (7) 0 (4) 2 (2) 28 (21) 27 (17) 63 (55)

Western Scotland 1 (3) 3 (1) 2 (13) 2 (9) 3 (16) 1 (1) 3 (3) 15 (46)
Total 4 (7) 5 (1) 3 (20) 2 (13) 5 (18) 29 (22) 30 (20) 78 (101)
MAM 37: 18-ashby – Press

disentangling trade 265

Mainland) (Grieg 1940, 81–82, fig 45), well-preserved examples appear to be absent
in Orkney. Comb fragments at sites such as the Brough of Birsay (Curle 1982, ill 49,
no 228) and Quoygrew, Westray (Ashby and Batey 2012) are possible representatives
of the type, but the paucity of clearly identifiable examples is notable. The same is
true in Shetland, while evidence from the Western Isles is similarly sparse, though
there are type 6 combs and fragments at Bornish on South Uist, and Dun Mor Vaul,
Tiree (Ashby 2006a, tab 8.5). The latter was found in the wall space of a broch (a pre-
Viking Age monumental dwelling), presumably deposited there at some point follow-
ing the building’s abandonment (MacKie 1974, 90–91).
In addition, comb cases from Earl’s Bu, Orphir (Batey and Morris 1992, 33),
Pool, Sanday (Smith 2007), Quoygrew, Westray (Ashby and Batey 2012), Freswick
Links, Caithness (Batey 1987, 227–228) and Scalloway, Shetland (Smith 1998, 157–
158) are suggestive of type 6. Intuitively, the compact, regular dimensions of type 6
combs do seem to lend themselves to use with a case, and a number of examples bear
endplate perforations that may have accommodated pegs, intended to allow the
combs to pivot against their cases. Moreover, at sites such as Birka (Ambrosiani 1981,
66–67) and Novgorod (Smirnova 2005, 77), the association of type 6 combs and cases
is clear. Nonetheless, even taking cases into consideration, the evidence for type 6
combs in Scotland is rather slim, such that one might suggest that the type was only in
rather irregular use in the 10th and 11th centuries. This is notable, particularly given
the possibility that in Scandinavia, type 6 was more common in the south and east
than the west (though see note 3).

distribution: type 7 combs

Type 7 represents a late Viking Age development from types 5 or 6, and seems to
have been most popular in England (eg MacGregor et al 1999, fig 889; Oakley and
Harman 1979, fig 137), Ireland (Dunlevy 1988, 364–365, fig 7; Hurley and Scully
1997, fig 17:2, no 18), and on the south coast of the Baltic (Cnotliwy 1973, 128, 314–
316, Ryc 48, d–e; Tempel 1969, Taf 23). In contrast, the type is poorly represented in
northern Scandinavia and its absence from the large Birka collection (Ambrosiani
1981; Ashby 2006a, tab 6.7), has led to its neglect in much subsequent debate. None-
theless, the type is of key significance in the study of contact between Britain, Ireland
and southern Scandinavia.
Conversely, the type is poorly represented in Scotland, even in comparison with
type 6. The furnished burial at Cnip, Lewis provides the clearest example, though
there are isolated occurrences at the South Uist settlements of Bornish and Cille
Pheadair (Ashby 2006a, 191), and a probable fragmentary specimen from the Brough
of Deerness, Orkney (Gerrard and Barrett 2010, 8).5 Given the paucity of examples
known from northern Scandinavia (see above; Ashby 2011), the infrequency of the
type in Scotland as a whole is notable, while the fact that a number of these ‘Scottish’
examples were recovered in the Western Isles is arguably indicative of an Irish
connection. Given their similarities with English and Irish type 7 combs, one might
speculate that the western examples were manufactured in the Irish Sea region, rather
than in Scandinavia. If this is the case, then the Cnip comb at least was made fairly
MAM 37: 18-ashby – Press

266 maritime societies of the viking and medieval world

soon after arrival in the region, as it was deposited in a furnished grave that probably
predates ad 950 (Graham-Campbell and Batey 1998, 154).

distribution: type 8 combs

Type 8 combs are best known in Ireland, England, and the southern Baltic area
(Cnotliwy 1973, 315, Ryc 56–57; Tempel 1969, 98–100). Key collections include those
from Waterford (Hurley and Scully 1997, 654–658) and Dublin,6 but examples are
also known from smaller Irish settlements such as Knowth and Ballinderry (Dunlevy
1988, 366–367, fig 9). In England, type 8 combs make up important components of
the collections at York (MacGregor et al 1999, fig 886) and Lincoln (Ashby 2006a,
tab 7.11) , and isolated examples have been recorded at sites across the country, as far
apart as Lancashire (Cook and Batey 1994) and Kent (Riddler 1991). Further afield,
the type is identifiable at Haithabu (Tempel 1969, 98–102, Taf 24, 25, 39), Wolin and
Gdansk (Cnotliwy 1973, Ryc 56–57), but is not common north of the Skagerrak.
Relative to types 6 and 7, type 8 combs are quite frequently recovered at late
Viking Age and medieval (‘Norse period’) sites in Atlantic Scotland, particularly in
the west. In the Northern Isles, examples have been identified at the following sites:
Snusgar, Sandwick, Orkney (see Chapter 16); Pool, Sanday (Smith 2007, ill 8.8.5,
PL0165, PL0356); Quoygrew, Westray (Ashby 2006a, fig 8.12; Ashby and Batey 2012);
and Jarlshof, Shetland (Hamilton 1956, 148, plate XXX, no 3). Possible fragmentary
examples are also known from Beachview, Birsay, Orkney (Batey 1996, ill 117), while
a well-preserved comb from the Brough Road, also in Birsay (Batey 1989, ill 154) is
unusual, but might represent a variant of the type. However, the densest distributions
of type 8 combs lie along Scotland’s western coast and islands, at Whithorn, Gallo-
way (Nicholson 1997, 480–485) and, in particular, at Bornish and Cille Pheadair,
South Uist (Ashby 2006a, 192–193; Parker Pearson et al 2004, fig 8; Sharples 2005,
fig 33, no 1438 and 1467; see Chapter 17). The connection between Ireland and
western Scotland suggested by the distribution of type 7 is thus rendered even more
tangible when we consider type 8.

distribution: type 9 combs

Type 9 combs were direct contemporaries of type 8 (Ashby 2006a, 130–132), and
from the late 10th century they are frequent finds across northern Scandinavia; so
much so that they are often considered type finds of ‘Late Norse’ (medieval rather
than Viking Age) settlement. Excavations in the city of Trondheim recovered a large
number of such combs from levels dated closely to between the late 10th and 15th
centuries. A similar situation exists at Norway’s other late Viking Age/medieval
towns (see Hansen 2005, 180–184, fig 44; Wiberg 1977; 1979; 1987), and we see the
same pattern at Sigtuna (Ashby 2006a, 125–126), Lund (Persson 1976, 323–326),
Kungähalla (Rytter 1991) and Visby (Carlsson 2002b). At Birka one can arguably see
the beginnings of the trend, as type 6 combs are here fixed with copper-alloy rivets,
and incised decoration declines in importance (Ambrosiani 1981, fig 33). The situ-
ation in southern Scandinavia is less clear-cut, as at sites in this region type 9 combs
MAM 37: 18-ashby – Press

disentangling trade 267

appear to occur alongside type 8. This is particularly true at Schleswig (see Ulbricht
1984, Taf 66, 67, 70, 71, 72). Nonetheless, type 9 combs spread far and wide in the
11th and 12th centuries, reaching as far west as Iceland (eg Amorosi 1992, fig 7), and
as far east as Estonia (Luik 1998; 2001) and European Russia (Smirnova 2005,
figs 3.46–3.48).7
As one might expect, type 9 combs also reached Atlantic Scotland, though their
distribution there is far from even. In contrast to types 7 and 8, type 9 combs are
much better represented in Caithness and the Northern Isles than in the west and the
Hebrides. In Orkney, where type 9 combs and fragments are particularly frequently
recorded, notable examples come from the Upper Norse Horizon at the Brough of
Birsay and sites at Skaill, Galilee, Ivar’s Knowe and Tofts Ness. There are several
examples from medieval contexts at Beachview, Birsay (Batey 1996, 143–144; Batey
and Freeman 1996, 59–62), and more recently, type 9 combs have been excavated
from 11th- to 14th-century deposits at Quoygrew (Ashby and Batey 2012), and from
the final occupation phase at Langskaill (Moore and Wilson 2005), both on Westray.
There are also 12 unstratified examples from Freswick Links, Caithness (Batey 1987,
205–211, 231, pl 33). In Shetland, there are seven combs (and four fragmentary
examples) from Jarlshof (Hamilton 1956, 166–168), three combs (and five fragments)
from Sandwick North, and another single example from elsewhere in Unst.
In western Scotland, the situation could hardly be more different. Other than
one comb and a single fragment from Bornish (Ashby 2006a, tab 8.5), the author
knows of no examples from Scotland’s western coasts and islands.8 The type is
equally poorly represented south of the Moray Firth. Indeed, few combs are known
from mainland Scotland in general, and composite combs seem to fade out of use in
England from around the 12th century, where their place was presumably taken by
combs of perishable materials such as horn and wood (MacGregor 1989; 1992;
cf Ashby 2006a, 227–228). With this in mind, the small number of type 9 combs south
of the Firth of Forth is perhaps unsurprising, and the very few examples of type 9 that
are present at sites in Yorkshire, Lincolnshire and London are individually note-
worthy (Ashby 2006a, 146–147; see also Smith 1909).
Given the diversity of type 9 sub-forms, it is particularly striking that all
examples known from Atlantic Scotland can be paralleled in the Scandinavian towns
(cf Hansen 2005, figs 44 and 45).9 This, together with the paucity of British and Irish
examples, and the absence of any evidence for the manufacture of type 9 combs in an
English, Scottish, Welsh or Irish context,10 tends to support the notion that such
combs reached Britain and Ireland as the products of travel and trade.

distribution: type 13 combs

Type 13 combs first appear in the archaeological record around the end of the
12th century. From this date on, they are very well known from excavations in Oslo,
Bergen, Trondheim, Tønsberg, Sigtuna and Lund (Ashby 2011; see also Flodin 1989,
30–33, fig 44; Persson 1976, 330–332; Wiberg 1987, 418–420). South of the Baltic,
they have been identified at Schleswig (Ulbricht 1984, Taf 76, 77), and further east
in towns such as Tallinn and Tartu, Estonia (Luik 1998, 168–170) and Novgorod,
MAM 37: 18-ashby – Press

268 maritime societies of the viking and medieval world

European Russia (Smirnova 2005, 297–304). Indeed, they are known across much of
the Norse diaspora (eg Amorosi 1992, 119), and have been accepted rather passively
as a marker of Scandinavian medieval ‘culture’. The situation is, as ever, more com-
plex than this, and it is instructive to briefly consider their distribution in Atlantic
Scotland.
In Orkney, there are examples from Langskaill, Westray (Moore and Wilson
2005), and Howar, North Ronaldsay, as well as several less well-provenanced spot
finds. Dated contexts are few, but examples are known from an early 14th-century
deposit at Quoygrew, Westray (Ashby and Batey 2012), and probable 11th- to 13th-
century phases at Pool, Sanday (Hunter 2007, 162). One might also note examples
from the Pentland Skerries, while there are 17 effectively unstratified finds from
Freswick Links, Caithness. In Shetland, examples have been identified at the 11th- to
13th-century site of Sandwick North (Stummann Hansen 2000, 95–96), the Sands of
Breckon, and medieval phases at Jarlshof (‘Viking and Later Norse VII’: Hamilton
1956, 179–187). In contrast, the presence of type 13 combs on Scotland’s western
coasts and islands is conspicuously small. There are three examples from Bornish,
together with three fragments, but the type appears to be absent elsewhere in the
record.11
Beyond Scotland, type 13 combs are poorly represented in Britain and Ireland,
though notable exceptions are known from York (MacGregor et al 1999, fig 892,
no 7682), Beverley (Foreman 1991, 185, 191, fig 128, no 1120) and Stafford (Ashby
2010) in England. Similarly, their representation in Ireland appears to consist of a
small number of isolated examples (Dunlevy 1988, 369, fig 10, no 1; Hurley and Scully
1997, fig 17:1, no 15). Their distribution then, bears much similarity to that of type 9.

discrete traits
The typology outlined in Ashby 2006a and 2011, and referred to above, provides
a broad framework for analysis. However, a deeper understanding of regional and
temporal variability is attainable through more detailed analysis, considering dis-
crete traits of ornament and manufacture, alongside overall morphology.12 Of over
50 discrete traits used in the definition of the type series, two proved to be of par-
ticular use in the study of late Viking Age and medieval combs, and are discussed in
detail herein. These traits relate to the use of (usually) metal rivets in constructing
and ornamenting the comb.
The first trait of significance relates to the materials used, which may be iron,
copper alloy or bone. Distinctive patterning was identified, suggesting that preference
for particular riveting materials was locally variable, dictated by a combination of
manufacturing traditions and raw material availability (Table 18.2). One may note
the invariable use of iron rivets with combs of types 6, 7 and 8, and the contrasting
employment of copper alloy rivets in types 9 and 13.13 These patterns are of key
significance.
The second aspect is the placement of rivets. One might assume that there would
be a standardized way of arranging rivets with respect to the toothplates they have to
hold in place, but in practice a number of distinctive styles are identifiable, and these
MAM 37: 18-ashby – Press

disentangling trade 269

Table 18.2 Rivet materials in combs from Scotland. Numbers in bold indicate examples in
which 50% or more of the total length is preserved; those in parentheses are fragments
representing less than 50% of total length

Iron Copper alloy Unknown

6 3 (5) 1 (1)
7 5 (1)
8a 3 (16) (4)
8b 2 (12) (1)
8c 5 (18)
9 1 (0) 28 (23)
13 0 (1) 30 (16) (3)
Total 19 (53) 59 (39) 0 (9)

are to a certain degree regionalized. For instance, rivets may be placed at alternating
edges of toothplates, a pattern almost ubiquitous in England (Ashby 2012, fig 3), and
also found in early combs from Atlantic Scotland (Ashby 2009), as well as Haithabu
(Tempel 1969, 71–107, Abb 27), Wolin and Gdansk (Cnotliwy 1973, 314). In con-
trast, combs from Viking Age northern Scandinavia were frequently fixed at every
edge, or centrally (through the longitudinal mid-point of each billet). The patterns
thus have great potential for identifying local traditions and schools of manufacture.
The trait needs to be considered more carefully in the medieval period, however,
as this context sees the mass production of large numbers of type 9 and 13 combs,
which (barring the earliest, probably late 10th century, examples of type 9) are fixed
with what we may consider supernumerary rivets, applied with aesthetics, rather
than mechanics, in mind. Further, more systematic research is necessary if regional
patterning is to be consistently identified in this material, but initial investigations
show some promise (Hansen 2005, 184).
Given the high level of comb fragmentation, the sample size for the Scottish
material is relatively small14 (Tables 18.3 and 18.4), and the most noticeable pattern-
ing relates to the difference between riveting styles in types 6, 7 and 8 on the one
hand, and types 9 and 13 on the other. This is not particularly instructive. However,
one may also note that combs of types 6, 7 and 8 seem to show a greater level of
uniformity in western Scotland than in the north. It is just possible that this is
indicative of the dominance of a single source or workshop in the west, contrasting
with the more mixed picture identifiable in the north. If this suggestion is to be tested
confidently, a greater quantity of combs from a wider range of sites is necessary; we
thus await further excavation in the area.

discussion
Though differences are apparent in the pre- and early Viking periods (see Ashby
2009; 2012), the trajectories of comb development in Atlantic Scotland and northern
England diverge markedly in the 10th and 11th centuries. Types 6 and 7 — relatively
MAM 37: 18-ashby – Press

270 maritime societies of the viking and medieval world

Table 18.3 Riveting techniques in combs from northern Scotland. Complete combs
and fragments representing over 50% of total comb length only

Alternating Central Decorative Every edge Mixed Unknown

6 1 1 1
7 1 1
8a 1
8b
8c 1 1
9 1 19 2 1 5
13 1 24 1 1
Total 2 2 43 4 4 8

Table 18.4 Riveting techniques in combs from western Scotland. Complete combs
and fragments representing over 50% of total comb length only

Alternating Central Decorative Every edge Mixed Unknown

6 1
7 3
8a 2
8b 2
8c 2 1
9 1
13 2 1
Total 9 0 2 3 1 0

common in this period at sites in northern England and Ireland — are infrequent
finds in Scotland, and, where present, may indicate contact with an Anglo-
Scandinavian or Hiberno-Norse milieu. The type 7 comb from the burial at Cnip,
Lewis, is a case in point, and the fact that this was associated with a Hiberno-Norse
ring-headed pin is perhaps of particular note (Welander et al 1987, 170).
While such isolated occurrences of types 6 and 7 provide considerable potential
for biographical study (eg Ashby 2006a, 253–258), more informative patterning is
evident in the distribution of types 8, 9 and 13. Type 8 combs are most common in the
west of Scotland, dominating the collections at Bornais and Cille Pheadair (see Parker
Pearson et al 2004, fig 8), but they are also known in both the Northern Isles and
England. However, from the end of the 10th century, the distinctive, copper-alloy
riveted type 9 combs imported from Norway began to dominate in northern
Scotland, and they continued to do so until the advent of type 13 200 years later. Both
types 9 and 13 are poorly represented in the west; one might even suggest that by this
point, Argyll and the Hebrides were effectively cut off from Scandinavian networks,
instead becoming increasingly integrated into an Irish Sea trading province that
united Ireland and western England and Scotland. In short, the combs are indicative of
MAM 37: 18-ashby – Press

disentangling trade 271

Figure 18.3 ‘Comb worlds’ of the 10th to 15th centuries: a schematic

interpretation. At sites within each region, the indicated types dominate assemblages.
Nonetheless, combs of other types are occasionally present in these regions,
particularly close to the edges of the suggested ranges (eg Orkney, northern Jutland).
A similar pattern is apparent in the distribution of riveting technologies (S P Ashby)

differing networks of contact, organized within more-or-less discrete regional arenas

(Figure 18.3). Notwithstanding regional differences in rates of typological change, the
potential manufacture of poor-quality local imitations of overseas fashions, and the
deposition of a small number of foreign ‘curios’ or travellers’ possessions, in general
terms there is greater unity within than between these zones.
How might one explain this apparent disarticulation? I have shown elsewhere
(Ashby 2009; 2012) that in the early Viking Age there was some disparity between the
comb repertoires of England, Scotland, Ireland and Scandinavia, both in terms of
morphology and manufacturing technique. Localized practice is evident in collections
from England and Scandinavia, with English and Scandinavian workshops manufac-
turing their own distinctive products according to local traditions and regional
fashions (Ashby 2012). By the 10th and 11th centuries, one might note some measure
of overlap between the combs being produced in Anglo-Scandinavian and Hiberno-
Norse workshops and those of southern Scandinavia, but the situation north of the
Baltic remains distinctive.
Interestingly, evidence for manufacture in Scotland is lacking for this period,
and northern Scotland in particular still seems tied into the Scandinavian system. It
appears that the ready supply of type 9 combs from Scandinavia, and type 8 combs
from the Irish Sea region, negated the production of local Orcadian or Shetlandic
MAM 37: 18-ashby – Press

272 maritime societies of the viking and medieval world

types. It is unlikely that antler was easily available in the Northern Isles at this date,
but an economically determinist explanation for what must have been a funda-
mentally social phenomenon is unsatisfactory. Type 8 and 9 combs are so distinctive
in form and ornament that their social connotations must have been widely under-
stood. Consumer choice may have been limited by the circles within which an indi-
vidual moved, or the markets to which they had access, but it may also have been
conditioned by the traditions with which they were most familiar or comfortable.
Thus, when a person used or wore their comb, they were (consciously or uncon-
sciously) making social, economic, geographical and/or political references to the
world in which they lived.
The situation in western Scotland seems to have been slightly different. Though
there is still a paucity of evidence for manufacture here, the dominance of type 8
combs is regionally distinctive. Whether these combs were being made in Dublin or in
Waterford, in Galloway or in the Hebrides, their regional character is clear (and is
not diminished by the popularity of the type in England and the south-western Baltic;
see below). Thus, while the influence of the Norwegian comb milieu stretched across
the North Sea, where it came head-to-head with the ‘type 8 world’ in Orkney and
Shetland, the rarity of combs of types 9 and 13 in the west clearly demonstrates the
relative separation of the Irish Sea area.
This stylistic dislocation is particularly tangible in the 13th century and later, by
which time Scandinavian workshops were producing combs in the new double-sided
fashion: type 13 (see Flodin 1989, ill 26–35; Ulbricht 1980; Wiberg 1987). These also
began to dominate in the Northern Isles, but must have been much less common in
the west.15 Indeed, it is interesting to note that type 8 combs were probably being
produced in Bornais’ short-lived workshop at this date, approximately a century after
their heyday in England and Ireland. Moreover, the production of composite combs
had all but ceased in England by the 13th century, presumably superseded by one-
piece combs in organic materials such as cattle horn and boxwood. While it is per-
haps a step too far to divide Insular Europe into two ‘worlds’ on the basis of
patterning in a single artefact form, we can nonetheless identify distinctive northern
and south-western trends, that may relate to particular ‘comb behaviours’.
Moreover, such distinctive patterning in habits of dress and practice may be
associated with broader social trends, and the combs may be suggestive of closer links
between political and economic geography than one might expect. The Western Isles
have a complex and dynamic political history, and though they officially became
Scottish territory at the Treaty of Perth in ad 1266, they were locally ruled for long
periods prior to this date. Trade in the Irish Sea began to flourish from the 10th and
11th centuries onward, and the connections between the ports of Bristol, Chester and
Dublin are certainly well attested at this time (eg Graham-Campbell 1992; Griffiths
1996; Mytum 2003; Sivier 2002). It does not seem inappropriate to add western
Scotland into this group, particularly given the status and scale of activity evidenced
at Bornais. In contrast, the Northern Isles continued in Scandinavian overlordship
until ad 1468, and — more importantly — direct contact with Norway seems to have
been maintained for much of this period. This persistence of close economic and
political ties with Scandinavia goes some way to explaining the profusion of Norse
MAM 37: 18-ashby – Press

disentangling trade 273

place names in the Northern Isles, as well as the suggestion of considerable Scan-
dinavian influence on the Orcadian gene pool (Barrett 2003, 78, 91).
Thus, while I do not suggest that the trends evident in the comb corpus may all
be explained away in historical terms, the correlation between the patterning in this
particular form of dress accessory (whether it relates to personal choice, or differ-
ential access to markets), and broader political trends, is striking. It may be that this
is an isolated phenomenon, particular to combs and their perceived raw material
requirements, but this should not be assumed a priori. Rather, research into other
areas of material culture (such as pins, dress accessories, textile equipment or stone
objects) is necessary, as it is only by means of such specialist analyses that we may
begin to identify further regionalities. Comparative work, considering the synchro-
nicities and conflicts of these various expressions of identity may in turn help us to
identify meaningful currents in material culture experience and political develop-
ment, and this must remain our goal. It is only through the focus on the particular
that local context and contingency can be appreciated, and it is only through such an
appreciation that we may attain a nuanced understanding of past social dynamics.16

notes
1 For the sake of brevity, detailed discussions of combs from Novgorod, and their apparent idio-
context, stratigraphy and dating are not included syncrasy, seem to be indicative of the form being
herein. See Ashby 2006a; 2009 for details. manufactured locally there (Smirnova 2005,
2 There is some debate regarding the Frisian or 305–308).
Scandinavian origins of type 5 (see, for example, 8 There are also examples from the Udal, North
Callmer 1998). This is not the venue to address Uist (illustrated in Crawford 1996), but numbers
the issue in detail, but whatever their mechanism are not given, and we await full publication of
of movement, the symmetry between their distri- the site.
bution and the expansion of the Norse colonies 9 Moreover, this applies to all examples from
in the 9th and 10th centuries is notable. Britain and Ireland seen by the author.
3 The chronology of types 8a, 8b and 8c has yet to 10 The archaeological signature for the manufac-
be subjected to scrutiny, though internal rela- ture of these combs is relatively well understood,
tionships are discussed in some detail in Ashby as towns such as Bergen and Lund have pro-
2011. For the sake of parsimony, the three sub- duced evidence in the form of debitage, semi-
types are treated here as a single class: ‘type 8’. manufactures, and tools. Nothing comparable is
However, sub-type data are provided in tabular known from Scotland, while material from
form, in order that they might be considered England (eg York) and Ireland (eg Dublin) is
more closely. clearly from the production of a different
4 Most of the combs known from western Scandi- product (type 7 and 8 combs).
navia come from interventions in the medieval 11 Again, one must note the exception of the Udal
towns. Excavations of earlier settlements, such (Crawford 1996).
as Kaupang (see Skre and Stylegar 2004), have 12 Investigations into raw material analysis (antler
produced few combs, though there are examples species) were also undertaken, but results from
from furnished graves, and from recent excava- macroscopic analyses proved equivocal, due to
tions at Bjørkum, Laerdal, Norway (Ramstad the number of combs for which raw materials
et al 2011). were unidentified. This problem relates to the
5 There is also a possible example from Jarlshof fact that medieval combs (particularly types 9
(Hamilton 1956, fig 69, no 12) though this comb and 13) are frequently finished to such a high
is morphologically distinct, and may simply rep- quality that areas of exposed macrostructure are
resent an attempt to imitate the type 5 template, rarely preserved (see Ashby 2013 for a critical
rather than being a genuine example of type 7. review of the bases, potentials, and practical
6 The publication of combs from recent excava- applications of these macroscopic techniques).
tions in Dublin is eagerly awaited. Destructive analyses such as aDNA (eg Arndt
7 However, the situation to the east of the Baltic et al 2003), or stable isotopes (eg Barrett et al
Sea is not directly comparable; the numbers of 2008; Sykes et al 2006) provide an alternative
MAM 37: 18-ashby – Press

274 maritime societies of the viking and medieval world

approach, but the greatest potential appears to attribution of style is impossible for small frag-
lie with collagen proteomics (Collins et al 2008). ments.
Indeed, recent work on combs from early his- 15 A notable exception comes from Bornais, South
toric (including Viking Age) Scotland has demon- Uist.
strated the potential of such an approach in the 16 This paper builds upon research undertaken as
confirmation of material identification, and infer- part of my AHRC-funded PhD thesis on bone
ence of provenance (von Holstein et al 2014). and antler hair combs from Viking Age and
13 However, note that in eastern Scandinavia type 6 medieval northern Britain (Ashby 2006a), and
combs are frequently fixed with copper-alloy thanks are due to my supervisor, James Barrett.
rivets (see in particular Ambrosiani 1981, 72). I must also thank Julian Richards, Terry
Small fragments are excluded from this analysis, O’Connor and Arthur MacGregor, and the
shrinking the sample size considerably. many individuals and institutions that allowed
14 This precaution is necessary, however, as rivet- me access to their collections, data, and conver-
ing technique is not always entirely consistent sation.
within a single comb, meaning that confident

references
Ambrosiani, K, 1981 Viking Age Combs, Comb Making and Comb Makers in the Light of Finds
from Birka and Ribe, Dept of Archaeology, University of Stockholm, Stockholm
Amorosi, T, 1992 ‘Climate impact and human response in northeast Iceland: archaeological
investigations at Svalbard, 1986–1988’, in C D Morris and D J Rackham (ed), Norse and
Later Settlement and Subsistence in the North Atlantic, Department of Archaeology, Univer-
sity of Glasgow, Glasgow, 103–148
Arndt, A, van Neer, W, Hellemans, B, Robben, J, Volckaert, F and Waelkens, M, 2003 ‘Roman
trade relationships at Sagalassos (Turkey) elucidated by ancient DNA of fish remains’,
Journal of Archaeological Science 30, 1095–1105
Ashby, S P, 2006a Time, trade and identity: bone and antler combs in northern Britain c. ad
700–1400, Unpublished PhD thesis, University of York, York
Ashby, S P, 2006b ‘Trade in Viking Age Britain: identity and the production and distribution of
bone and antler combs’, in J Arneborg and B Grønnow (ed), Dynamics of Northern Societies:
Proceedings of the SILA/NABO Conference on Arctic and North Atlantic Archaeology,
Copenhagen, May 10th–14th, 2004, National Museum of Denmark, Copenhagen, 273–279
Ashby, S P, 2007 ‘Bone and antler combs’, Finds Research Group Datasheet 40
Ashby, S P, 2009 ‘Combs, contact, and chronology: reconsidering hair combs in early-historic and
Viking-Age Atlantic Scotland’, Medieval Archaeology 53, 1–33
Ashby, S P, 2010 ‘Comment on the comb’, in M O H Carver and C Spall (ed), The Birth of a
Borough: An Archaeological Study of Anglo-Saxon Stafford, Boydell, Woodbridge, 116–117
Ashby, S P, 2011 ‘An atlas of medieval combs from northern Europe’, Internet Archaeology 30,
http://dx.doi.org/10.11141/ia.30.3
Ashby, S P, 2012 ‘A study in regionality: hair combs and bone/antler craft in the northern
Danelaw’, in D Petts and S Turner (ed), Early Medieval Northumbria, Brepols, Turnhout,
303–319
Ashby, S P, 2013 ‘Some comments on the identification of cervid species in worked antler’, in
A Choyke and S O’Connor (ed), From These Bare Bones: Raw Materials and the Study of
Worked Osseous Objects, Oxbow, Oxford, 208–222
Ashby, S P and Batey, C E, 2012 ‘Worked antler and bone’, in J H Barrett (ed), Quoygrew: Being an
Islander in Late Viking Age and Medieval Orkney, McDonald Institute for Archaeological
Research, Cambridge, 229–243
Barrett, J H, 2003 ‘Culture Contact in Viking Age Scotland’, in J H Barrett (ed), Contact,
Continuity, and Collapse: The Norse Colonization of the North Atlantic, Brepols, Turnhout,
74–111
MAM 37: 18-ashby – Press

disentangling trade 275

Barrett, J H, Johnstone, C, Harland, J, van Neer, W, Ervynck, A, Makowiecki, D, Heinrich, D,

Hufthammer, A K, Enghoff, I B, Amundsen, C, Christiansen, J S, Jones, A K G, Locker, A,
Hamilton-Dyer, S, Jonsson, L, Lougas, L, Roberts, C and Richards, M, 2008 ‘Detecting the
medieval cod trade: a new method and first results’, Journal of Archaeological Science 35,
850–861
Batey, C E, 1987 Freswick Links, Caithness: A Reappraisal of the Late Norse Site in its Context,
British Archaeological Reports, British Series 179, Archaeopress, Oxford
Batey, C E, 1989 ‘Excavations beside the Brough Road, Birsay: the artefact assemblage’, in
C D Morris (ed), The Birsay Bay Project, Volume 1: Coastal Sites Beside the Brough Road,
Birsay, Orkney Excavations 1976–1982, Department of Archaeology, University of Durham,
Durham, 191–229
Batey, C E, 1996 ‘7.3. Area 1: the artefactual assemblage’, in C D Morris (ed), The Birsay Bay
Project, Volume 2: Sites in Birsay Village and on the Brough of Birsay, Orkney, Department of
Archaeology, University of Durham, Durham, 133–147
Batey, C E and Freeman, C, 1996 ‘5.3. Excavations at Beachview Burnside Area 2: the artefactual
assemblage’, in C D Morris (ed), The Birsay Bay Project, Volume 2: Sites in Birsay Village and
on the Brough of Birsay, Orkney, Department of Archaeology, University of Durham,
Durham, 56–64
Batey, C E and Morris, C D, 1992 ‘Earl’s Bu, Orphir, Orkney: excavation of a Norse horizontal
mill’, in C D Morris and D J Rackham (ed), Norse and Later Settlement and Subsistence in the
North Atlantic, Department of Archaeology, University of Glasgow, Glasgow, 33–41
Callmer, J, 1998 ‘Archaeological sources for the presence of Frisian agents of trade in northern
Europe ca. ad 700–900’, in A Wesse (ed), Studien zur Archäologie des Ostseeraumes Von der
Eisenzeit zum Mittelalter: Festschrift für Michael Müller-Wille, Wachholtz, Neumünster,
469–481
Carlsson, D, 2002a Combs and Comb Making in Viking Age and Middle Ages: A Short Resumé,
ArkeoDoc, Visby
Carlsson, D, 2002b Viking and Medieval Combs from the Island of Gotland, Sweden (CD-Rom),
ArkeoDoc, Visby
Christophersen, A, 1980 Håndverket i Forandring. Studier i Horn- og Beinhåndverkets Utvikling i
Lund ca. 1000–1350, Gleerup, Lund
Clarke, D and Heald, A, 2002 ‘Beyond typology: combs, economics, symbolism and regional
identity in Late Norse Scotland’, Norwegian Archaeological Review 35, 81–93
Cnotliwy, E, 1973 Rzemiosło Rogownicze na Pomorzu Sczesnośredniowiecznym [Antler Handi-
craft in Early Medieval Pomerania], Zakład Narodowy imienia Ossolińskich Wydawnictwo
Polskiej Akademii Nauk, Wrocław
Collins, M, Cappellini, E, Buckley, M, Penkman, K E H, Griffin, R C and Koon, H E C, 2008
‘Ancient proteins: what remains to be detected?’, in H Schutkowski (ed), Between Biology and
Culture, Cambridge Studies in Biological and Evolutionary Anthropology, Cambridge
University Press, Cambridge, 165–185
Cook, A and Batey, C E, 1994 ‘The bone comb’, in T W Potter and R D Andrews, ‘Excavation and
survey at St Patrick’s chapel and St Peter’s church, Heysham, Lancashire, 1977–78’, The
Antiquaries Journal 85, 122–124
Crawford, I A, 1996 ‘The Udal’, Current Archaeology 147, 84–94
Curle, C L, 1982 Pictish and Norse Finds from the Brough of Birsay 1934–74, Society of
Antiquaries of Scotland, Edinburgh
Dunlevy, M M, 1988 ‘A classification of early Irish combs’, Proceedings of the Royal Irish
Academy 88, 341–422
MAM 37: 18-ashby – Press

276 maritime societies of the viking and medieval world

Flodin, L, 1989 Kammakeriet i Trondheim ca. 1000–1600, Folkebibliotekstomten, Meddelelser Nr

14. Riksantikvarien, Trondheim
Foreman, M, 1991 ‘The bone and antler’, in P Armstrong, D G Tomlinson and D H Evans (ed),
Excavations at Lurk Lane Beverley, 1979–82, J R Collis, Sheffield, 183–196
Galloway, P and Newcomer, M, 1981 ‘The craft of comb-making: an experimental enquiry’,
University of London Institute of Archaeology Bulletin 18, 73–90
Gerrard, J and Barrett, J, 2010 ‘An enigmatic Viking-age nucleated settlement in Orkney’,
Medieval Archaeology 54, 423-429
Graham-Campbell, J (ed), 1992 Viking Treasure from the North West: The Cuerdale Hoard in its
Context, National Museums and Galleries on Merseyside, Liverpool
Graham-Campbell, J and Batey, C E, 1998 Vikings in Scotland: An Archaeological Survey,
Edinburgh University Press, Edinburgh
Grieg, S, 1940 Viking Antiquities in Great Britain and Ireland, Part II: Viking Antiquities in
Scotland, H. Aschehoug, Oslo
Griffiths, D, 1996 ‘The maritime economy of the Chester region in the Anglo-Saxon period’, in
P Carrington, ‘Where Deva Spreads her Wizard Stream’: Trade and the Port of Chester.
Papers from a Seminar held at Chester, November 1995, Chester Archaeological Society,
Chester, 49–60
Hamilton, J R C, 1956 Excavations at Jarlshof, Shetland, HMSO, Edinburgh
Hansen, G, 2005 Bergen c. 800–1170: The Emergence of a Town, Fagbokforlaget, Bergen
Hunter, J R, 2007 Investigations in Sanday Orkney, Volume 1: Excavations at Pool, Sanday:
A Multi-period Settlement from Neolithic to Late Norse Times, The Orcadian, Kirkwall
Hurley, M F and Scully, O M B, 1997 Late Viking Age and Medieval Waterford: Excavations
1986–1992, Waterford Corporation, Waterford
Luik, H, 1998 Muinas- ja keskaegsed luukammid Eestis, Muinasaja Teadus 6, Ajaloo Instituut,
Tallinn
Luik, H, 2001 ‘Bone combs from medieval Tallinn, from the excavations in Sauna Street’, in
A M Choyke and L Bartosiewicz (ed), Crafting Bone: Skeletal Technologies through Time
and Space, Proceedings of the 2nd Meeting of the (ICAZ) Worked Bone Research Group
Budapest, 31 August–5 September 1999, Archaeopress, Oxford, 321–330
MacGregor, A, 1985 Bone, Antler, Ivory and Horn: the Technology of Skeletal Materials Since the
Roman Period, Croom Helm, London
MacGregor, A, 1989 ‘Bone, antler and horn industries in the urban context’, in D Serjeantson and
T Waldron (ed), Diet and Crafts in Towns: The Evidence of Animal Remains from the
Roman to the Post-medieval Periods, British Archaeological Reports, British Series 199, John
and Erica Hedges, Oxford, 107–128
MacGregor, A, 1992 ‘Antler, bone and horn 400–1500: industries in transition’, in G De Boe and
F Verhaeghe (ed), Art and Symbolism: Medieval Europe 1992 Pre-printed Papers Volume 7,
Medieval Europe 1992, York, 163–168
MacGregor, A, Mainman, A J and Rogers, N S H, 1999 Craft, Industry and Everyday Life: Bone,
Antler, Ivory and Horn from Anglo-Scandinavian and Medieval York, The Archaeology of
York 17/2, Council for British Archaeology, York
Mackie, E, 1974 Dun Mor Vaul: An Iron Age Broch on Tiree, Glasgow University Press, Glasgow
Mainman, A J and Rogers, N S H, 1999 ‘Craft and industry: the range of the evidence’, in
A MacGregor, A J Mainman and N S H Rogers (ed), Craft, Industry and Everyday Life: Bone,
Antler, Ivory and Horn from Anglo-Scandinavian and Medieval York, The Archaeology of
York 17/2, Council for British Archaeology, York, 1903–16
Moore, H and Wilson, G, 2005 ‘The Langskaill Souterrain’, Current Archaeology 199, 333–335
MAM 37: 18-ashby – Press

disentangling trade 277

Mytum, H C, 2003 ‘The Vikings and Ireland’, in J H Barrett (ed), Contact, Continuity, and
Collapse: The Norse Colonization of the North Atlantic, Brepols, Turnhout, 113–137
Nicholson, A, 1997 ‘The antler’, in P Hill (ed), Whithorn and St Ninian: The Excavation of a
Monastic Town, 1984–91, Sutton, Stroud, 474–495
Oakley, G E and Harman, M, 1979 ‘The worked bone’, in J H Williams (ed), St. Peter’s Street
Northampton: Excavations 1973–1976, Northampton Development Corporation, North-
ampton, 308–318
Parker Pearson, M, Smith, H, Mulville, J and Brennand, M, 2004 ‘Cille Pheadair: the life and times
of a Norse-period farmstead c. 1000–1300’, in J Hines, A Lane and M Redknap (ed), Land,
Sea and Home: Settlement in the Viking Period, Maney/Society for Medieval Archaeology,
Leeds, 235–254
Persson, J, 1976 ‘Kammar’, in A V Mårtensson (ed), Uppgrävt förflutet för PK-banken i Lund,
Kulturhistoriska museet, Lund, 317–332
Ramstad, M, Synnøve Halvorsen, L, and Olsen, A B, 2011 ‘Bjørkum: feasting, craft production
and specialisation on a Viking Age rural site in Norway’, Antiquity Project Gallery 86 (328),
http://antiquity.ac.uk/projgall/ramstad328/
Riddler, I, 1990 ‘Saxon handled combs from London’, Transactions of the London and Middlesex
Archaeological Society 41, 9–20
Riddler, I, 1991 ‘Two Late Saxon combs from Longmarket’, Canterbury’s Archaeology 1989/90,
The 14th Report of Canterbury Archaeological Trust Limited, Canterbury Archaeological
Trust Limited, Canterbury, http://www.hillside.co.uk/arch/longmarket/comb.html
Riddler, I, 1998 ‘Combs with perforated handles’, The Archaeology of Canterbury 8, 189–198
Rogers, N S H, 1993 Anglian and Other Finds from Fishergate, The Archaeology of York 17/9,
Council for British Archaeology, London
Rytter, J, 1991 ‘Kamme fra Kungahälla’, in K Carlsson (ed), Kungahälla arkeologi 1989, Forn-
lämning 53 Ytterby socken, La?nsstryrelsen i Go?teborg och Bohus La?n, Göteborg, 39–51
Sharples, N, 2005 A Norse Farmstead in the Outer Hebrides: Excavations at Mound 3, Bornais,
South Uist, Oxbow, Oxford
Sivier, D, 2002 Anglo-Saxon and Norman Bristol, Tempus, Stroud
Skre, D and Stylegar, F-A, 2004 Kaupang: The Viking Town. The Kaupang Exhibition at UKM,
Oslo, 2004–5, University of Oslo, University Museum of Cultural History, Oslo
Smirnova, L, 2005 Comb-making in Medieval Novgorod (950–1450): An Industry in Transition,
British Archaeological Reports International Series 1369, Archaoepress, Oxford
Smith, A N, 1998 ‘Miscellaneous bone tools’, in N Sharples (ed), Scalloway: A Broch, Late Iron
Age Settlement and Medieval Cemetery in Shetland, Oxbow, Oxford, 152–159
Smith, A N, 2007 ‘Worked bone’, in J R Hunter (ed), Investigations in Sanday Orkney, Volume 1:
Excavations at Pool, Sanday: A Multi-period Settlement from Neolithic to Late Norse Times,
The Orcadian, Kirkwall, 499–514
Smith, R A, 1909 ‘Anglo-Saxon remains’, in W Page (ed), Victoria History of the Counties of
England: London 1, Constable, London, 147–170
Stummann Hansen, S, 2000 ‘Viking settlement in Shetland: chronological and regional contexts’,
Acta Archaeologica 71, 87–103
Sykes, N, White, J, Hayes, T and Palmer, M, 2006 ‘Tracking animals using strontium isotopes in
teeth: the role of fallow deer (Dama dama) in Roman Britain’, Antiquity 80, 948–959
Tempel, W-D, 1969 Die Dreilagenkämme aus Haithabu. Studien zu den Kämmen der Wikingerzeit
im Nordseeküstengebiet und Skandinavien, Unpublished dissertation, Göttingen
Ulbricht, I, 1980 ‘Middelalderlig kamproduktion i Slesvig’, Hikuin 6, 147–152
Ulbricht, I, 1984 Die Verarbeitung von Knochen, Geweih und Horn im mittelalterlichen Schleswig,
Ausgrabungen in Schleswig, Wachholtz, Neumünster
MAM 37: 18-ashby – Press

278 maritime societies of the viking and medieval world

von Holstein, I C C, Ashby, S P, van Doorn, N L, Sachs, S M, Buckley, M, Meiri, M, Barnes, I,

Brundle, A and Collins, M J, 2014 ‘Searching for Scandinavians in pre-Viking Scotland:
molecular fingerprinting of early medieval combs’, Journal of Archaeological Science 41, 1–6
Welander, R D E, Batey, C E and Cowie, T G, 1987 ‘A Viking burial fom Kneep, Uig, Isle of
Lewis’, Proceedings of the Society of Antiquaries of Scotland 117, 147–174
Wiberg, C, 1977 ‘Horn og Beinmaterialet fra ‘‘Mindets tomt’’’, in H I Høeg, H-E Lidén, A Liestøl,
P B Molaug, E Schia and C Wiberg (ed), De arkeologiske utgravninger i Gamlebyen, Oslo,
Bind 1, Alvheim and Eide, Oslo, 202–213
Wiberg, C, 1979 ‘Beinmaterialet’, in E Schia (ed), De arkeologiske utgravninger i Gamlebyen, Oslo,
Bind 2: Feltene ‘Oslogate 3 og 7’, Alvheim and Eide, Oslo, 59–63
Wiberg, T, 1987 ‘Kammer’, in E Schia (ed), De Arkeologiske utgravninger i Gamlebyen, Oslo, Bind
3, Alvheim and Eide, Oslo, 413–422
MAM 37: 19-mulville – Press

chapter 19
DEALING WITH DEER: NORSE
RESPONSES TO SCOTTISH ISLES
CERVIDS

By Jacqui Mulville

introduction
Understanding the response of Viking period raiders and settlers to novel
landscapes and to different environments is a challenge. In Atlantic Scotland, the
manner in which they assimilated, adapted or rejected local practices has been
explored through material culture, housing, domestic stock management and fishing
(eg Barrett 2003; 2012a; Barrett et al 1999; ; Bond 1998; Parker Pearson and Sharples
1999; Sharples and Smith 2009). This research expands these studies to include the
relationship between the new settlers and endemic wild terrestrial mammals within
the archipelagos of Atlantic Scotland. The predominant exploited insular wild species
were cervids, particularly red deer, and this paper examines the evidence for adapted
or enacted cervid exploitation strategies with reference to practices documented in
the Scandinavian ‘homelands’.
The Viking Age migrants who arrived and settled in the North Atlantic islands
were predominantly linked to Norway where a wide range of cervids was available to
exploit. Zooarchaeological evidence indicates that, during the Viking Age and medi-
eval periods, assemblages in southern Scandinavia were broadly characterized by a
predominance of red deer (Ashby 2007, 25). Further north elk was important
(Vretemark 1997, 200, 204), with reindeer common at higher latitudes and/or eleva-
tions (eg Hufthammer et al 2011). Cervids were utilized for their skins, fur and
antlers, with the antler of both red deer and reindeer worked to produce artefacts; for
example, large quantities of reindeer antler objects and waste were recovered at
Bergen, Oslo, Kungahälla and Trondheim (Vretemark 1998).
Cervids, and other wild species, played a comparatively minor role in Norwegian
food procurement at this time (eg Barrett et al 2007; Hufthammer 2003) and this pat-
tern is mirrored across mainland Britain. From the earliest Neolithic onwards, foods
on British sites became almost exclusively domestic in origin, with the sparse wild
faunal remains dominated by red deer antler (eg Serjeanston 2011). In riverine, coastal
and insular communities this emphasis on domestic resources was not as pronounced,
MAM 37: 19-mulville – Press

280 maritime societies of the viking and medieval world

and a rich diversity of wild terrestrial, marine and aerial fauna provided a major
resource for both food and materials — continually exploited from earliest prehistory
until the post-medieval period. Thus when the Norse arrived on the Scottish islands
they encountered a different world, farmers also hunted and gathered with exploita-
tion focused on deer (red and, to a minor extent, roe deer), marine mammals (seals,
whales and dolphins) and birds (mostly gulls and auks) (Best and Mulville 2013).
The unusual continued reliance on deer across Scottish islands has led many to
speculate on the nature of the deer–human relationship in prehistory (Armit 2006,
237; Clutton-Brock 1979; McCormick and Buckland 1997; McCormick 2007a;
Mulville 2010; Mulville and Thoms 2005; Noddle 1981; 1982; Sharples 2000; Thoms
2003). By combining the substantial data sets available from fieldwork over the past 35
years, the chronological and regional scope of this debate can now be extended to the
Norse period. By characterizing the evidence for the human:deer relationship prior to
and after Norse settlement the impact of new populations and ideas on the extant
strategies can be assessed. Were the endemic practices adapted, enhanced or replaced,
and what impact did differential access to resources such as landscapes, environment
and the availability of wild resources have? For example, by the time the Norse arrived
red deer, once relatively common, were becoming rare in the archipelago of Orkney.
The Scottish islands can be divided into four groups: the Inner Hebrides, the
Outer Hebrides (also known as the Western Isles), Orkney and Shetland (jointly
referred to as the Northern Isles). The islands lying at some distance from the British
mainland (the Outer Hebrides and the Northern Isles) have an impoverished ter-
restrial fauna as a result of a separation that has endured since the end of the last
glaciation (although there is some debate regarding when the Orcadian land bridge
flooded, eg Haynes et al 2003). Due to a paucity of early Holocene sites it is unclear
when, after the glacial retreat, deer or other terrestrial fauna first arrived on the outer
islands. Nevertheless, re-colonization must have been human-assisted (Fairnell and
Barrett 2007; Serjeantson 1990). Identifying when introduced species became estab-
lished is difficult. For example, small amounts of deer bone and antler could represent
traded joints of venison or raw materials for crafts. Nevertheless, local deer popu-
lations can be inferred from a suite of indicators: an age profile that includes calves
and juveniles, a sustainable population size and a range of body parts present on
archaeological sites. Insular populations are also known to become reduced in stature
compared to mainland populations (as a result of environmental and/or genetic
effects, see Clutton-Brock et al 1982; Grigson and Mellars 1987; Lister 1995) and the
red deer recovered from the islands demonstrate this trait (Mulville 2010). As roe deer
remains are sparse in the more distant islands, with only a few bones recorded in the
Outer Hebrides and Orkney (although more common in the Inner Hebrides), the
major focus of this paper is red deer. It is to their pre-Norse history that we now turn;
site details, including references, are listed in Table 19.1.

pre-norse deer
Red deer remains dating to the Mesolithic are recovered from mainland sites on
the west coast, such as Sand (Parks and Barrett 2009) and from the inner Hebridean
MAM 37: 19-mulville – Press

dealing with deer 281

island of Oronsay (Grigson and Mellars 1987). For the latter, metrical data suggest
two different sized groups of deer were present, one a similar stature to mainland
animals and the other smaller (Grigson and Mellars 1987, 260). This has been inter-
preted as the exploitation of two populations of deer, one from the islands reduced in
size and the other larger individuals imported from the mainland.
On the outer islands, the low visibility of Mesolithic sites (Wickham Jones 2004;
Gregory et al 2005; Lee and Woodward 2012) combined with poor bone preservation
has resulted in sparse evidence for any early faunal exploitation (Melton and
Nicholson 2007). Here the first substantial faunal assemblages are associated with the
tombs and settlements of Neolithic Orkney (Table 19.1) and the presence of both
young red deer calves and older individuals suggests a locally breeding population at
this time. The few available measurements (Figure 19.1) indicate that deer at some of
the earliest sites are comparable in size to mainland animals and represent imports,
either as joints of venison or as live translocations. The magnitude of the island deer
populations remains unclear, with many of the assemblages very different in char-
acter and the possibility of trade and movement between islands. On the Neolithic
settlements of Mainland Orkney, deer make up between 1% and 5% (Knap of
Howar, Skara Brae and Links of Noltland) of the assemblages (Table 19.1), whilst
they are of equal or higher abundance in the quantified tomb assemblages of
Pierowall (4%) and Quanterness (8%) and reported as ‘numerous’ at the Knowes of
Ramsay and Yarso, Midhowe and Blackhammer on Rousay (also burial monu-
ments). There is no evidence for the introduction of deer to Shetland at this time,
with only a few fragments of bone and antler reported from Neolithic sites such as
Scord of Brouster and little evidence for later populations. In the Western Isles, a
small assemblage of bones recovered from Northton, Harris, provides metrical
evidence for a Neolithic red deer population already diminished in size (Figure 19.1).
The much larger Irish landmass provides an interesting comparison to the
Scottish islands. Red deer are absent from Irish Mesolithic assemblages, present
solely as worked antler in the Early Neolithic with skeletal remains only appearing
from the Late Neolithic onward (Carden et al 2012). A combination of zooarchaeo-
logical and genetic analysis has suggested that Irish deer were either late natural
colonizers (Searle 2008) or were introduced by humans from mainland Britain (Carden
et al 2012). The first relatively substantial assemblage of red deer is recovered from the
exterior of the Newgrange Late Neolithic passage tomb in Co Meath. Of the full
assemblage of 12,000 bones, however, only 3% represent red deer (100 bones and
antler fragments) (van Wijngaarden-Bakker 1974, 1986). The few measurements
available from Newgrange indicate that Irish deer were similar in size to the con-
temporaneous Scottish island deer. Their small size in this extensive and possibly
wooded environment (McCormick and Murray 2007, 25), could suggest that reduc-
tions in body size occurred prior to their introduction.
In the Scottish isles the red deer populations established in the Neolithic increased
in numbers over time. By the middle of the Bronze Age substantial quantities of deer
are recovered on Hebridean and Orcadian sites (Table 19.1) and measurements indi-
cate a continual reduction in size. There are minor exceptions with some sites showing
a decline in red deer numbers, for example between the Neolithic and Bronze Age
 Table 19.1 The representation of red deer in a variety of animal bone assemblages from Atlantic Scotland 282
Island Group Site Phase Island/Area NISP* % Red Deer Reference

Mainland and Inner Hebrides

Mainland Oban Caves Mesolithic Present Kitchener et al 2004
MAM 37: 19-mulville – Press

An Corran Mesolithic Present Kitchener et al 2004

Carding Mill Bay Mesolithic Present Kitchener et al 2004
Duntroon Mesolithic Present Kitchener et al 2004
Sand Mesolithic 143 72 Parks and Barrett 2009
Crosskirk Iron Age Caithness 1879 6 McCartney 1984
Inner Hebrides Oronsay Middens Mesolithic Oronsay 168 66 Grigson and Mellars 1987
Killelan Early Bronze Age 107 12 Serjeantson et al 2005
Ardnave Bronze Age Islay 243 5 Harman 1983
Dun Mor Vaul Iron Age Tiree 1232 21 Noddle 1978–80
Dun Ardtrek Iron Age Tiree 1302 28 Noddle 1978–80
Dun Cul Bhuirg Iron Age Tiree 180 17 Noddle 1978–80
Killelan Iron Age 607 18 Serjeantson et al 2005
High Pastures Cave Iron Age Skye 1352 2 Drew 2004
Fiskavaig Iron Age Skye 4420 <1% Mulville and Ingrem 2014
Iona Vallum 7th century ad Iona 100 10 Noddle 1981
Iona Guest House Iron Age Iona 928 35 McCormick 1981
Dun Mor Vaul Norse Tiree 139 28 McCormick 2007a

Northern Isles: Orkney and Shetland

Orkeny Skara Brae Neolithic W Mainland 30169 3 Armour-Chelu 1992
Point of Cott Neolithic W Mainland 513 2 Halpin 1997
Isbister Neolithic W Mainland 601 4 Barker 1983
Pierowall Quarry Neolithic W Mainland 1809 4 McCormick 1984
Midhowe Neolithic W Mainland Present Platt 1934
Eday Neolithic W Mainland Present Platt 1937a
Blackhammer Neolithic W Mainland Present Platt 1937b
Pool Neolithic W Mainland 2065 <1 Bond 2007
Knap of Howar Late Neolithic W Mainland 6720 1 Noddle 1983
Quanterness cairn Neolithic W Mainland 222 8 Clutton-Brock 1979
Yarso Neolithic W Mainland MNI 36 Platt 1935
Knowe of Ramsay Neolithic W Mainland MNI 14 Platt 1936
Links of Noltland Late Neolithic W Mainland 8678 5 Armour-Chelu 1992
Tofts Ness Neolithic/Early Bronze Age W Mainland 1 Nicholson and Davies 2007
Point of Buckquoy Neolithic/Early Bronze Age W Mainland 188 74 Rackham 1989
Skaill, Deerness Late Bronze Age W Mainland 1084 25 Noddle 1997
maritime societies of the viking and medieval world

Tofts Ness Bronze Age W Mainland 1323 1 Nicholson and Davies 2007
Moaness, Rousay Bronze Age W Mainland 1 individual Mainland unpublished
Howe Iron Age W Mainland 1905 28 Smith 1994
Warebeth Broch Iron Age W Mainland 2098 17 Sellar 1989
Pierowall Quarry Early Iron Age W Mainland 219 2 McCormick 1984
Skaill, Deerness Iron Age W Mainland 12,623 4 Noddle 1997
 Mine Howe Iron Age W Mainland 10,849 6 Mainland and Ewens 2005;
Mainland et al 2004
Pool Late Iron Age/Pictish W Mainland 4505 1 Bond 2007
Pool Pictish/Viking W Mainland 13,535 1 Bond 2007
Buckquoy Norse (9th–12th C ad) W Mainland 2738 Absent Noddle 1976–77
Brough Road (areas 1 & 2) Viking W Mainland 259 Absent Rackham 1989
MAM 37: 19-mulville – Press

Skaill, Deerness Viking/Post Viking W Mainland 5490 1 Noddle 1997

Brough of Deerness Late Norse/medieval W Mainland 253 Absent Rackham 1986
Earl’s Bu Late Norse/medieval W Mainland 1263 1 Mainland 1995
Shetland Jarlshof Prehistoric Sumburgh, Mainland Absent Platt 1956
Scord of Brouster Neolithic Walls, Mainland Antler only Noddle 1986
Scalloway Iron Age Scalloway, Mainland 4271 Absent O’Sullivan 1998
Scatness Iron Age Sumburgh, Mainland 641 Absent Bond pers. comm.
Jarlshof Norse Sumburgh, Mainland Present Platt 1956
Outer Hebrides
Lewis/Harris Northton Neolithic Harris 608 4 Finlay 1984
Northton Beaker Harris 726 29 Finlay 1984
Northton Iron Age Harris 393 57 Finlay 1984
Berie Iron Age Lewis 1878 32 Thoms 2003
Cnip Iron Age Lewis 1489 25 McCormick 2007a
Bostadh Iron Age Lewis 351 8 Thoms 2003
Bostadh Norse Lewis 1511 24 Thoms 2003
The Uists Udal Neolithic N Uist 358 3 Finlay 1984
Udal Early Bronze Age N Uist 277 Absent Finlay 1984
Udal Iron Age N Uist 2399 Absent Finlay 1984
Udal North Iron Age N Uist 5504 1 Serjeantson 2013
Sollas Iron Age N Uist 812 1 Finlay 1984
Rosinish Beaker N Uist 199 21 Finlay 1984
Baleshare Iron Age Benbecula 2040 1 Halstead 2003
dealing with deer

Cladh Hallen Late Bronze Age/Iron Age S Uist 889 7 Mulville and Powell
forthcoming
Hornish Point Late Bronze Age/Iron Age S Uist 440 1 Halstead 2003
A’Cheardach Bheag Iron Age S Uist 68 13 Finlay 1984
A’Cheardach Mhor Iron Age S Uist 223 2 Finlay 1984
Cille Donnain III Iron Age S Uist 5131 1 Vickers et al 2014
Dun Vulan Iron Age S Uist 2882 1 Mulville 1999
Bornais Iron Age S Uist 2890 16 Mulville 2005
Pabbay Iron Age Pabbay 2274 3 Mulville and Ingrem 2000
Dunan Ruadh Iron Age Barra 1423 1 Mulville and Ingrem 2000
Sandray Iron Age Sandray 144 Absent Mulville and Ingrem 2000
Mingulay Iron Age Mingulay 416 2 Mulville and Ingrem 2000
Udal North Viking N Uist 1942 1 Serjeantson 2013
Bornais Norse S Uist 13,243 6 Mulville 2005; Mulville and
Powell 2012
Cille Pheadair Norse S Uist 7115 3 Mulville forthcoming
283

* Number of Identified Specimens of cattle, sheep, pig and red deer bone (antler has been excluded where possible)
MAM 37: 19-mulville – Press

284 maritime societies of the viking and medieval world

Figure 19.1 The means and ranges for two measurements on deer bones, calculated for
samples with more than five values. If only a square is present this represents a single
measurement. The values run from mainland sites, to Inner Hebridean sites, to the Northern Isles
and finally the Western Isles. Chronology is indicated by: (M) Mesolithic, (N) Neolithic,
(BA) Bronze Age, (IA) Iron Age, (Nor) Viking Age and/or medieval ‘Norse’ (J Mulville)
MAM 37: 19-mulville – Press

dealing with deer 285

deposits at the multi-period site of Tofts Nest on the small island of Sanday, Orkney.
This decline is not seen at the contemporaneous Orcadian sites of the Bay of Skaill and
Buckquoy where red deer dominate. In the Outer and Inner Hebrides the numbers of
sites with deer, and the proportion of deer, also generally increases, although again
individual sites differ, for example at the Early Bronze Age phases on the Udal, Uist,
no red deer bone was recovered. Deer continue to be rarely exploited in Ireland at this
time; at Late Bronze Age Haughey’s Fort (Co Armagh) only six fragments of bone and
antler were found in an assemblage of over 700 bones (Mallory et al 1999).
By the beginning of the Iron Age, deer were present on the majority of Orcadian
sites, but in variable amounts with evidence of a decline in numbers over time (Table
19.1). The large Iron Age assemblage at the broch settlement of Howe, Mainland,
clearly illustrates this; deer make up 35% of the Early Iron Age assemblage (around
200 bc), but this figure falls to only 4% by the Late Iron Age. Once again there are
exceptions, with deer becoming most numerous during the Late Iron Age and Pictish
periods at Pool, on the island of Sanday, Orkney (Table 19.1). In Shetland, red deer
are not present in the large Iron Age assemblages at Scalloway and Scatness, re-
affirming the lack of a local deer population.
On the Inner Hebrides, the proportion of red deer increases and they account for
up to one-fifth of assemblages in (pre-Viking) Late Iron Age deposits. Unusually, roe
deer are also present in large numbers at Dun Mor Vaul (up to 30%) on Tiree and at
Dun Ardtreck (7%) on Skye (Noddle 1974). For the Outer Hebrides, differences in
the intensity of deer exploitation between the northerly landmass of Lewis/Harris and
the southerly islands that make up the Uists emerge. On the Bhaltos peninsula of
north-west Lewis, red deer make up one-quarter to one-third of the larger terrestrial
mammal assemblages at Cnip, but only one-tenth of the assemblage at Bostadh on
the nearby island of Great Bernera. Deer were superabundant in the small assem-
blage at Northton, south Lewis, but this quantification includes numerous fragments
of antler that are excluded from other assessments. These sites can be contrasted
with those on the Uists where, although deer are reported from all sites across the
islands, they generally make up a small proportion of the assemblages (1–2%).
Exceptions to this pattern are the pre-Viking settlement at Late Iron Age Bornais and
the small assemblage within the roundhouse at A’ Cheardach Bheag, where a hearth
surround was constructed of deer mandibles. In Ireland the few late Bronze Age and
Iron Age sites with bone assemblages continue to contain small quantities of deer. At
Dún Aillinne (Co Kildare), a Bronze to Early Iron Age settlement, red deer were
represented by only three items out of an assemblage of over 2400 specimens.

viking age and medieval deer

Thus by the time the Norse began to disperse into the Scottish island groups,
around or after ad 800, they were faced with variable situations with regard to red
deer populations and their utilization. During the Norse period red deer were rarely
procured on Orkney, accounting for less than 1% of all assemblages, and there is
little evidence for an endemic breeding population. This pattern is noted across the
island group and includes sites on the largest island, Mainland, where populations
MAM 37: 19-mulville – Press

286 maritime societies of the viking and medieval world

were previously substantial (for example, at Earl’s Bu and Skaill in Deerness), as well
as on the smaller islands. At Pool on Sanday deer numbers declined during the
Pictish/Norse interface, dwindled further during the Viking phase, then persisted but
in very low numbers to the end of the 11th century (Bond 2007, 214). Elsewhere, at
Viking Age and medieval Quoygrew, on Westray, almost no red deer remains were
recovered (Harland 2012).
This lack of deer on Orkney is referred to in Orkneyinga saga, with the earls of
Orkney travelling to Caithness to hunt (Palsson and Edwards 1978, 209) and the
complete absence of red deer remains at Earl’s Bu, an earl’s residence, attests to the
loss of local herds. In Shetland, rare finds of deer bones were reported from Viking
Age Jarlshof (Platt 1956, 212–215), but were not present at medieval Sandwick
(Bigelow 1985; Smith 2008), suggesting that the former may represent traded items.
Unlike Orkney, the Outer Hebridean populations of red deer did not demon-
strate a steep decline during the Iron Age, and they continued to be exploited,
although in small proportions, on both Viking Age and medieval Norse settlements.
There are a number of sites that show continuity into the Norse period, but it is only
at Bostadh, on Lewis, that deer relative abundance increased. At the Udal on North
Uist deer remained a small but persistent presence whilst at Bornais, on South Uist,
deer numbers halved but remained significant. At the newly established settlement of
Cille Pheadair, only a few kilometres south of Bornais, red deer were also exploited in
much smaller numbers. The only published Inner Hebridean Norse material comes
from Dun Mor Vaul, Tiree, and deer become predominant in the assemblage at this
time (McCormick 2007a, 164; Noddle 1974).
The varying proportions of red deer have been linked to environmental con-
straints. McCormick (2007a, 163, 165) considers the rich agricultural potential of the
extensive Western Isles machair to be a factor in reducing the proportions of hunted
species on Uist and the lack of good agricultural land around the sites of Cnip and
Northton on Lewis to focus exploitation on the extensive Lewis uplands and higher
quantities of deer. In Orkney, deer abundance has been linked to island size, with
their presence on smaller islands thought to reflect inter-island trade (eg Bond 2007).
These links between land area, quality and cervid exploitation are difficult to recon-
cile with the large number of red and roe deer reported from sites on Tiree in the
Inner Hebrides. This relatively small and woodless island has extensive machair and
high agricultural potential (McCormick 2007, 163), and less need to exploit upland
areas for wild food.
It is also apparent that deer body size does not reflect the overall availability of
land. Insular deer are generally smaller than mainland deer, but those on the larger
Outer Hebridean land mass are on average, smaller (between two-thirds and one-half
the size of mainland deer) than earlier and contemporary animals found in Orkney
(average size reduction 15%) or the Inner Hebrides (average size reduction 10 to
20%). The movement of larger mainland animals to the Inner Hebrides may have
boosted the size of the latter, but the persistently greater size of Orcadian deer up
until their extirpation is perplexing. The smaller Orcadian land mass (at 990km2 less
than one-third of either the Inner or Outer Hebrides) should have exacerbated
MAM 37: 19-mulville – Press

dealing with deer 287

the effects of an increasing human population and greater competition with domestic
crops and animals causing a decrease in deer size. The body size of Orcadian deer
could be a result of the characteristics of foundation animals (ie larger individ-
uals introduced in the Neolithic), a faster expansion in initial populations, re-
introductions throughout prehistory, differences in plant ecology and nutrition and
ultimately the eradication of deer before they became further reduced in size. This
interplay between phenotypic and genotypic effects in the North Atlantic islands is
the focus of on-going research exploring insular cervid genetic and metrical histories.
So what role did red deer play on these islands throughout the Viking Age and
medieval Norse settler phases? In Orkney, deer populations were either small or
absent and archaeological evidence suggests a very minor role for any deer products
at this time. On the Western Isles, deer populations, and their exploitation for meat
as well as other products, persisted throughout the Iron Age and with the arrival of
the Norse new divergent strategies emerged. At Bostadh, on Lewis, the emphasis
shifted towards the procurement of fully mature red deer with only a small amount of
bone coming from sub-adults. On the Uists, the site of Bornais demonstrated
continuity with Iron Age practices with a bimodal age pattern of red deer exploita-
tion — but the focus shifted to younger animals with the result that nearly one-third
of the harvested individuals were new-borns (less than one month old). This targeting
of neonates must have occurred during June/July, when suckling deer calves could
have been gathered from their hiding places whilst their mothers foraged elsewhere.
At the newly founded settlement of Cille Pheadair small numbers of neonatal deer
were also represented, but the majority of animals exploited were mature. This per-
sistent use of red deer, particularly in the Western Isles, up until the medieval period
indicates that Scottish insular communities, including the colonizing Norse, con-
structed and maintained a relationship with deer that was at odds with most post-
Neolithic communities across Britain and Ireland.

different ways with deer

As already noted, the degree of wild food exploitation seen on Scottish islands is
unequalled in British mainland assemblages (Mulville 2010). However, the signifi-
cance of red deer goes beyond unusually high levels of exploitation for food. They
had a role in defining and redefining place (Jones 1998, 315, 316; Mulville and Thoms
2005; Sharples 2000). Moreover, throughout time they were subject to unusual age-
selection strategies that differ from expected hunting patterns. In Neolithic Orkney,
for example, foetal/neonatal deer were placed in tombs (eg Quanterness Cairn:
Clutton-Brock 1979, 119), young deer were also present in settlements (eg Pool on
Sanday), with 15, mostly female, yearling deer deposited unbutchered at the Links
of Noltland (Amour-Chelu 1992). On the Uists in the Outer Hebrides there was a
focus on neonatal deer at the Late Bronze Age to Early Iron Age settlement of Cladh
Hallan, and very young animals were exploited throughout the Iron Age at Dun
Vulan (Mulville 1999) and Bornais (Mulville 2005). These practices were continued at
the Norse sites of Bornais and Cille Pheadair. Furthermore, deer uniquely appear as
MAM 37: 19-mulville – Press

288 maritime societies of the viking and medieval world

animal motifs on Hebridean pottery and other objects, such as the carved head on a
wooden handle from Dun Bharavat, Lewis (Armit 2006, 239 ill 7.10). As noted above,
the significance of deer is also highlighted in the use of 17 adult mandibles as a hearth
surround at A’Cheardach Bheag in the Uists (Mulville and Thoms 2005).
Traditionally, zooarchaeologists take the presence of neonatal animals to indi-
cate on-site breeding, domestication and intensive exploitation. A high proportion of
cattle neonates is relatively common in prehistoric Hebridean assemblages and has
been linked to milk production and/or poor husbandry conditions (Craig et al 2005;
Mulville et al 2005; Serjeantson and Bond 2007; Serjeanston 2013, 50). However, the
appearance of large numbers of neonates of any wild species is highly unusual and
reviews of prehistoric European red deer mortality have revealed few parallels
(Carter 2001a; 2001b; Steele 2004). At Ertebølle Ringkloster, Denmark, the inclusion
of very young deer calves was interpreted as reliance on seasonal exploitation with
a focus on skins and fur (Rowley-Conwy 1994–95) whilst in Bosnia-Herzegovina,
where 2% of the population was foetal rather than neonatal, it was concluded that
pregnant females were hunted over the winter (Miracle and O’Brien 1998). At Meso-
lithic Star Carr, around half of all red deer mandibles were from sub-adults, but
neonates were rare, whilst at Thatcham neonates were absent (Carter 1997; 1998).
The exploitation of very young animals by human populations could be a product of
seasonal hardship, but as neonatal deaths occurred in early summer other foods
would have been available.
During the Viking Age and Middle Ages the island strategy of slaughtering neo-
natal cattle intensified, suggesting increased dairy production (Bond 2007, 221–225;
Harland 2007; 2012; Mulville et al 2005; Serjeantson and Bond 2007, 202–206,
Serjeantson 2013). Gadid fishing became more important and new herring fisheries
were established (Barrett 2012a; Barrett et al 1999; Ingrem 2005). At the same time the
exploitation of extant red deer populations continued and on the Uists in particular
there was an increase in the strategy of targeting young deer. There is no evidence for
similar strategies of neonatal deer exploitation in the Scandinavian homelands and
the increase in neonates of wild species at this time remains an enigma. The native
islanders’ solutions to the challenges provided by the local environment had been suc-
cessfully met for many centuries, and the new Norse settlers appear to have both
assimilated and augmented these techniques.
The presence of red deer neonates in special contexts, such as the main burial
chamber at Quanterness and a well at Howe (Smith 1994, 145), implies that food
scarcity was not the only force driving this exploitation. A number of authors have
argued that the introduction of deer to the islands, their persistence in restricted
environments and the on-site presence of neonates all indicate close management and
possibly the culling of very young deer to restrict herd size (Clutton-Brock 1979, 113;
McCormick 2007a, 163–164; Noddle 1981; Sharples 2000; Thoms 2003). In the light
of milk-production models, for domestic species at least, being linked to neonatal
slaughter (Mulville et al 2005) it is timely to reconsider the human–deer relationship.
Examples of alternative relationships with cervids are found among the reindeer
herders of the circumpolar region, for example the Sami people of Fennoscandia and
MAM 37: 19-mulville – Press

dealing with deer 289

the Kola Peninsula (eg Ingold 1980) and the Olenok Evenk people of Siberia (Grøn
and Turov 2007) who farm and fish in addition to hunting, herding and milking
reindeer. The antiquity of reindeer-herding culture remains under debate, but pre-
cedes the Scandinavian settlement of Atlantic Scotland (eg Andersen 2011, Bergman
et al 2008). Domestic reindeer are marginally tamed and identified as property,
but their breeding is not fully controlled. Reindeer are exploited to varying degrees
from the hunting of wild animals (often using tame decoys) to the herding, milking
and pack animal usage of ‘domesticated’ deer. Even within one group, both hunter-
gatherers (hunting mainly elk, with reindeer used only for transport), and pastoralists
who milk deer can co-exist. Both groups only eat their ‘domesticated’ deer in special
rituals or extreme economic circumstances and these animals may be exposed when
they die (often hung in a bag from a large tripod), rather than consumed (Grøn and
Turov 2007, 69). Only small amounts of the very fatty milk is used, in the summer as
a luxury (Grøn and Turov 2007), and there is no evidence for intensive milking allied
to deliberate neonatal slaughter (Grøn pers comm). However, there are problems in
drawing direct comparisons between red deer management and exploitation to that
of reindeer as the two species differ in both behaviour and physiology.
In Norway, Sami products were valuable to Norse chieftains in their external
trade and were integrated into chiefly economies (Bergman et al 2008, 97). Thus
whilst deer exploitation and management appear to be island innovations, the Norse
settlers may have already been familiar with a wide range of human:cervid interac-
tions. Research indicates that deer held a central role in the insular communities of
Scotland over thousands of years, and the range of ethnographically attested inter-
actions possible with reindeer also provides a useful tool in constructing new
accounts of human:red deer relations. For example, it may be possible to reconsider
the significance of the deer mandible hearth surround at A’Cheardach Bheag in the
light of the Evenk tradition that the heads of reindeer are considered sacred, and in
particular the animal’s soul is said to sit in its mandible (Grøn 2005).

the importance of antler

Thus far this paper has focused on the role of deer as a food species, looking
only at the bone recovered from sites, but throughout prehistory and history antler
was a vital resource for the production of artefacts (eg Hallen 1994). This section
considers the Norse response to the varying access to antler as a function of red deer
availability and compares and contrasts the strategies employed for procuring antler
and producing artefacts on islands with differing access to deer. Whilst the earls of
Orkney crossed the sea to hunt red deer, how was the production of essential items
made from antler, such as composite combs, sustained? Composite combs were com-
mon on sites from the Late Iron Age through to the medieval period across northern
Europe and their culturally distinctive styles and forms makes them an excellent
resource for understanding culture contact. In the case of the Scottish Isles they
provide evidence of Scandinavian imports, influence and local innovation (Ashby
2007; 2009; von Holstein et al 2014; see Chapter 18).
MAM 37: 19-mulville – Press

290 maritime societies of the viking and medieval world

Typical (pre-Viking Age) Late Iron Age composite combs have been recovered
from Orcadian, Shetlandic and Hebridean sites (Ashby 2007, 108; 2009). Artefact-
production techniques and materials can provide evidence regarding antler avail-
ability, for example antler waste material at Late Iron Age Bostadh in the Outer
Hebrides provides evidence for comb-making and access to red deer resources.
Early Viking Age combs (Ashby Type 5) are found across Atlantic Scotland and
provide evidence of contact with Scandinavia from the mid-9th century ad onwards
(2007, 231). Later, by the 12th to 13th centuries, the types of combs present and the
evidence for their production separates the Atlantic fringe into distinct northern and
western zones. The northern zone, taking in Norway and the Northern Isles, is
characterized by the use of two types of comb (Ashby 9 and 13), frequently made of
reindeer antler and invariably fixed with copper alloy rivets (Ashby 2007, 236). Whilst
these combs are recovered from Orkney, it is unclear if local comb production was
occurring. The identification of a very small number of blanks (Ashby 2007) could
indicate production on an incidental scale. Given the small quantity of deer, and
therefore antler, in Orkney at this time, and the paucity of comb-making waste, it is
likely that most combs were produced in Norway and imported to the islands (Ashby
and Batey 2012). Barrett (2012b, 275) goes so far as to suggest that the last remaining
Orcadian deer were quickly hunted out during the Viking Age and Middle Ages as
antler combs were easily acquired from Scandinavia.
A western zone, incorporating England, Ireland and western Scotland, is char-
acterized by combs made of red deer antler (type 8), the raw material and style of
riveting suggest local production (Ashby 2007, 232). Perplexingly, the finds from
Bostadh, Lewis, in the Outer Hebrides indicate that Iron Age comb manufacture did
not continue into the Norse period, despite evidence for increased exploitation of
local red deer as a food source. Further south, on the Uists, Bornais provides sub-
stantial evidence for comb manufacture even until the late 13th century and is the
only major British comb-production site known outside an urban context (Sharples
2005). Here antler was pivotal as a resource with both Type 8 (western zone) and
type 13 (northern zone) combs, with iron rather than copper alloy rivets, manu-
factured at the site.
The evidence from the islands demonstrates that the relationship between red
deer abundance, antler availability and comb production can be complex. In the case
of Orkney, the lack of deer, style of combs, absence of waste, presence of reindeer
antler (Ashby 2009; Ashby and Batey 2012; von Holstein et al 2014) and good trading
links to Norway all support importation. The need for antler artefacts appears to
have been overcome, not by a trade in the resource, but by a trade in finished
products. In the Western Isles, deer are present and comb production indicates that
local antler was exploited, at least on the Uists. However, access to red deer herds did
not result in comb production at Bostadh, nor at Cille Pheadair and other materials
were imported to produce Norse artefacts; walrus ivory has been recovered from
Bornais (see Chapter 17) suggesting that the availability of a resource was not the
only driver of local production. At a minimum, the availability of craftsmen to pro-
duce combs locally and their level of specialism would also affect production strate-
gies and styles (Ashby 2007, 33; Chapter 18).
MAM 37: 19-mulville – Press

dealing with deer 291

living with and without deer

This paper has provided an account of red deer exploitation both prior to and
during Norse arrival and occupation in Atlantic Scotland. As noted above, the
inhabitants of the Scottish isles developed particular adaptations to insular life, inter-
acting with their environment through a reliance on wild resources and developing a
close and complex relationship with cervids. This relationship petered out in Orkney
by the time the Norse arrived, but continued in the Hebrides, and both situations
would have provided novel situations for Norse immigrants. On one hand, the few
Orcadian deer were quickly extirpated, leaving little or no antler for comb produc-
tion, and quarry for hunting only available on adjacent landmasses. On the other
hand, there were plenty of Hebridean deer and a pre-existing indigenous tradition of
close management and comb production.
The Hebridean reliance on deer, which continued into the Viking Age and
Middle Ages, made cervids an intimate part of insular lifeways. Deer would have been
perceived in a number of ways — from competitors with domestic stock and con-
sumers of crops, to valuable exploiters of less-productive land, to providers of antler
— a valuable, renewable, workable resource. The pre-existing special relationship
between humans and deer, across the islands, is demonstrated by the placement of
their remains in tombs, in wells, in houses (with mandibles protecting hearths) and in
post holes, in addition to their depictions in art (Mulville and Thoms 2005). The
parallels between their exploitation and that of cattle — both were prey to culls of
their young, required management to protect crops and inhabited similar upland
areas for part of the year — suggest a similar and close relationship with elements of
‘domestication’. This relationship is highly unusual when viewed from the British
mainland, but may have been less unusual to Viking Age Scandinavian incomers. The
reindeer-using Sami were an integral part of the Viking Age Scandinavian world
(eg Price 2002) and reindeer harvesting was also part of Norse Fennoscandian tradi-
tion (eg Bang-Andersen 2013). The nature of the relationships between humans and
deer out on the islands would, of course, have been very different, but the level of
complexity seen in circumpolar reindeer relations provides a good parallel for the
islanders’ attitudes to ‘wild’ cervids. Scandinavian migrants may well have recognized,
understood and easily assimilated the islanders’ close relationship with red deer.
The story of the island deer is without parallel in British archaeology, and
gaining an understanding of how and why this relationship with humans developed
and persisted is a challenge. To then attempt to comprehend how this novel situation
was perceived and integrated by past settlers adds further complexity. This paper has
demonstrated that there are useful contrasts and parallels that can be drawn, and that
by considering both domestic and wild animals a deeper understanding of Norse
lifeways can be reached.

references
Andersen, O, 2011 ‘Reindeer-herding cultures in northern Nordland, Norway: methods for
documenting traces of reindeer herders in the landscape and for dating reindeer-herding
activities’, Quaternary International 238(1), 63–75
MAM 37: 19-mulville – Press

292 maritime societies of the viking and medieval world

Armit, I (ed) 2006 Anatomy of an Iron Age Roundhouse: The Cnip Wheelhouse Excavations,
Lewis, Society of Antiquaries of Scotland, Edinburgh
Armour-Chelu, J M, 1992 Neolithic Britain, with special reference to the sites of Links of Noltland,
Etton, and Maiden Castle, Unpublished PhD thesis, University College London Institute of
Archaeology and The British Museum (Natural History), London
Ashby, S P, 2007 Time, trade and identity: bone and antler combs in northern Britain c
ad 700–1400, Unpublished PhD thesis, University of York, York
Ashby, S P, 2009 ‘Combs, contact and chronology: reconsidering hair combs in early-historic and
Viking-Age Atlantic Scotland’, Medieval Archaeology 53, 1–33
Ashby, S P and Batey, C E, 2012 ‘Evidence of exchange networks: the combs and other worked
skeletal material’, in J H Barrett (ed), Being an Islander: Production and Identity at
Quoygrew, Orkney, ad 900–1600, McDonald Institute for Archaeological Research,
Cambridge, 229–244
Bang-Andersen, S, 2013 ‘Prehistoric reindeer hunting in south-west Norway with emphasis on the
period 1000 bc to ad 1000: overview, retrospect and perspective’, in O Grimm and
U Schmölcke (ed), Hunting in Northern Europe until 1500 ad: Old Traditions and Regional
Developments, Continental Sources and Continental Influences, Wachholtz Verlag, Neu-
münster, 41–55
Barker, G, 1983 ‘The animal bones’, in J W Hedges (ed), Isbister: A Chambered Tomb in Orkney,
British Archaeological Reports British Series 115, Archaeopress, Oxford, 133–150
Barrett, J H, 2003 ‘Culture contact in Viking Age Scotland’, in J H Barrett (ed), Contact, Con-
tinuity and Collapse: The Norse Colonization of the North Atlantic, Brepols, Turnhout,
73–111
Barrett, J H (ed), 2012a Being an Islander: Production and Identity at Quoygrew, Orkney,
ad 900–1600, McDonald Institute for Archaeological Research, Cambridge
Barrett, J H, 2012b ‘Being an Islander’, in J H Barrett (ed), Being an Islander: Production and
Identity at Quoygrew, Orkney, ad 900–1600, McDonald Institute for Archaeological
Research, Cambridge, 275–291
Barrett, J, Nicholson, R A and Ceron-Carassco, R, 1999 ‘Archaeo-icthyological evidence for long
term socioeconomic trends in northern Scotland: 3500 bc to ad 1500’, Journal of Archaeo-
logical Science 26, 353–388
Barrett, J, Hall, A, Johnstone, C, Kenward, H, O’Connor, T and Ashby, S, 2007 ‘Interpreting the
plant and animal remains from Viking-age Kaupang’, in D Skre (ed), Kaupang in Skiringssal,
Aarhus University Press, Aarhus, 283–319
Bergman, I, Liedgren, L, Östlund, L and Zackrisson, O, 2008 ‘Kinship and settlements: Sami
residence patterns in the Fennoscandian Alpine regions about ad 1000’, Arctic Anthropology
45, 97–100
Best, J and Mulville, J, 2013 ‘Between the sea and sky: the archaeology of avian resource exploita-
tion in Scottish island environments’, in M-Y Daire, C Dupont, A Baudry, C Billard, J-M
Large, L Lespez, E Normand and C Scarre (ed), Anciens Peuplements Littoraux Et Relations
Homme/Milieu Sur Les Cotes De L’Europe Atlantique / Ancient Maritime Communities and
the Relationship Between People and Environment along the European Atlantic Coasts,
British Archaeological Reports International Series 2570, Archaeopress, Oxford, 417–426
Bigelow, G F, 1985 ‘Sandwick Unst and the Late Norse Shetlandic economy’, in B Smith (ed),
Shetland Archaeology: New Work in Shetland in the 1970’s, Shetland Times Press, Lerwick,
95–127
Bond, J M, 1998 ‘Beyond the fringe? Recognising change and adaptation in Pictish and Norse
Orkney’, in C Mills and G Coles (ed), Life on the Edge: Human Settlement and Marginality,
Oxbow Books, Oxford, 81–90
MAM 37: 19-mulville – Press

dealing with deer 293

Bond, J M, 2007 ‘The faunal remains from Pool’, in J Hunter, J M Bond and A N Smith (ed),
Investigations in Sanday, Orkney, Volume 1: Excavations at Pool, Sanday, a Multi-period
Settlement from Neolithic to Late Norse Times, The Orcadian/Historic Scotland, Kirkwall,
207–262
Carden, R F, McDevitt, A D, Zachos, F E, Woodman, P C, O’Toole, P, Rose, H, Monaghan, N,
Campana, M G, Bradley, D G and Edwards, C J, 2012 ‘Phylogeographic, ancient DNA, fossil
and morphometric analyses reveal ancient and modern introductions of a large mammal: the
complex case of red deer (Cervus elaphus) in Ireland’, Quaternary Science Reviews 42, 74–84
Carter, R J, 1997 ‘Age estimation of the roe deer (Capreolus capreolus) mandibles from the
Mesolithic site of Star Carr, Yorkshire, based on radiographs of mandibular tooth develop-
ment’, Journal of Zoology 241, 495–502
Carter, R J, 1998 ‘Reassessment of seasonality at the Early Mesolithic Site of Star Carr, Yorkshire
based on radiographs of mandibular tooth development in red deer (Cervus elaphus)’, Journal
of Archaeological Science 25, 851–856
Carter, R J, 2001a ‘Dental indicators of seasonal human presence at the Danish boreal sites of
Holmegaard I, IV,V and Mullerup and the Atlantic sites of Tybrind Vig and Ringkloster’, The
Holocene 11, 359–365
Carter, R J, 2001b ‘New evidence for seasonal human presence at the Early Mesolithic site of
Thatcham, Berkshire, England’, Journal of Archaeological Science 28, 1055–60
Clutton-Brock, J, 1979 ‘Report of the mammalian remains other than rodents from Quanterness’,
in A C Renfrew (ed), Investigations in Orkney, Society of Antiquaries of London, London,
112–134
Clutton-Brock, J, Guinness, F E and Albon, S D, 1982 Red Deer: Behavior and Ecology of Two
Sexes, Edinburgh University Press, Edinburgh
Craig, O E, Taylor, G, Mulville, J, Collins M J and Parker Pearson, M, 2005 ‘The identification of
prehistoric dairying activities in the Western Isles of Scotland: an integrated biomolecular
approach’, Journal of Archaeological Science 32, 91–103
Drew, C, 2004 in S Birch, M Wildgoose and G Kozikowski (ed), Uamh an Ard Achadh (High
Pasture Cave) Strath, Isle of Skye 2004 (NGR NG 5943 1971): the preliminary assessment and
analysis of late prehistoric cultural deposits from a limestone cave and associated surface
features, West Coast Archaeological Services, Isle of Skye, 56–59
Fairnell, E H and Barrett, J H, 2007 ‘Fur-bearing species and Scottish islands’, Journal of
Archaeological Science 34, 463–484
Finlay, J, 1984 Faunal evidence for prehistoric economy and settlement in the Outer Hebrides to
ad 400, Unpublished PhD thesis, University of Edinburgh, Edinburgh
Gregory, R A, Murphy, E M, Church, M J, Edwards, K J, Guttmann, E B and Simpson, D D, 2005
‘Archaeological evidence for the first Mesolithic occupation of the Western Isles of Scotland’,
The Holocene 15, 944–950
Grigson, C and Mellars, P, 1987 ‘The mammalian remains from the middens’, in P A Mellars (ed),
Excavations on Oronsay: Prehistoric Human Ecology on a Small Island, Edinburgh University
Press, Edinburgh, 243–289
Grøn, O, 2005 ‘A Siberian perspective on the north European Hamburgian Culture: a study in
applied hunter-gatherer ethnoarchaeology’, Before Farming 2005(1), article 3
Grøn, O and Turov, M, 2007 ‘Resource pooling and resource management’, in B Hårdh,
K Jennbert and D Olausson (ed), On the Road: Studies in Honour of Lars Larsson, Almqvist
& Wiksell International, Stockholm, 67–72
Hallen, Y, 1994 ‘The use of bone and antler at Foshigarry and Bac Mhic Connain, two Iron Age
sites on North Uist, Western Isles’, Proceedings of the Society of Antiquaries of Scotland 124,
189–231
MAM 37: 19-mulville – Press

294 maritime societies of the viking and medieval world

Halpin, E, 1997 ‘Animal, bird and fish bone’, in J Barber (ed), The Excavation of a Stalled Cairn at
Point of Cott Westray, Orkney, Scottish Trust for Archaeological Research, Edinburgh
Halstead, P, 2003 ‘The animal bones from Baleshare and Hornish Point’, in J Barber (ed), Bronze
Age Farms and Iron Age Farm Mounds of the Outer Hebrides, Society of Antiquaries of
Scotland, Edinburgh, 142–148
Harland, J, 2007 Zooarchaeology in the Viking Age to medieval Northern Isles, Scotland, Unpub-
lished PhD thesis, University of York, York
Harland, J F, 2012 ‘Animal husbandry: the mammal bone’, in J H Barrett (ed), Being an Islander:
Production and Identity at Quoygrew, Orkney, ad 900–1600, McDonald Institute for
Archaeological Research, Cambridge, 139–154
Harman, M, 1983 ‘Animal remains from Ardnave, Islay’, in G Ritchie and H Welfare (ed), ‘Exca-
vations at Ardnave, Islay’, Proceedings of the Society of Antiquaries of Scotland 113, 302–336
Haynes, S, Jaarola, M and Searle, J B, 2003 ‘Phylogeography of the common vole (Microtus
arvalis) with particular emphasis on the colonization of the Orkney archipelago’, Molecular
Ecology 12, 951–956
Hufthammer, A K, 2003 ‘Med kjøtt og fisk på menyen’, in O Skevik (ed), Middelalder-Gården i
Trøndelag, Stiklestad Nasjonale Kultursenter, Verdal, 182–196
Hufthammer, A K, Bratbak, O F and Indrelid, S, 2011 ‘A study of bone remains and butchery
patterns from medieval mass-hunting of reindeer in the south Norwegian mountain districts’,
Quaternary International 238, 55–62
Ingold, T, 1980 Hunters, Pastoralists and Ranchers: Reindeer Economies and Their Transforma-
tions, Cambridge University Press, Cambridge
Ingrem, C, 2005 ‘The fish remains’, in N Sharples (ed), A Norse Farmstead in the Outer Hebrides:
Excavations at Mound 3, Bornais, South Uist, Oxbow Books, Oxford, 157–158
Jones, A, 1998 ‘Where Eagles Dare: landscape, animals and the Neolithic of Orkney’, Journal of
Material Culture 3, 301–324
Kitchener, A C, Bonsall, C and Bartosiewicz, L, 2004 ‘Missing mammals from Mesolithic middens:
a comparison of the fossil and archaeological records from Scotland’, in A Saville (ed),
Mesolithic Scotland and its Neighbours: The Early Holocene Prehistory of Scotland, its
British and Irish Context, and some Northern European Perspectives, Society of Antiquaries
of Scotland, Edinburgh, 73–82
Lee, D and Woodward, N L, 2012 Mesolithic occupation at Links House, Stronsay, Orkney
c 8000 BP: the first structural evidence for the Northern Isles of Scotland, Unpublished report,
ORCA, Kirkwall
Lister, A M, 1995 ‘Sea-levels and the evolution of island endemics: the dwarf red deer of Jersey’, in
R C Preece (ed), Island Britain: A Quaternary Perspective, Geological Society, London,
151–172
Mainland, I, 1995 A preliminary discussion of the animal bone assemblage from the 1979–1993
excavation at the Earl’s Bu, Orphir, Orkney, Unpublished report, University of Bradford,
Bradford
Mainland, I, nd The mammalian, avian and amphibian remains from Moaness, Unpublished
report, University of Bradford, Bradford
Mainland, I and Ewens, V, 2005 A preliminary report on the hand collected mammal bone assem-
blages from 2003 excavations at the Iron Age site of Mine Howe, Tankerness, Orkney,
Unpublished report, University of Bradford, Bradford
Mainland, I, Ewens, V and Davis, G 2004, A preliminary report on the hand collected mammal
bone assemblages from 2000 and 2002 excavations at the Iron Age site of Mine Howe,
Tankerness, Unpublished report, University of Bradford, Bradford
MAM 37: 19-mulville – Press

dealing with deer 295

Mallory, J P, Brown, D M and Baillie, M G L 1999, ‘Dating Navan Fort’, Antiquity 73, 427–431
McCartney, E, 1984 ‘Analysis of faunal remains’, in H Fairhurst (ed), Excavations at Crosskirk
Broch Caithness, Society of Antiquaries of Scotland, Edinburgh, 133–147
McCormick, F, 1981 ‘The animal bones from Ditch 1’, in J Barber (ed), ‘Excavations in Iona’,
Proceedings of the Society of Antiquaries of Scotland 111, 282–380
McCormick, F, 1984 ‘Large mammal bone’, in N Sharples (ed), ‘Excavations at Pierowall Quarry,
Westray, Orkney’, Proceedings of the Society of Antiquaries of Scotland, 75–125
McCormick, F, 2007a ‘Subsistence and environment’, in I Armit (ed), Anatomy of an Iron Age
Roundhouse: The Cnip Wheelhouse Excavations, Lewis, Society of Antiquaries of Scotland,
Edinburgh, 161–172
McCormick, F, 2007b ‘Mammal bones from prehistoric Irish sites’ in E M Murphy and N J White-
house (ed), Environmental Archaeology in Ireland, Oxbow Books, Oxford, 77–101
McCormick, F and Buckland, P C, 1997 ‘The Vertebrate Fauna’, in K J Edwards and I B Ralston
(ed), Scotland: Environment and Archaeology, 8000bc – ad1000, John Wiley & Sons, New
York, 83–103
McCormick, F and Murray, E V, 2007 Knowth and the Zooarchaeology of Early Christian Ireland,
Royal Irish Academy, Dublin
Melton, N D and Nicholson, R A, 2007 ‘A Late Mesolithic–Early Neolithic midden at West Voe,
Shetland’, in N Milner, O E Craig and G N Bailey (ed), Shell Middens in Atlantic Europe,
Oxbow Books, Oxford, 94–100
Miracle, P T and O’Brien, C, 1998 ‘Seasonality of resource use and site occupation at Badanj,
Bosnia-Herzegovina: subsistence stress in an increasingly seasonal environment?’ in T Rocek
and O Bar-Yosef (ed), Seasonality and Sedentism: Archaeological Perspectives from Old and
New World Sites, Peabody Museum of Archaeology and Ethnology, Harvard University,
Cambridge MA, 41–74
Mulville, J, 1999 ‘The faunal remains’, in M Parker Pearson and N Sharples (ed), Between Land
and Sea: Excavations at Dun Vulan, South Uist, Sheffield Academic Press, Sheffield, 235–246
Mulville, J, 2005 ‘Mammalian bone: resource exploitation, site activities and discussion’, in
N Sharples (ed), A Norse Farmstead in the Outer Hebrides: Excavations at Mound 3, Bornais,
South Uist, Oxbow Books, Oxford, 165–168
Mulville, J, 2010 ‘Red Deer on Scottish Islands’, in T O’Connor and N Sykes (ed), Extinctions and
Invasions: A Social History of British Fauna, Windgatherer Press, Oxford, 43–50
Mulville, J, forthcoming ‘The mammalian bone’, in M Parker Pearson, H Smith, J Mulville and
M Brennand (ed), A Norse-period Farmstead at Cille Pheadair, South Uist, c. 950–1300,
Sheffield Academic Press, Sheffield
Mulville, J, Bond, J and Craig, O, 2005 ‘The white stuff: milking in the North Atlantic’, in
J Mulville and A Outram (ed), The Zooarchaeology of Fats, Oils, Milk and Dairying, Oxbow
Books, Oxford, 167–183
Mulville, J and Ingrem, C, 2000 ‘Faunal Remains’, in K Brannigan and P Foster (ed), From Barra to
Berneray: The Archaeology of the Southern Isles of the Outer Hebrides, Sheffield Academic
Press, Sheffield, 250–265, 285–290, 299–305
Mulville, J and Ingrem, C, 2014 Animal remains from Uamh An Eich Bhric (Cave of the Speckled
Horse), Fiskavaig, Isle of Skye, Unpublished report, University of Cardiff, Cardiff
Mulville, J and Powell, A, 2012 ‘The moorland. 1. red deer’, in N Sharples (ed), A Late Iron Age
Farmstead in the Outer Hebrides: Excavations at Mound 1, Bornais, South Uist, Oxbow
Books, Oxford, 246
Mulville, J and Powell, A, forthcoming ‘Animals at Cladh Hallan: land, sea and air’, in M Parker
Pearson, J Mulville and H Smith (ed), Cladh Hallan, Sheffield Academic Press, Sheffield
MAM 37: 19-mulville – Press

296 maritime societies of the viking and medieval world

Mulville, J and Thoms, J, 2005 ‘Animals and ambiguity in the Iron Age of the Western Isles’, in
V Turner (ed), Tall Stories? Broch Studies Past, Present and Future, Shetland Amenity Trust,
Lerwick, 235–245
Nicholson, R and Davies, G, 2007 ‘The mammal bone’, in S J Dockrill, J M Bond, R Nicholson
and A Smith (ed), Investigations in Sanday, Orkney: Island Landscape Through 3000 years of
Prehistory, Volume 2: Tofts Ness, Sanday, The Orcadian, Kirkwall, 169–227
Noddle, B, 1974 ‘Report on the animal bone found at Dun Mor Vaul’, in E W Mackie (ed), Dun
Mor Vaul: An Iron Age Broch on Tiree, University of Glasgow Press, Glasgow, 187–198
Noddle, B, 1976–77 ‘The animal bones from Buckquoy, Orkney’, in A Ritchie (ed), ‘Excavations of
Pictish and Viking-age farmsteads at Buckquoy, Orkney’, Proceeding of the Society of
Antiquaries of Scotland 108, 174–227
Noddle, B, 1978–80 ‘Animal bones from Dun Cul Bhuirg, Iona’, in J N Ritchie and A M Lane (ed),
‘Dun Cul Bhuirg, Iona, Argyll’, Proceedings of the Society of Antiquaries of Scotland 110,
209–229
Noddle, B, 1981 ‘A comparison of mammalian bones found in the midden deposit with others from
the Iron Age site of Dun Bhuirg’, in R Reece (ed), Excavations in Iona 1964 to 1974, University
College London Institute of Archaeology, London, 225–227
Noddle, B, 1982 ‘The size of red deer in Britain — past and present, with some reference to fallow
deer’, in M Bell and S Limbray (ed), Archaeological Aspects of Woodland Ecology, British
Archaeological Reports International Series 146, British Archaeological Reports, Oxford,
315–333
Noddle, B, 1983 ‘Animal bones from Knap of Howar’, in A Ritchie (ed), ‘Excavation of a Neolithic
farmstead at Knap of Howar, Papa Westray, Orkney’, Proceedings of the Society of
Antiquaries of Scotland 113, 92–100
Noddle, B, 1986 ‘Animal bones’, in A Whittle (ed), Scord of Brouster an Early Agricultural
Settlement on Shetland, Excavations 1977–79, Oxford University Press, Oxford, 132
Noddle, B 1997 ‘Animal bone’, in S Buteux (ed) Settlements at Skaill, Deerness, Orkney:
Excavations by Peter Gelling of the Prehistoric, Pictish, Viking and Later Periods, 1963–1981,
British Archaeological Reports, British Series 260, Archaeopress, Oxford, 235–243
O’Sullivan, T, 1998 ‘The mammal bone’, in N Sharples (ed), Scalloway: A Broch, Late Iron Age
Settlement and Medieval Cemetery in Shetland, Oxbow Books, Oxford, 91
Palsson, H and Edwards, P, 1978 Orkneyinga Saga: The History of the Earls of Orkney, Hogarth
Press, London
Parker Pearson, M and Sharples, N, 1999 ‘Norse settlement on the Outer Hebrides’, Norwegian
Archaeological Review 21, 41–62
Parks, R and Barrett, J, 2009 ‘The zooarchaeology of Sand’, in K Hardy and C Wickham-Jones
(ed), Mesolithic and Later Sites around the Inner Sound, Scotland: The Work of the Scotland’s
First Settlers Project 1998–2004, Society of Antiquaries of Scotland, Edinburgh, section 3.11
Platt, M I, 1934 ‘Report on the animal bones’, in G J Callander and G G Walter (ed), ‘A long
stalled chambered cairn or mausoleum (Rousay type) near Midhowe, Rousay, Orkney’,
Proceedings of the Society of Antiquaries of Scotland 10, 321–349
Platt, M I, 1935 ‘Report on the animal bones’, in G J Callander and G G Walter (ed), ‘A stalled
chambered cairn, the Knowe of Ramsay, at Hullion, Rousay, Orkney’, Proceedings of the
Society of Antiquaries of Scotland 69, 341–343
Platt, M I, 1936 ‘Report on the animal bones found in the chambered cairn, Knowe of Ramsay,
Rousay, Orkney’, Proceedings of the Society of Antiquaries of Scotland 70, 415–419
Platt, M I, 1937a ‘Report on the animal bones’, in G J Callander and G G Walter (ed), ‘Long stalled
cairn at Blackhammer, Rousay, Orkney’, Proceedings of the Society of Antiquaries of
Scotland 11, 306–308
MAM 37: 19-mulville – Press

dealing with deer 297

Platt, M I, 1937b ‘Report on the animal bone’, in C Calder (ed), ‘A Neolithic double-chambered
cairn of the stalled type and later structures on the Calf of Eday, Orkney’, Proceedings of the
Society of Antiquaries of Scotland 11, 152–154
Platt, M I, 1956 ‘Report on the animal bone’, in J R C Hamilton (ed), Excavations at Jarlshof, Her
Majesty’s Stationery Office, Edinburgh, 213–215
Price, N, 2002 The Viking Way: Religion and war in late Iron Age Scandinavia, Uppsala University
Press, Uppsala
Rackham, J, 1986 ‘An analysis of the animal remains’, in C D Morris and N Emery (ed), ‘The
chapel and enclosure on the Brough of Deerness, Orkney: survey and excavation, 1975–1977’,
Proceedings of the Society of Antiquaries of Scotland 116, 301–374, fiche A1–4, G7
Rackham, J D, 1989 ‘Excavation 1960: the biological assemblage’, in C D Morris (ed), The Birsay
Bay Project, Volume 1, University of Durham, Durham, 77–80
Rowley-Conwy, P, 1994–95 ‘Meat, furs and skins: Mesolithic animal bones from Ringkloster, a
seasonal hunting camp in Jutland’, Journal of Danish Archaeology 12, 87–98
Searle, J B, 2008 ‘The colonization of Ireland by mammals’, in J L Davenport, D P Sleeman and
P C Woodman (ed), Mind the Gap: Postglacial Colonization of Ireland, Irish Naturalists’
Journal Special Supplement, 109–115
Sellar, T J, 1989 ‘Bone report’, in B Bell and C Dickson (ed), ‘Excavations at Warebeth (Stromness
cemetery) Broch, Orkney’, Proceedings of the Society of Antiquaries of Scotland 119,
101–131, fiche C1–G2
Serjeantson, D, 1990 ‘The introduction of mammals to the Outer Hebrides and the role of boats in
stock management’, Anthropozoologica 13, 7–18
Serjeantson, D, 2011 Review of animal remains from the Neolithic and Early Bronze Age of
Southern Britain, English Heritage, Portsmouth
Serjeantson, D, 2013 Farming and Fishing in the Outer Hebrides ad 600 to 1700: The Udal, North
Uist, The Highfield Press, Southampton
Serjeantson, D and Bond, J, 2007 ‘Cattle and sheep husbandry’, in S J Dockrill, J M Bond,
R Nicholson and A Smith (ed), Investigations in Sanday, Orkney: Island Landscape Through
3000 Years of Prehistory, Volume 2: Tofts Ness, Sanday, The Orcadian, Kirkwall, 169–227
Serjeantson, D, Smithson, V and Waldron, T, 2005 ‘Animal husbandry and the environmental
context’, in A Ritchie (ed), Kilellan Farm, Ardnave, Islay, Society of Antiquaries of Scotland,
Edinburgh, 151–168
Sharples, N, 2000 ‘Antlers and Orcadian rituals: an ambiguous role for red deer in the Neolithic’,
in A Ritchie (ed), Neolithic Orkney in its European Context, McDonald Institute for Archaeo-
logical Research, Cambridge, 107–116
Sharples, N M, 2005 A Norse Farmstead in the Outer Hebrides: Excavations at Mound 3, Bornais,
South Uist, Oxbow Books, Oxford
Sharples, N and Smith, R, 2009 ‘Norse settlement in the Western Isles’, in A Woolf (ed), Scandi-
navian Scotland — Twenty Years After, The Committee for Dark Age Studies, University of
St Andrews, St Andrews, 103–130
Smith, C, 1994 ‘Animal bone report’, in B Ballin Smith (ed), Howe: Four Millennia of Orkney
Prehistory, Excavations 1978–1982, Society of Antiquaries of Scotland, Edinburgh, 139–153
Smith, C, 2008 The Mammal and Bird Remains from Milla Skerra, Sandwick, Unst, Unpublished
report, GUARD, Glasgow
Steele, T E, 2004 ‘Variation in mortality profiles of red deer (Cervus elaphus) in Middle
Palaeolithic assemblages from western Europe’, International Journal of Osteoarchaeology
14, 307–320
Thoms, J E, 2003 Aspects of economy and environment of north west Lewis in the first millen-
nium ad: the non-marine faunal evidence form Bostadh and Beirgh considered within the
MAM 37: 19-mulville – Press

298 maritime societies of the viking and medieval world

framework of North Atlantic Scotland, Unpublished PhD thesis, University of Edinburgh,

Edinburgh
van Wijngaarden-Bakker, L H, 1974 ‘The animal remains from the Beaker settlement at New-
grange, Co. Meath: first report’, Proceedings of the Royal Irish Academy 74C, 313–383
van Wijngaarden-Bakker, L H, 1986 ‘The animal remains from the Beaker settlement at New-
grange, Co. Meath: final report’, Proceedings of the Royal Irish Academy 86C, 17–111
Vickers, K, Ul Haq, S and Hanshaw-Thomas, J, 2014 ‘The faunal remains’, in M Parker Pearson
and M Zvelebil (ed), Excavations at Cill Donnain: A Bronze Age Settlement and Iron Age
Wheelhouse in South Uist, Oxbow Books, Oxford, 155–173
von Holstein, I C C, Ashby, S P, van Doorn, N L, Sachs, S M, Buckley, M, Meiri, M, Barnes, I,
Brundle, A and Collins, M J, 2014 ‘Searching for Scandinavians in pre-Viking Scotland:
molecular fingerprinting of early medieval combs’, Journal of Archaeological Science 41, 1–6
Vretemark, M, 1997 Från ben till boskap. Kosthåll och djurhållning med utgångspunkt i medeltida
benmaterial från Skara, Skaraborgs La?nsmuseum, Nossebro
Vretemark, M, 1998 Raw materials and urban comb manufacturing in medieval Scandinavia,
Anthropozoologica 25/26, 201–206
Wickham Jones, C, 2004 ‘Searching for the first settlers of Orkney’, in Papers in Honour of Daphne
Home Lorimer MBE, http://www.orkneyjar.com/archaeology/dhl/papers/cwj/index.html
MAM 37: 20-harrison – Press

chapter 20
‘WARRIOR GRAVES’? THE WEAPON
BURIAL RITE IN VIKING AGE BRITAIN
AND IRELAND

By Stephen H Harrison

In the winter of 1846–47, the Danish archaeologist J J A Worsaae was in Dublin.

While he was unimpressed by the quality of much of the antiquarian research he
found there (Henry 1995, 1, 13), he did have an opportunity to examine a remarkable
assemblage of artefacts, principally weapons, which had been discovered during
the recent excavation of a railway cutting at Kilmainham, then on the western edge of
the city. When invited to address the Royal Irish Academy on the then hotly debated
‘three age system’ (Rowley-Conwy 2007), Worsaae took the opportunity to expand
his discussion of the Iron Age to include the Kilmainham assemblage and some
related material. These artefacts, he asserted, were ‘Danish, or rather Norwegian’ in
origin and ‘perhaps the very weapons by which Norsemen had shed Irish blood’.
Their ‘buried owners’, in turn, could be equated with ‘a small number of vikings from
Scandinavia’ (Worsaae 1847, 331–333).
Worsaae’s work was to have a profound influence on the study of the Viking
Age not just in Ireland but across these islands, particularly after the publication of an
English translation of his Minder om de Danske og Normændene i England, Skotland
og Irland in 1852. While Worsaae went out of his way to stress some of the more
positive aspects of the Scandinavian presence in Britain and Ireland, it is interesting to
note that his few references to ‘Viking’ graves consistently stress their military char-
acter. Those buried with weapons, or indeed in mounds, in these islands were self-
evidently ‘vikings’ and ‘warriors’ (Worsaae 1852, 242, 252, 328). In Ireland, these
‘Viking’ graves have been consistently interpreted as those of ‘warriors’ and as poten-
tial evidence for raiding activity ever since. By the early 20th century, this inter-
pretative model had become so strong that it was even applied to women’s graves
(ie those containing oval brooches). Early discussion of the woman’s grave found
close to the beach at Ballyholme, Co Down in 1903, for example, linked it to a raid on
a nearby monastery at Bangor that was recorded in ad 824 (Cochrane 1906, 452).
Despite a few dissenting voices, notably Sheehan (1988, 68–70), even some of the
most recently discovered burials have been called ‘warrior graves’ and linked to a
possible failed ‘raiding party’ (Simpson 2005, 53). This approach is in no way
MAM 37: 20-harrison – Press

300 maritime societies of the viking and medieval world

confined to Ireland. Since the 1960s, those few ‘Viking’ graves which have been iden-
tified in eastern and southern England, from Repton, Derbyshire (Biddle and Kjølbye-
Biddle 1992, 40), to Reading, Berkshire (Graham-Campbell 2001, 115), have been
linked to the perambulations of the Great Army and the deaths of some of its less
fortunate members. In Scotland, commentators have been a little more willing to
acknowledge the relationship between burial and settlement, but the martial char-
acter of these graves has been emphasized by many commentators, beginning with
Anderson (1874, 562) and continued by prominent figures such as Eldjárn (1984, 8),
who suggested that the graves of the Western Isles at least represented ‘real Vikings in
the true sense of the word . . . [unlikely] to stay in these small unassuming islands’. On
the Isle of Man, too, while many furnished burials have been linked to settlement
rather than more transient activity, those with weapons are nonetheless seen as
‘warrior’s graves’ (Bersu and Wilson 1966, xi).
For more than 150 years, antiquarians and archaeologists across these islands
have based their interpretations both on the surviving historical record and on the
empirical evidence provided by ‘Viking’ graves themselves. The various recensions of
the Irish annals and the Anglo-Saxon chronicle record and arguably emphasize the
military exploits of lay patrons and their opponents, and the compilers of the Cogadh
Gáedhel re Gallaibh had even stronger motivations for emphasizing the military char-
acter of the Viking presence in these islands (Nı́ Mhaonaigh 2007, 66). In their graves,
too, these ‘Vikings’ seem to have left a tangible record of their militarism. Of the 379
potential Viking Age furnished burials which were identified in a recent study of
Britain and Ireland, 200 contained weapons, and of the 1154 grave goods or artefact
groups identified in the same study, 366 were weapons (Harrison 2008, 65, 105).
Graves containing weapons outnumber those containing oval brooches or other
‘female’ artefacts by almost four to one, and weapons — swords, spearheads, axe-
heads, arrowheads and shield bosses — outnumber every other category of artefacts
found in these ‘Viking’ graves, including dress-fasteners and jewellery. Graves, too,
provide some of the best evidence for the plundering of monasteries and the wealth of
these islands, even if the majority of modified ecclesiastical artefacts come from
women’s graves in Norway rather than ‘Viking’ (weapon) graves in Britain or Ireland
(Wamers 1998, 42, 44). Nonetheless, a superficial examination of these graves sug-
gests that these insular Scandinavians have been caught ‘red-handed’, buried with
their plunder and the weapons with which they won it, having (according to docu-
mentary sources at least) terrorized entire indigenous populations in the process. To
all too many commentators, these graves represent the dead of an essentially transient
population who buried their fallen before moving on to pillage other unfortunate
indigenous communities elsewhere.
There have, of course, been dissenting voices. As early as 1945, Shetelig (1945, 2)
argued that the ‘Viking’ graves of Britain and Ireland were actually the remains of
settlers rather than raiders, and represented the first generations of incoming popula-
tion groups who thought themselves here to stay. This interpretation was entirely
in keeping with contemporary Scandinavian interpretations of similar graves in
Norway. There, too, weapons were frequent discoveries in graves, but their occu-
pants were rarely, if ever, seen as ‘Vikings’. Instead, they were ‘peasants’ (bønder) or
MAM 37: 20-harrison – Press

‘warrior graves’? 301

perhaps more accurately ‘free farmers’ (Brøgger 1929, 11–13), and these weapon
graves therefore represented the backbone of Viking Age Norwegian society. While
these individuals clearly had the right to carry weapons, the ‘folk weapons’ (folkvápn)
of the rather later Norwegian laws, and to attend local weapon musters (Solberg 1985,
68–69), their military role was generally underemphasized. Far from being seen as in
any way transient, those buried in these graves have consistently been seen as firmly
fixed within the landscape, buried close to the places where they had lived and died
(eg Skre 2001, 10–11). On occasion, the interpretation of these graves as the product
of a more or less egalitarian society creaked under the strain: Brøgger (1929, 13, 20),
for example, explained the contents of a grave from Aamot (Åmot), Hedmark, purely
in terms of the abundance of iron in the period. Equipped with a sword, two axeheads,
a spearhead, a shield boss, multiple arrowheads, a penannular brooch, a bridle, a
sickle and a cauldron, this grave was more richly furnished than any in Britain or
Ireland, and comments such as this inevitably set the tone for the interpretation of all
insular ‘Viking’ graves as those of individuals of relatively modest status, albeit with
some minor, often local, variations (Brøgger 1929, 126–127).
The interpretation of burials with weapons as part of a widespread ‘folk custom’
came under increasing assault in the 1980s. In a detailed study of furnished graves
around Sognefjord, Dommasnes (1982, 71, 73) argued that the rite was confined to a
relatively small proportion of the population. While increasing numbers of artefacts
corresponded to individuals of progressively higher status, all furnished graves, how-
ever modest, corresponded to individuals of ‘relatively high rank’. In 1985, Solberg
published a more wide-ranging study of ‘male’ and ‘female’ burial assemblages across
much of central and southern Norway. Despite using very generous definitions, she
identified no more than 4692 potential burial assemblages, 3796 of which contained
at least one weapon, with the remaining 833 containing one or more ‘female’ artefacts
(1985, 64–65). While Hagen (1967, 397) produced a slightly higher estimate of more
than 7000 graves, and Stylegar (2010, 71) has recently suggested a little less than 8000,
these figures remain exceptionally low. If it is assumed that approximately 7500 graves
represent 30-year generations spread over a 400-year period, then the population
represented at any one time is little more than 564 individuals, spread across an area
some 1800km from north to south (Harrison 2008, 104). Furthermore, the fact that
male graves outnumber those of women by a figure approaching five to one (Solberg
1985, 66, 67) is a statistic that, if these furnished graves represented entire population
groups, would give an entirely new impetus to Viking Age expansion! These statistics
are, of course, a gross oversimplification which ignores both regional and chrono-
logical variations in the record. These variations, however, provide further evidence
that, even allowing for the fact that an unknown proportion of furnished graves
remain undiscovered, only a fraction of the Norwegian population can have been
buried in this way. Both Solberg (1985, 73–74) and Dommasnes (1982, 83) concluded
that these graves represented individuals with a relatively high rank or status.
Solberg went on to argue that, if graves containing weapons were expressions of
rank, then the number and type of weapons placed in them might well reflect
variations in status among those using this burial rite. Among her ‘group one’ graves
MAM 37: 20-harrison – Press

302 maritime societies of the viking and medieval world

— those containing a single weapon — she argued that individuals buried with a
sword were of a higher status than those buried with one of the other ‘folk weapons’
— the spear or axe. Those in her ‘group two’, buried with two weapons, represented
a higher status again, while ‘group three’ graves, which contained all three weapons
— sword, spear and axe — represented individuals of the very highest status (Solberg
1985, 66). While there were some regional variations, graves with two weapons were
consistently rarer than those with one, and group three burials were the rarest of all,
statistics which supported the hypothesis that these graves represented individuals of
the highest status (Figure 20.1). Although Solberg’s attempts to link her three burial
groups to social groups described in the Norwegian law tracts were not particularly
successful, her analysis indicated that the weapons placed in Norwegian graves could
be interpreted as expressions of rank rather than simple military ability, and con-
firmed that all such graves represented individuals of some considerable standing.
If this is the case in Norway, then it is almost certainly true for Britain and
Ireland as well. Far from being the graves of ‘ordinary’ Vikings, the 200 weapon
graves identified in these islands must represent the graves of a select group, presum-
ably individuals of relatively high rank. There are, however, some intriguing differ-
ences between the numbers and types of weapons placed in insular Viking graves,
and those groups which Solberg identified in Norway (Figure 20.1). Of the insular
weapon graves, 186 were suitable for direct comparison with Solberg’s findings, with
104 (56%) corresponding to her group one, 65 (35%) to her group two, and 17 (9%)

Figure 20.1 Proportions of Solberg’s groups 1, 2 and 3 graves in western, central and
eastern Norway, and in Britain and Ireland (S H Harrison; Norwegian comparisons after
Solberg 1985, fig 5)
MAM 37: 20-harrison – Press

‘warrior graves’? 303

to her group three (Harrison 2008, 106). In Norway, on the other hand, group one
graves formed somewhere between 61% and 74% of the total corpus, by region
(Solberg 1985, 66). Individuals buried with weapons in Britain and Ireland were there-
fore more likely to be buried with multiple weapons than were their counter-
parts in Norway. There were also variations within group one. While in Norway, the
majority of these graves contained an axe, and burials with swords were rarest,
these statistics were completely reversed in insular contexts (Figure 20.2). Seventy-
three (39% of the total) contained a single sword, while just eight (4%) contained a
single axe. In Norway, the figures are 14–21% for single swords, and 28–32% for
single axes (Solberg 1985, 66). As there is a general consensus that swords were the
most prestigious of the weapon types (Davidson 1962, 10), these figures suggest one
of two things. It may be that those whose social peers in Norway would have been
buried with a single axe or spearhead, or indeed with a single weapon, were the first
to abandon the weapon burial rite in Britain and Ireland. Those whose peers buried
their dead with multiple weapons in Scandinavia, on the other hand, were perhaps
the last to abandon the practice in these islands. Alternatively, those choosing to
bury their dead with weapons in Britain and Ireland may have chosen to invest
greater resources in individual burials, by depositing either more weapons or more
prestigious weapons (swords) in graves. In either case, it seems reasonable to infer
that these weapon graves represent an even smaller proportion of the insular Scandi-
navian population than that of Norway itself.

Figure 20.2 Proportions of graves containing single swords, spears and axeheads in western,
central and eastern Norway, and in Britain and Ireland (S H Harrison; Norwegian comparisons
after Solberg 1985, tab 1)
MAM 37: 20-harrison – Press

304 maritime societies of the viking and medieval world

Of course, the direct statistical comparison of such large areas represents the
oversimplification of a much more complex evidence base. Solberg’s research clearly
demonstrated strong regional variations within the Norwegian corpus, with the
figures for western Norway being rather different to those of the two other regions
which she studied. It also seems clear that more localized variations in the burial rite
could and did occur (Price 2008). The dead, after all, do not bury themselves (Parker
Pearson 1999, 3), and every burial involving the deposition of artefacts must have
involved a subtle balance between prestige, pride and necessity. The relationship
between perceived and actual status was further complicated by the fact that no arte-
fact has an entirely ‘fixed’ value or meaning, but rather has meaning imposed upon it.
While the widespread use of weapons clearly demonstrates some shared value across
most of the Viking world, the perceived importance of individual graves must have
been strongly influenced by the local burial tradition. There is certainly local vari-
ation in the insular burial rite. The evidence from the Kilmainham-Islandbridge
burial complex at Dublin, for example, is notoriously difficult to interpret, but new
research by the Irish Viking Graves Project, based at the National Museum of
Ireland, suggests that many of these graves contained multiple weapons. It is at the
very least clear that the complex contains a remarkable number of graves with
swords — 42 at the most recent count — and the quality of their hilt decoration
further demonstrates the importance of Dublin in the 9th and early 10th centuries
(Harrison and Ó Floinn 2014, 75–93). The sheer quantity of graves may also reflect
the role of competitive display within the population of what was almost certainly a
nucleated settlement.
At the opposite extreme are those few weapon graves which have been identified
within the former Danelaw. Weapon burials, like ‘Viking’ graves generally, are
fundamentally a feature of the western seaboard, or more specifically northern and
western Scotland and the Irish Sea basin. The paucity of Viking graves to the east of
the Pennines has been recognized by many commentators, from Graham-Campbell
(1980, 379) to Richards (2000, 142). Using the most generous criteria possible, no
more than 29 graves with weapons can be identified in this area. Of these, 19 (66%)
contained just one weapon, and only one contained all three (Harrison 2008, 110).
Thus, the Danelaw is characterized not just by fewer weapon burials, but also by the
fact that those graves which do occur contain fewer weapons. For Scandinavians
practising and witnessing the weapon burial rite in this area, the perceived value of
individual weapons may well have been rather different to that at Dublin or
elsewhere on the western seaboard.
These local variations also provide further evidence that while the weapon
burial rite practised in Britain and Ireland was self-evidently related to that used in
Scandinavia, it was not simply a passive reflection of that rite. Instead, the number
and type of weapons incorporated within graves clearly changed to reflect local
circumstances, situations and traditions. Much the same comments can be made
about the Scandinavian furnished burial rite generally, both in Scandinavia and
elsewhere. To many post-processualists, Solberg’s assumption of any constant rela-
tionship between the perceived value of artefacts is problematic, but the sheer scale
of her research suggests that, despite local variations, broad values underlay the
MAM 37: 20-harrison – Press

‘warrior graves’? 305

processes by which artefacts were selected for deposition in graves. Nonetheless,

some aspects of her methodology are problematic. Leaving aside the identification of
graves, her study considered just three weapons — the aforementioned sword, spear
and axe — largely ignoring shields, because they were not ‘offensive’ weapons, and
entirely omitting arrowheads, presumably because they were not part of the ‘folk
weapon’ set. Nor did her methodology make allowance for the occurrence of mul-
tiple weapons of the same type in individual graves, even though this pattern clearly
occurs in some Norwegian graves. A study of insular weapon graves which includes
these five weapon types, and which treats groups of arrowheads as ‘single’ artefacts,
reveals a pattern which closely reflects that established using Solberg’s methodology
(Figure 20.3). The inclusion of shield bosses and multiple weapons, notably pairs of
spears, reduced the figure for graves with single weapons further, to just under 50%
of the corpus. It should be noted, however, that graves with four or five weapons
remained rare, with just 9% falling into this category (Harrison 2008, 109), and only
one grave, a boat burial from Westness, Orkney, contained all five weapon types
(Kaland 1995, 314–316). The distribution pattern confirms that graves with multiple
weapons tend to focus in specific areas (Figure 20.4). Dublin has already been men-
tioned, but there are other concentrations in the Inner Hebrides, and on the Isle of
Man. While allowing for regional variations and elements of competitive display, it

Figure 20.3 Proportions of Viking Age furnished graves containing one to five weapons in
Britain and Ireland (S H Harrison)
MAM 37: 20-harrison – Press

306 maritime societies of the viking and medieval world

Figure 20.4 The

distribution of insular
‘Viking’ graves containing
one to five weapons,
based on all possible
weapon graves identified
to date (S H Harrison)

remains very likely that these areas represent definite power foci in the period when
furnished burial was in use.
By the time these graves were created, Scandinavian groups were the only ones in
Britain and Ireland, and indeed northern Europe, who still deposited weapons in
graves. Nonetheless, it is important to realize that they were drawing on a far more
widespread tradition which placed weapons at the centre of a complex symbolic
web that went far beyond simple military function (Härke 1990, 42–43). In Anglo-
Saxon England, swords in particular were clearly seen as suitable royal or noble
gifts, bestowed on favourites and presented to peers (Davidson 1962, 110, 118–121).
‘Lesser’ weapons, notably spears and axes, although undoubtedly easier to manu-
facture and clearly more common, were expressions of the rank of other influential
MAM 37: 20-harrison – Press

‘warrior graves’? 307

members of society across Europe. Individuals had both the right and obligation to
carry and potentially use weapons on certain occasions (Halsall 1998, 30–31).
Shields, carefully decorated and visible from a considerable distance, could have been
associated with specific individuals and groups long before the development of
formal heraldry (Harrison 1995, 137–138, 141–143). All weapons, not just swords,
were potentially ‘both a symbol of power and a means of acquiring that power’
(Walsh 1995, 37).
Weapons could also represent other aspects of social identity, however. Archae-
ologists have consistently associated them with a specifically male identity, and, while
the precise relationship between grave goods and biological sex has been questioned
by some (Hadley and Moore 1999, 28–30), the relationship between weapons and
masculinity was clearly close, at least in the Viking Age. In Britain and Ireland, all
extant skeletal material which can be associated with weapon graves has proved to be
male (Harrison 2008, 102). It is also striking that, while Viking Age women could
apparently be represented by a range of artefact forms, all insular graves containing
any artefacts conventionally linked to men also contained at least one weapon.
Smith’s tools, for example, have always been found with weapons rather than in
specific ‘smith’s graves’, a situation which can be contrasted with the evidence from
Anglo-Saxon England, such as Tattershall Thorpe, Lincolnshire (Hinton 2000).
Research by Härke (1989, 146–147) has further demonstrated that Anglo-Saxon
knives could function as statements of male identity, and potentially as symbols of
social maturity. In Viking Age Britain and Ireland, new evidence from Balnakeil,
Caithness, points to the complexity of social identity in this later period. Here an
individual aged between 8 and 13 and no more than 1.52m tall was buried with a
sword, spear and shield, even though there is some debate as to whether or not he
could have wielded the sword which was buried with him (Batey 1995, 157–158).
This burial in particular raises a whole series of questions about the nature of social
identity and inherited rank. While it is often assumed that the period was one of great
social instability, with documentary sources making it clear that social ranks could be
transformed as a result of single military encounters, the Balnakeil burial suggests
that social power could be successfully transferred from one generation to the next,
and implies at least some degree of familial or societal stability.
The selection of specific weapons for burial may also reflect concerns that go
beyond simple numerical counts and which take account of local usages and prac-
tices. The increasing prominence of swords in the insular burial rite has already been
noted, and may reflect the concerns of individuals who sought to maintain a distinct
identity at least in part through burial rituals. Of all the various ‘Viking’ weapon
forms classified by Petersen in 1919, swords, or perhaps more specifically sword hilts,
have the widest geographical distribution. The repeated occurrence of specific ‘Scan-
dinavian’ hilt types, perhaps most obviously Petersen’s type H, in graves across the
Viking world gives an impression of cultural homogeneity (eg Walsh 1998, 229–230)
and suggests that those burying their dead outside Scandinavia were choosing a
specifically Scandinavian artefact type as a grave good. The reality is, of course, more
complex, as by no means all of the hilt types which Petersen identified were neces-
sarily of Scandinavian origin, or even manufacture. Petersen’s type K, for example, is
MAM 37: 20-harrison – Press

308 maritime societies of the viking and medieval world

of Frankish origin (Walsh 1998, 230), and there is no particular reason to believe
those examples found in Britain and Ireland had arrived via Scandinavia (Harrison
and Ó Floinn 2014, 90–91). Similarly, type L swords are usually seen as Anglo-Saxon
(but see Pierce 2002, 74–82), while the later type X is a common ‘Germanic’ form
(Petersen 1919, 167). If there are few appreciable differences between the sword hilts
found in insular and Norwegian contexts, this is perhaps as much a reflection of the
scale of imports to Norway as the export of Norwegian swords elsewhere. It is the
funerary context of these swords, not their origin, which makes them ‘Scandinavian’,
and, while the absence of an indigenous Irish furnished burial tradition makes it
difficult to draw direct comparisons, finds such as the exceptional Petersen type K
sword found at Ballinderry crannog 1, Co Westmeath (Wallace 2002, 228), strongly
suggest that these artefacts were potent symbols of aristocratic power that crossed
social and cultural boundaries, from Frankish manufacturers to Scandinavian
importers, and Irish users — all terms which simplify the complexity of ethnic
identities in this period.
If, however, swords enjoyed near-universal prestige, the same cannot be said of
all ‘Viking’ weapon types. Perhaps the single most striking difference between
contemporary weapon assemblages in insular and Norwegian graves is the almost
total absence of axeheads in Britain and Ireland. Just 35 examples have been recorded
in the latter context, spread between 34 graves, or 17% of the total. In Norway,
Solberg’s research indicates that somewhere between 48% and 63% of weapon graves,
by region, contained axeheads (1985, 66). It is the rarity of axeheads that makes
Solberg’s group 3 graves (those with all three ‘folk weapons’) relatively uncommon in
insular contexts (above). In 1852, the Irish antiquarian George Petrie noted that he had
been forced to ‘withdraw’ an iron axe from his collection because of ‘the ridicule it
created’ (Anon 1852, 242). While this may go some way towards explaining the rarity
of axeheads at Dublin, it is difficult to believe that such selective preservation could
have influenced the occurrence of axeheads across such broad geographical areas.
There is no particular reason to suggest that most insular antiquarians collected
‘different’ weapons to their Norwegian counterparts, and consequently this distribu-
tion must reflect actual deposition patterns in the Viking Age. This pattern is partic-
ularly puzzling because the 12th-century writer Giraldus Cambrensis (Gerald of
Wales) recorded a tradition that the use of the axe in Irish warfare was a practice
which had been introduced by the Ostmen — the Hiberno-Norse descendants of the
first Scandinavian settlers in these islands (O’Meara 1982, 122). Nonetheless, it seems
clear that their 9th- and early 10th-century antecedents chose only occasionally to
include axeheads in their burial rites, and, while this does not necessarily mean that
these artefacts were not then in use, it does suggest that they were relatively rare.
Given that the axe was the one ‘Viking’ weapon which seems to have been unknown
to insular groups before the arrival of Scandinavian groups, it is tempting to link
the rarity of axes in Viking graves to some early, local influence, but this possibility
cannot be demonstrated conclusively.
There is, however, conclusive evidence for local influence on other ‘Scandi-
navian’ weapon forms, or rather weapons placed in insular Scandinavian graves.
Across Scotland, the Isle of Man and north-west England, for example, the
MAM 37: 20-harrison – Press

‘warrior graves’? 309

overwhelming majority of spearheads from Viking graves appear to correspond to

‘standard’ Petersen types. In Dublin, however, a completely different spearhead type
is found. Characteristically smaller than most Scandinavian types, many examples of
this ‘Dublin type’ were secured to their shafts using a series of copper pins which
were clearly decorative as well as functional (Bøe 1940, 26). These spearheads are
little studied (but see Harrison and Ó Floinn 2014, 102–110) and are only sporadic-
ally noted outside the Dublin area (eg Wilson 2008, 31, 37), but presumably reflect
an insular manufacturing tradition. A few examples have been found at crannog
sites such as Lagore, Co Meath (Hencken 1950, 94–98), but these artefacts (like the
Ballinderry sword) may represent a ‘Viking’ influence on these royal sites rather than
an insular development. Whatever their precise origin, the quantity of such ‘local’
spearheads in insular Viking graves is almost certainly underestimated by archae-
ologists. During the last systematic study of these burials (Shetelig 1940), Grieg (1940)
and Bjørn (Bjørn and Shetelig 1940) expected to find Scandinavian spearhead types,
and it was only the sheer quantity of these spearheads at Dublin which forced Bøe
(1940) to consider the possibility of local influences and origins. Similar, local spear-
heads may well lie unrecognized in English as well as Scottish and Manx collections.
In addition to these ‘Dublin type’ spearheads, the shield bosses recovered from
the Kilmainham-Islandbridge burial complex in particular are very different to those
found elsewhere in the Viking world (Harrison and Ó Floinn 2014, 122–125). Small
and conical rather than large and hemispherical, these bosses seem to represent the
effective fusion of an Irish tradition of small shield bosses with an on-going Anglo-
Saxon tradition of conical forms (Harrison 1995, 116). Unlike the spearheads, no boss
of this type has ever been found on an Irish settlement site, or any other ‘indigenous’
context — indeed, only three definite examples have been found outside Dublin. At
Dublin, and particularly within the Kilmainham-Islandbridge complex, however,
these ‘Dublin type’ bosses are entirely dominant, with only a handful of the larger
Scandinavian and Irish Sea types known (Bøe 1940, 33–38). This geographical focus,
coupled with the remarkably consistent size and shape of these artefacts, suggests a
local point of manufacture and perhaps even a single workshop. Far less clear, how-
ever, are the reasons why the inhabitants of 9th-century Dublin chose to develop this
boss type, let alone to bury their dead with it. Shields were presumably replaced
far more frequently than swords, but this does not in itself explain the development
of an entirely new boss type. Military considerations are perhaps the most obvious
explanation, and Dublin’s population may have been responding to the same tactics
which led the Irish elite to use equally small shield bosses (Harrison 1995, 34–44).
Nonetheless, it is difficult to imagine any specific factors which could have affected
the population of Dublin so deeply, while apparently having almost no impact on
other insular Scandinavian groups. The possibility of more subtle local influences
must therefore be considered. Whatever the precise explanation for the adoption of
these bosses (and indeed spearheads), the end result was that many of Dublin’s
leading figures were buried with weapons which are specifically linked to the local
area. If these artefacts were also used in daily life — and there is no reason to believe
they were not — then Dublin’s male elite were equipped with shields and spearheads
which were used at that settlement and almost nowhere else. Weapons, it would
MAM 37: 20-harrison – Press

310 maritime societies of the viking and medieval world

seem, could be used to express a distinct, local identity, even when incorporated
within what had originally been a Scandinavian burial rite.
The evidence from Dublin also suggests that such local identities could and did
exist, and that many of those buried with weapons were actually members of more or
less permanently established communities and groups. Duffy (1999, 5) and others
have argued that the populations of military bases such as Dublin were relatively
transient, with the arrival of substantial fleets often balanced by heavy losses in the
field. The evidence from Dublin’s graves, however, would suggest that these newly
arrived groups were rarely buried with artefacts, let alone weapons. Instead, the
preponderance of Dublin-type spearheads and shield bosses would suggest that those
buried with these artefacts at Kilmainham-Islandbridge and at other sites around
Dublin were those with some extended association with the settlement — or, at the
very least, that those controlling the burial ceremony wished observers to think this.
In the midst of short-term influxes, an elite group or groups more or less permanently
established at Dublin sought to control the settlement. The weapon burial rite seems
to have been an important mechanism by which this was achieved. If this was true of
Dublin, then it may equally have been true of the individuals buried in the much
smaller cemeteries and single graves which are scattered around the western littoral
of Britain and Ireland. The fact that many of these burials are coastal and situated
close to sheltered, if shallow, bays and inlets (Harrison 2007, 176) does not neces-
sarily mean that the individuals buried in them were part of an entirely transient
population who simply rowed ashore to bury their dead. While this model has been
applied to graves in regions as diverse as Talacre, Wales (Smith 1933, 48), Eyrephort,
Co Galway (Raftery 1960, 5–6), and Sonning, Berkshire (Evison 1969, 342), weapon
graves, which represent a considerable sacrifice for those involved with the funerary
rite, are far more likely to represent the leaders of families, kin-groups and com-
munities who had strong vested interests in creating these memorials, and who
sought to establish a permanent presence in these areas.
In recent years, much has been made of the social importance of early medieval
burial mounds, which could function not just as expressions of status (Solberg 1985,
66) and land ownership (Wilson 2008, 55), but also as statements of ‘pagan’ identity
(van de Noort 1993, 70) and foci for local expressions of ritual practice (Price 2010).
Fundamentally, these monuments had a permanent impact on the landscape and
served as a reminder of the (ancestral) dead beneath them. In insular contexts, the
evidence for mounds is much more limited than is the case in Norway, a statistic
which may represent an early Christian (or at least indigenous) influence on burial
practices (Harrison 2008, 200–201). Nonetheless, it seems clear that weapon graves,
and indeed most Viking Age furnished burials in Britain and Ireland, were marked
on the surface in some way. Viking graves seem rarely, if ever, to have intercut each
other, even in cemeteries where graves were relatively close together, and this
phenomenon can only be explained if each grave were marked in some way. As well
as extant mounds, such as Ballateare, Man (Bersu and Wilson 1966, 46–48) and the
kerbstones which surrounded others, as at Kneep (or Cnip), Lewis (Dunwell et al
1995, 731), graves could be marked by individuals stones, as at Westness, Orkney
(Kaland 1995, 312); cairns, as at Woodstown, Co Waterford (Russell and Harrison
MAM 37: 20-harrison – Press

‘warrior graves’? 311

2011, 66); wooden posts, as apparently occurred at Repton (Biddle and Kjølbye-
Biddle 1992, 41) — or a combination of these elements. With or without associated
mounds, the sites of these graves were clearly remembered long after the funeral
ceremony itself.
When discussing burial mounds at Norwegian farm sites, Skre (2001, 10) has
argued that the creation of a mound, an event which occurred roughly once every
generation, was not simply part of the funeral ceremony, but a physical symbol of the
inheritance process — a means of demonstrating the new-found authority of the heir
through actions which physically displayed the perceived status of the deceased.
Many of these Norwegian mounds seem to have covered furnished burials. Is it not
therefore possible that the formal deposition of artefacts within the grave pit or
chamber formed an equally important part of this inheritance process? While the pit
or chamber was ultimately backfilled, vanishing from public view, some memory of
the artefacts involved must have remained. In Anglo-Saxon England, it has been
argued that the key point of a high-status funeral involving grave-goods was the
display of a ‘tableau’: the deceased surrounded by artefacts which represented his
role and status within a given community (Geake 2003, 260). Given the comparable
care with which artefacts were deposited in Viking Age graves, it seems entirely
reasonable to suggest that weapons formed a key part of that tableau, at least in the
case of high-status males.
Despite having evidence for some two hundred insular weapon graves, the
precise deposition of grave-goods is recorded in only a handful of cases and even
fewer have been professionally excavated. In general, however, it would seem that
weapons were placed in prominent positions around the body. The sword was gen-
erally placed to one side, point down, with the hilt positioned either at the hip, as in
the case of a grave found at Ballinaby, Islay in 1932 (Edwards 1934, 75), and allegedly
at Eyrephort, Co Galway (Raftery 1960, 3), or the shoulder, as was the case with a
grave found at Inchicore, Dublin, in 1934 (Bøe 1940, 60–62), and apparently at Cronk
Moar, Man (Bersu and Wilson 1966, 69). At Larne, Co Antrim, on the other hand, the
sword is said to have been placed across the skeleton’s breast, with the hilt towards
the right hand (Smith 1840, 42). The position of spearheads is rarely recorded, but in
the case of graves with multiple weapons they seem often to have been placed on the
opposite side of the body to the sword, as in the coffin or chamber at Ballateare, Man
(Bersu and Wilson 1966, 50). However, a position at the hip is only noted in the
aforementioned graves at Eyrephort and Inchicore, and this may be the result of their
excavators mistaking these spear blades for ‘daggers’. Only three axeheads have had
their positions within the grave recorded. In the 1932 Ballinaby find, the axehead was
placed opposite the sword, close to the ‘right elbow’ (Edwards 1934, 75), in a grave
found at Reay, Caithness, in 1926, it was positioned at the left knee (Edwards 1927,
203), and in the grave found at Woodstown, Waterford, in 2004, it was found in a
disturbed area close to the body’s feet (Harrison and Ó Floinn 2014, 666). Shield
bosses could be placed across the feet, as at Eyrephort (Raftery 1960, 3), or on the
chest, as was the case with the aforementioned Reay burial (Edwards 1927, 203), as
well as a more recent discovery at South Great George’s Street, Dublin (Simpson
2005, 39–40). In all cases, it will be noted that these weapons were placed around
MAM 37: 20-harrison – Press

312 maritime societies of the viking and medieval world

the corpse, rather than ‘worn’ by it. This positioning is perhaps best explained by
considering these artefacts as part of Geake’s ‘tableau’, displayed for the benefit of
observers (Harrison and Ó Floinn 2014, 280–282).
An interesting local variation seems to have occurred on the Isle of Man, where
in at least two cases weapons were placed on or beside the coffin or chamber, rather
than beside the corpse itself. At Ballateare, while a sword and spearhead were placed
beside the body, two more spears were placed on top of the ‘coffin’ and a shield was
placed beside it (Bersu and Wilson 1966, 50). At Cronk Moar, while a sword seems to
have been placed inside the ‘coffin’, a spear and shield were placed on top of it (Bersu
and Wilson 1966, 67–68). The original report argues that these large artefacts were
left outside the coffin/chamber because they were too large to fit inside it, but the
labour investment involved in the creation of these graves and their associated
mounds was such that the construction of slightly larger chambers can hardly have
been a concern. The positioning of these artefacts outside the coffin/chamber would,
however, have meant that they remained visible until the point when backfilling
began. Whether inside or outside the chamber, witnesses could potentially have
remembered the positioning of these artefacts within the tableau for years after the
funeral itself (Williams 2010, 118–121). Even if precise details became blurred with
time, it should also be remembered that this formal display occurred at precisely that
time when, according to Skre’s inheritance model, the heir first established and
demonstrated his authority — a process which may at least occasionally have
involved the use of artefacts very similar to those deposited alongside his predecessor.
A model which emphasizes the importance of furnished graves in general, and
burials with weapons in particular, to the process of inheritance is, of course, par-
ticularly relevant to the environment within which insular Scandinavian burials were
created. It has long been accepted that ‘Viking’ graves represent only the first gener-
ations of Scandinavians active in these islands (Shetelig 1945, 23–24, 36). These
graves were created over a period of no more than 150 years, between ad c800 and
ad c950, and there is no typological reason why these graves could not have been
created within a century of each other (Harrison 2008, 79). Furthermore, individual
burial sites rarely contain graves which are significantly older or younger than their
peers, a detail which suggests that, at specific sites, no more than a single generation
was buried in this way. It would be wrong, however, to assume that these graves are a
simple, passive reflection of a folk tradition to which these settlers clung. Instead,
they represent the efforts of the first insular Scandinavian generation inheriting land
to establish and affirm their rights to that inheritance. In this context, the creation
of these graves may well have taken on a particular urgency. While it is generally
acknowledged that most insular graves are less well furnished than their Norwegian
counterparts (Wilson 1976, 99), the increased deposition of weapons identified in the
present paper may reflect a need to demonstrate legitimacy, perhaps by emphasizing
the military ability of the original ‘conqueror’. Burial sites with multiple weapon
graves of approximately the same date, such as Ballinaby, Islay (Graham-Campbell
and Batey 1998, 122–125), Westness, Orkney (Kaland 1995, 312–316), and/or the
more recent finds at Cumwhitton, Cumbria (Paterson et al 2014), may well represent
family (or at least kin-) groups, but the frequency of furnished burial rites at these
MAM 37: 20-harrison – Press

‘warrior graves’? 313

sites also suggests particularly troubled inheritance processes, or at least the regular
transfer of land during the period when furnished burials were in vogue. A link
between burial and inheritance also provides a new perspective on the abandonment
of the furnished burial ritual. While it would be foolish to deny the influence of
Christian (or at least indigenous) burial practices on this process, an emphasis on the
social rather than the religious significance of the rite provides an alternative
explanation. In many cases, a single grave could have provided sufficient connection
between a kin-group and a given landholding. Later generations may not have had
the same need to create these burials, and may therefore have adopted other
(unfurnished) burial practices more easily.
The more sceptical may argue that this ‘inheritance’ model is unduly mechan-
istic, and reduces what was undoubtedly a complex process of evaluation and selec-
tion on the part of those preparing the grave to a single over-arching social concern. It
may also be suggested that a model emphasizing the inheritance of land seems to
break down in the case of larger cemeteries such as Pierowall, Orkney (Thorsteins-
son 1968), and of course the Kilmainham-Islandbridge burial complex at Dublin
(Harrison and Ó Floinn 2014, 242–266). At the latter site in particular, there are
simply too many graves to represent individual landowners, and their focus at a
single site seems to reduce any potential associations with specific landholdings.
Nonetheless, the graves at this site remained the preserve of a limited proportion of
the population, and they may have had a social role which was closely related to
furnished burials elsewhere in these islands. Throughout the 9th and early 10th cen-
tury, Dublin had a ruling elite that not infrequently divided into distinct factions
(Downham 2007, 17–27). While documentary evidence is lacking, it may be assumed
that each pretender to the rule of Dublin was surrounded by followers of comparable,
if not equal, social rank. Like their rural counterparts, these individuals controlled
resources of some kind, and their deaths would have necessitated the renegotiation of
social relationships. The control of a specific funerary ceremony may have provided
an immediate advantage to specific groups, and provided an ideal arena for the dis-
play of wealth and power. Dublin’s graves are exceptionally well furnished not just
because the settlement was wealthy in this period, but because its leaders chose to
commit particularly elaborate artefacts and many weapons to these graves. This pro-
cess was undoubtedly influenced by competitive display, and Dublin’s graves conse-
quently represent a local variation in the broader pattern of isolated burials and small
cemeteries which are characteristic of areas of insular Scandinavian settlement at
this time.
Similar patterns may also be seen in areas with concentrations of furnished
graves, and weapon burials in particular. The graves in the southern part of the
Western Isles, particularly Colonsay, Oronsay and western Islay, where there is a
definite focus of Viking graves (Figure 20.3), and numerous examples of multiple
weapons in burials, may have been at least partially driven by a similar impetus. As a
result, these concentrations do not necessarily correspond to areas of more intensive
settlement, but rather to areas with particularly intense social rivalries. In order for
these rivalries to develop, however, those using furnished burial must have been
established in these areas for some time.
MAM 37: 20-harrison – Press

314 maritime societies of the viking and medieval world

The one obvious exception to this rule is, of course, Repton, Derbyshire, where
it is argued that all five graves containing weapons (and indeed several other fur-
nished burials) can be linked to the overwintering of the ‘Great Army’ at the site in
ad 873–874 (Biddle and Kjølbye-Biddle 1992, 40). By ad 873, the Great Army had
been active in England for some eight years. Given its (relatively) large size and lon-
gevity, its internal politics were undoubtedly complex, and it may be that furnished
burial at Repton, as at Dublin, was a means of reinforcing social order within this
group. It should perhaps be pointed out, however, that Repton also lies close to the
western boundary of the Danelaw as it was defined at a slightly later date, and
remained a site of considerable strategic and political importance throughout the
Viking Age (Richards 2003, 387, 390). As such, it could also have provided a suitable
location for furnished burial by a locally based group or groups at some point soon
after the enclosure at the site had been abandoned by the Great Army. The Repton
graves would then form part of the general pattern of furnished ‘weapon’ burial
across these islands, being associated with a settled rather than a transient com-
munity. Similar interpretations can be proposed for the furnished burials found at or
near other overwintering sites such as Reading and Nottingham (Graham-Campbell
2001, 106, 115). The Great Army, however, represented an exceptional gathering in
the context of 9th-century England, and it should be remembered that the furnished
burial rite was clearly characterized as much by flexibility and adaptation as by any
inherent conservatism. These burials could therefore represent a specific adaptation
to a set of unusual social stresses associated with this military enterprise.
Even at Repton, it is clear that Viking graves containing weapons are far from
‘simple’ interments and contain the remains of individuals who were much more than
simple ‘warriors’. More generally, there is strong evidence to suggest that furnished
burials in insular contexts, as the deliberate creations of families, kin-groups and/or
communities who had established themselves in the local area, represent efforts to
manipulate social perceptions, very often in the context of inheritance or the transfer
of power from one generation to the next. By establishing a physical link between
these groups and the local area, these ‘Viking’ graves anchored them within the local
context (Harrison 2007, 179–180). In the case of the most elaborately furnished
burials in particular, this process may have gone beyond local groups, ‘establish[ing]
a presence in the landscape of Norse-dominated territories, from which . . . the living
population could derive a sense of the historical legitimacy of their leaders’ power’
(Griffiths 2004, 127). Fundamentally, these burials represent a belief in a long-term
presence in the local area, even if they were not in themselves a guarantee of success.
It is also important to realize that while the use of weapons in graves is highly
visible in archaeological terms, the practice was almost by definition restricted to a
limited segment of the population, and was by no means the only high-status burial
rite, let alone burial form, practised by insular Scandinavian groups. Space does not
allow discussion of the closely related group of elaborately furnished ‘female’ graves,
dominated by examples containing oval brooches, although it should be noted that
there is considerable evidence for overlap with the present argument. Then there is
the range of more poorly furnished Viking Age graves, with neither brooches nor
MAM 37: 20-harrison – Press

‘warrior graves’? 315

weapons, which are so little studied that almost no definite comments can be made
(Harrison 2008, 166–190). At present, too, the cremation cemetery at Heath Wood,
Derbyshire, occupies a unique position in insular contexts (Richards 2004). Again
associated with the Great Army’s camp at Repton in ad 873–874, it is entirely pos-
sible that cemeteries containing similar burials and monument forms exist in other
parts of these islands, but have not yet been recognized as such. It would be foolhardy
in the extreme to suggest that these variations in the ‘Scandinavian’ burial rite are
attributable to simple status issues, particularly as it is clear that by no means all
members of the insular Scandinavian elite made use of the furnished burial rite,
let alone the weapon burial rite in particular. Across Man (Wilson 2008, 57–86),
Cumbria, the Danelaw (Bailey 1980) and even parts of Scotland (Lang 1976), the
erection of stone monuments provided an alternative mechanism for marking high-
status graves, albeit at a slightly later date, although there is increasing evidence for
some chronological overlap (Griffiths 2004, 133). Other 9th- and 10th-century insular
Scandinavian groups may have forgone high-status burial rites entirely, for reasons
more complex than simple poverty. Whatever the reasons for the process, it is clear
that all insular Scandinavian graves merged into the broader background of indig-
enous burial practices, very often within a single generation. The distribution of
weapon graves provides at best a partial picture of Scandinavian activity in these
islands. Nonetheless, as a relatively common burial rite, they provide a useful
evidence source which is in urgent need of review.
Some commentators will argue that the interpretation presented here forms part
of a general tendency to ‘sanitize’ the Viking Age by removing as many elements of its
inherent violence as possible, explaining weapons largely in terms of social prestige.
This is unintentional. The Viking Age was inherently violent, and the fact that
weapons seem to have functioned as one of the most important status symbols for
men in this period is by no means coincidental (Halsall 1998, 3–4). Osteological
evidence from Viking weapon graves is surprisingly limited, but of the 13 skeletons
which have been scientifically examined, two produced evidence that they met violent
deaths. One of these, ‘hit on the head and then killed by a massive cut into the head of
the left femur’ came from Repton (Biddle and Kjølbye-Biddle 2001, 61). The other,
‘shot by four arrows in his back, arm, belly and thighbone’, was buried at Westness,
Orkney, one of the few Viking burial sites to be consistently interpreted as a ‘family’
cemetery (Kaland 1995, 316). With the arguable exception of the burial at Balnakeil
(above), there can be little doubt that those buried with weapons had access to them,
or that they were prepared to use them on occasion. Whatever the evidence from
Copeland, Cumbria (Griffiths 2004, 129), there can be little doubt that most ‘Vikings’
inheriting land or authority in this period acquired it from individuals who had
themselves won it using at the very least the threat of violence. Nonetheless, the view
that these individuals were primarily, let alone exclusively, ‘warriors’ needs to be
called into question. Weapons were incorporated within their burial rituals, just as
they were carved on the Middleton Cross at a slightly later date, because they were
symbols of authority (Bailey 1980, 209–214). Broader arguments about perceptions of
the afterlife and religious belief are perhaps best left to another occasion, although it
MAM 37: 20-harrison – Press

316 maritime societies of the viking and medieval world

should be pointed out that the interpretation offered here does not in itself negate
such beliefs, or their importance to the construction of burial assemblages and graves
(Price 2010).
Fundamentally, it is time to reconsider the significance of these ‘Viking’ graves.
The limited number of graves containing weapons indicates that only a fraction of
the population was buried in this way. Moreover, they represent a response to local
insular conditions in terms of the number and type of weapons which they contain.
It seems very likely that most, if not all, are in some way related to the transfer of
property or power from one generation to the next, and at some level served as a
mechanism to ensure social stability. As such they were created by individuals, kin-
groups or communities who had vested local interests, and who used the furnished
burial rite as a means to further these interests. If these are ‘Viking’ graves, then these
‘Vikings’ were here to stay.1

note
1 Much of this paper is based on sections of a PhD carried out on behalf of the Irish Viking Graves
carried out under the supervision of Prof T B Barry Project, based at the National Museum of
at Trinity College, Dublin, and on subsequent Ireland, but the interpretation of this material
research as an IRCHSS Post-doctoral Research does not necessarily reflect the views of the Pro-
Fellow at the School of Archaeology, University ject. I am also grateful for the support of the
College Dublin. Some elements, particularly Medieval History Research Centre at Trinity
those relating to Dublin, are based on research College Dublin.

references
Anderson, J, 1874 ‘Notes on the relics of the Viking period of the Northmen in Scotland’,
Proceedings of the Society of Antiquaries of Scotland 10, 536–594
Anon, 1852 ‘Proceedings Monday, February 9th’, Proceedings of the Royal Irish Academy 5,
241–245
Bailey, R N, 1980 Viking Age Sculpture in Northern England, Collins, London
Batey, C E, 1995 ‘The Viking and late Norse graves of Caithness and Sutherland’, in C E Batey,
J Jesch and C D Morris (ed), The Viking Age in Caithness, Orkney and the North Atlantic:
Select Papers from the Proceedings of the Eleventh Viking Congress, Edinburgh University
Press, Edinburgh, 148–172
Bersu, G and Wilson, D M, 1966 Three Viking Graves from the Isle of Man, Society for Medieval
Archaeology, London
Biddle, M and Kjølbye-Biddle, B, 1992 ‘Repton and the Vikings’, Antiquity 46, 36–51
Biddle, M and Kjølbye-Biddle, B, 2001. ‘Repton and the ‘‘great heathen army’’, 873–4’, in
J Graham-Campbell, R Hall, J Jesch and D N Parsons (ed), Vikings and the Danelaw: Select
Papers from the Proceedings of the Thirteenth Viking Congress, Oxbow Books, Oxford,
45–69
Bjørn, A and Shetelig, H, 1940 Viking Antiquities in England with a Supplement of Viking
Antiquities on the Continent of Western Europe: Viking Antiquities in Great Britain and
Ireland 5, Aschehoug, Oslo
Bøe, J, 1940 Norse Antiquities in Ireland: Viking Antiquities in Great Britain and Ireland 3,
Aschehoug, Oslo
Brøgger, A W, 1929 Ancient Emigrants: A History of the Norse Settlements of Scotland, Oxford
University Press, Oxford
MAM 37: 20-harrison – Press

‘warrior graves’? 317

Cochrane, R, 1906 ‘Exhibit and description of bronze brooches and bowl found at Ballyholme,
Co. Down’, Journal of the Royal Society of Antiquaries of Ireland 36, 450–454
Davidson, H E, 1962 The Sword in Anglo-Saxon England, Boydell, Woodbridge
Dommasnes, L H, 1982 ‘Late Iron Age in western Norway: female roles and ranks as deduced from
an analysis of burial customs’, Norwegian Archaeological Review 15, 70–84
Downham, C, 2007 Viking Kings of Britain and Ireland: The Dynasty of Ívarr to ad 1014,
Dunedin, Edinburgh
Duffy, S, 1999 ‘The historical background’, in L Simpson (ed), Director’s Findings: Temple Bar
West, Temple Bar Properties, Dublin, 5–8
Dunwell, A J, Cowie, T G, Bruce, M F, Neighbour, T and Rees, A R, 1995 ‘A Viking Age cemetery
at Cnip, Uig, Isle of Lewis’, Proceedings of the Society of Antiquaries of Scotland 125,
719–752
Edwards, A J H, 1927 ‘Excavation of graves at Ackergill and of an earth-house at Freswick Links,
Caithness, and a description of the discovery of a Viking grave at Reay, Caithness’, Proceed-
ings of the Society of Antiquaries of Scotland 41, 196–207
Edwards, A J H, 1934 ‘A Viking cist-grave at Ballinaby, Islay’, Proceedings of the Society of
Antiquaries of Scotland 68, 74–78
Eldjárn, K, 1984 ‘Graves and grave goods: survey and evaluation’, in A Fenton and H Pálsson (ed),
The Northern and Western Isles in the Viking World: Survival, Continuity and Change,
Donald, Edinburgh, 2–11
Evison, V I, 1969 ‘A Viking grave at Sonning, Berks’, Antiquaries Journal 49, 330–345
Geake, H, 2003 ‘The control of burial practice in Anglo-Saxon England’, in M Carver (ed), The
Cross Goes North: Processes of Conversion in Northern Europe ad 300–1300, York Medieval
Press, York, 259–269
Graham-Campbell, J, 1980 ‘The Scandinavian Viking-age burials of England – some problems of
interpretation’, in P Rahtz, T Dickenson and L Watts (ed), Anglo-Saxon Cemeteries 1979,
British Archaeological Reports British Series 82, Archaeopress, Oxford, 379–382
Graham-Campbell, J, 2001 ‘Pagan Scandinavian burial in the central and southern Danelaw’, in
J Graham-Campbell, R Hall, J Jesch and D N Parsons (ed), Vikings and the Danelaw: Select
Papers from the Proceedings of the Thirteenth Viking Congress, Oxbow Books, Oxford,
104–124
Graham-Campbell, J and Batey, C E, 1998 Vikings in Scotland: An Archaeological Survey,
Edinburgh University Press, Edinburgh
Grieg, S, 1940 Viking Antiquities in Scotland, Viking Antiquities in Great Britain and Ireland 2,
Aschehoug, Oslo
Griffiths, D, 2004 ‘Settlement and acculturation in the Irish Sea region’, in J Hines, A Lane and
M Redknap (ed), Land, Sea and Home: Proceedings of a Conference on Viking-period
Settlement, at Cardiff, July 2001, Maney, Leeds, 125–138
Hadley, D M and Moore, J M 1999 ‘‘‘Death makes the man’’? Burial rite and the construction of
masculinities in the early Middle Ages’, in D Hadley (ed), Masculinity in Medieval Europe,
Longman, London, 21–39
Hagen, A, 1967 Norges Oldtid, Cappelen, Oslo
Halsall, G, 1998 ‘Violence and society in the early medieval west: an introductory survey’, in
G Halsall (ed), Violence and Society in the Early Medieval West, Boydell, Woodbridge, 1–45
Härke, H, 1989 ‘Knives in early Saxon burials: blade length and age at death’, Medieval Archae-
ology 33, 144–148
Härke, H, 1990 ‘‘‘Warrior graves’’? The background of the Anglo-Saxon weapon burial rite’, Past
and Present 126, 22–43
MAM 37: 20-harrison – Press

318 maritime societies of the viking and medieval world

Harrison, S H, 1995 Viking Age shield bosses in Dublin and the Irish Sea area, Unpublished MA
thesis, National University of Ireland (University College Cork), Cork
Harrison, S H, 2005 ‘College Green — a neglected ‘‘Viking’’ cemetery at Dublin’, in A Mortensen
and S V Arge (ed), Viking and Norse in the North Atlantic: Select Papers from the Proceedings
of the Fourteenth Viking Congress, Føroya Fróðskaparfelag, Tórshavn, 329–339
Harrison, S H, 2007 ‘Separated from the foaming maelstrom: landscapes of insular ‘‘Viking’’
burial’, in S Semple and H Williams (ed), Early Medieval Mortuary Practices, Anglo-Saxon
Studies in Archaeology and History 14, Oxford University School of Archaeology, Oxford,
173–182
Harrison, S H, 2008 Furnished Insular Scandinavian burial: artefacts and landscape in the early
Viking Age, Unpublished PhD thesis, University of Dublin (Trinity College), Dublin
Harrison, S H and Ó Floinn, R, 2014 Irish Viking Graves and Grave-Goods, National Museum of
Ireland, Dublin
Hencken, H O’N, 1950 ‘Lagore crannog: an Irish royal residence of the seventh to tenth centuries
ad’, Proceedings of the Royal Irish Academy 53C, 1–247
Henry, D (ed) 1995 Viking Ireland: Jens Worsaae’s Accounts of his Visit to Ireland 1846–47,
Pinkfoot, Balgavies
Hinton, D A, 2000 A Smith in Lindsey: The Anglo-Saxon Grave at Tattershall Thorpe, Lincoln-
shire, Society for Medieval Archaeology, London
Kaland, S H H, 1995 ‘The settlement of Westness, Rousay’, in C E Batey, J Jesch and C D Morris
(ed), The Viking Age in Caithness, Orkney and the North Atlantic: Select Papers from the
Proceedings of the Eleventh Viking Congress, Edinburgh University Press, Edinburgh,
308–317
Lang, J, 1976 ‘Hogback monuments in Scotland’, Proceedings of the Society of Antiquaries of
Scotland 105, 206–235
Nı́ Mhaonaigh, M, 2007 Brian Boru: Ireland’s Greatest King? Tempus, Stroud
O’Meara, J J (ed), 1982 The History and Topography of Ireland by Gerald of Wales, Penguin,
London
Parker Pearson, M, 1999 The Archaeology of Death and Burial, Sutton, Stroud
Paterson, C, Parsons, A J, Newman, R M, Johnson N and Howard Davis, C, 2014 Shadows in the
Sand: Excavation of a Viking-Age Cemetery at Cumwhitton, Cumbria, Lancaster Imprints,
Lancaster
Petersen, J, 1919 De Norske Vikingesverd: En Typologisk-Kronologisk Studie over Vikingetidens
Vaaben, Skrifter Utgit av Videnskapsselskapet i Kristiania 1919, Historisk-Filosofisk Klasse 1,
Dybwad, Kristiania
Pierce, I G, 2002 Swords of the Viking Age, Boydell, Woodbridge
Price, N, 2008 ‘Dying and the dead: Viking Age mortuary behaviour’, in S Brink with N Price, The
Viking World, Routledge, London, 257–273
Price, N, 2010 ‘Passing into poetry: Viking-age mortuary drama and the origins of Norse mythol-
ogy’, Medieval Archaeology 54, 123–156
Raftery, J, 1960 ‘A Viking burial from County Galway’, Journal of the Galway Archaeological and
Historical Society 29, 3–6
Richards, J D, 2000 Viking Age England, Tempus, Stroud
Richards, J D, 2003 ‘Pagans and Christians at a frontier: Viking burial in the Danelaw’, in
M Carver (ed), The Cross Goes North: Processes of Conversion in Northern Europe
ad 300–1300, York Medieval Press, York, 383–395
Richards, J D, 2004 ‘Excavations at the Viking barrow cemetery at Heath Wood, Ingleby,
Derbyshire’, The Antiquaries Journal 84, 23–116
MAM 37: 20-harrison – Press

‘warrior graves’? 319

Rowley-Conwy, P, 2007 From Genesis to Prehistory: The Archaeological Three Age System and its
Contested Reception in Denmark, Britain and Ireland, Oxford University Press, Oxford
Russell, I and Harrison S H, 2011 ‘Woodstown 6 — Viking age enclosed settlement and grave’, in
J Eogan and E Shee Twohig (ed), Cois tSiúire — Nine Thousand Years of Human Activity in
the Lower Suir Valley: Archaeological Excavations on the N25 Waterford City Bypass,
National Roads Authority, Dublin
Sheehan, J, 1988 ‘A re-assessment of the Viking burial from Eyrephort, Co. Galway’, Journal of
the Galway Archaeological and Historical Society 41, 60–72
Shetelig, H (ed) 1940 Viking Antiquities in Great Britain and Ireland, 5 vols, Aschehoug, Oslo
Shetelig, H, 1945 ‘The Viking graves in Great Britain and Ireland’, Acta Archaeologica 16, 1–53
Simpson, L, 2005 ‘Viking warrior burials in Dublin: is this the longphort?’, in S Duffy (ed),
Medieval Dublin 6, Four Courts, Dublin, 11–62
Skre, D, 2001 ‘The social context of settlement in Norway in the first millennium ad’, Norwegian
Archaeological Review 34, 1–12
Smith, G, 1933 ‘Talacre and the Viking grave’, Proceedings of the Llandudno, Colwyn Bay and
District Field Club 17, 42–49
Smith, J H, 1840 ‘An account of the discovery, in the month of November last, of a human skeleton
accompanied with weapons, ornaments, &c., interred on the sea shore, in the vicinity of
Larne, in the county of Antrim’, Proceedings of the Royal Irish Academy 2, 40–46
Solberg, B, 1985 ‘Social status in the Merovingian and Viking Periods in Norway from archaeo-
logical and historical sources’, Norwegian Archaeological Review 18, 61–76
Stylegar, F-A, 2010 ‘Hvorfor er det færre kvinne- enn mannsgraver fra vikingtiden i Norge?’,
Primitive Tider 12, 71–79
Thorsteinsson, A, 1968 ‘The Viking burial place at Pierowall, Westray, Orkney’, in B Niclasen
(ed), The Fifth Viking Congress, Føroya Fróðskaparfelag, Tórshavn, 150–173
van de Noort, R, 1993 ‘The context of early medieval barrows in western Europe’, Antiquity 67,
66–73
Wallace, P F, 2002 ‘Viking Age Ireland ad 850–1150’, in P F Wallace and R Ó Floinn, Treasures of
the National Museum of Ireland, Gill & Macmillan, Dublin, 213–256
Walsh, A, 1995 ‘Viking swords in Ireland’, Archaeology Ireland 9, 37–38
Walsh, A, 1998 ‘A summary classification of Viking Age swords in Ireland’, in H B Clarke, M Nı́
Mhaonaigh and R Ó Floinn (ed), Ireland and Scandinavia in the Early Viking Age, Four
Courts, Dublin, 222–235
Wamers, E, 1998 ‘Insular finds in Viking Age Scandinavia and the state formation of Norway’, in
H B Clarke, M Nı́ Mhaonaigh and R Ó Floinn (ed), Ireland and Scandinavia in the Early
Viking Age, Four Courts, Dublin, 37–72
Williams, H, 2010 Death and Memory in Early Medieval Britain, Cambridge, Cambridge Univer-
sity Press
Wilson, D M, 1976 ‘Scandinavian settlement in the north and west of the British Isles — an
archaeological point-of-view’, Transactions of the Royal Historical and Archaeological
Society (5th series) 26, 95–113
Wilson, D M, 2008 The Vikings in the Isle of Man, Aarhus University Press, Aarhus
Worsaae, J J A, 1847 ‘A review of the different descriptions of Danish and Irish antiquities, and of
several historical events associated with the invasion of Ireland’, Proceedings of the Royal
Irish Academy 3, 327–344
Worsaae, J J A, 1852 An Account of the Danes and Norwegians in England, Scotland and Ireland,
Murray, London
MAM 37: 21-jesch – Press

chapter 21
THE THREATENING WAVE: NORSE
POETRY AND THE SCOTTISH ISLES

By Judith Jesch

The maritime basis of the economy, culture and identity of the Scottish Isles goes
without saying. In an archipelagic environment, and before the advent of air travel in
the 20th century, much local communication and all more far-flung communication
inevitably happened by sea. In the Viking and medieval periods, the imperative for
long-distance communication arose from the political, economic and cultural links
the islands had with a wider Norwegian world, and maintaining such links required
long and often dangerous sea journeys. Thus, both the establishment and the
preservation of a Norse identity, as well as the local identities of the islanders in
relation to each other and to neighbouring regions, depended on successful conquest
of the sea. What I should like to explore in this paper are the ways in which this
conquest was verbalized and conceptualized.
Although sometimes neglected in current theoretical discussions of past iden-
tities, language is an important constitutive element of identity and, for at least one
linguist, ‘identity’ is fundamentally ‘a linguistic phenomenon’ (Joseph 2004, 11–14).
Although we no longer have direct access to Viking Age speakers, it is possible to
investigate their linguistic behaviour in a number of ways. The specialized linguistic
medium of poetry is an important vehicle for the formation of identities and the con-
ceptualization of a Norse world-view, in both the pre-literate and the literate periods.
Although much of the surviving poetry in Old Norse is associated with Iceland, it is
well known that Orkney was also a centre of literary endeavour in at least the 11th,
12th and early 13th centuries (Jesch 2005; 2013). There has however been little
detailed study of the ways in which this literature might express a more localized,
rather than a pan-Norse, or North Atlantic, identity or world-view, and little atten-
tion has been paid to poetic activity in other parts of Scandinavian Scotland. There is
a small, but not negligible, body of poetry that deals, precisely and in some detail,
with the perils and joys of sailing in the archipelagic waters of north Britain and thus
with the maritime identities of this region. Next to actually building a Viking ship
and sailing it, the best way of experiencing such a journey is to read some of the
poetry composed by those who did just that.
MAM 37: 21-jesch – Press

the threatening wave 321

ro˛ gnvaldr: sailor, poet and earl

Of course, the sailors of Viking and medieval Scotland did not restrict their
movements to north British waters, and the first relevant example is not directly
about sailing in Scotland, though it was composed by an Orcadian, Earl Ro˛gnvaldr
himself. The stanza is preserved in Orkneyinga saga which says he spoke it as he
sailed through the Mediterranean towards Byzantium (Jesch 2009, 607–608):
Rı́ðum Ræfils Vakri!
Rekuma plóg af akri!
Erjum úrgu barði
út at Miklagarði!
Þiggjum þengils mála!
Þokum framm ı́ gný stála!
Rjóðum gylðis góma!
Gerum rı́ks konungs sóma!
Let’s ride the Vakr khorsel of Ræfill ksea-kingl [SHIP]! Let’s not drive the plough
from the field! Let’s plough with a drenched prow out to Constantinople! Let’s
receive the wages of the prince! Let’s move forward in {the din of weapons}
[BATTLE]! Let’s redden the gums of the wolf! Let’s create the honour for the
powerful king!
This stanza sums up the character of Ro˛gnvaldr’s poetry, showing its combina-
tion of tradition and innovation. It is in the traditional skaldic metre of dróttkvætt
and largely follows the rules of this form. However, instead of the usual internal
rhyme, it has end-rhyme, which is further emphasized by the fact that seven of its
eight lines are end-stopped. Its imagery also combines the traditional with the inno-
vatory. Ræfils Vakr is a traditional type of kenning for ‘ship’, which relies on the
audience knowing that Ræfill is the name of a sea-king and Vakr the name of a
legendary horse. The ‘sea-king’s horse’ or ‘horse of the sea’ is thus a ‘ship’ and
naturally what one does with a horse is to ride it. This image is extremely common in
skaldic verse but what is new in Ro˛gnvaldr’s stanza is the comparison of the bold
equestrian/seafarer with the stay-at-home ploughman. Just as the ploughman cuts
through the soil with his plough, so the sailor furrows the sea with the prow of his
ship.1 This is a very ancient comparison found in many cultures and literatures but
oddly not much in Old Norse poetry, perhaps because Iceland is not a great country
for ploughing. But Ro˛gnvaldr, born in southern Norway and living in Orkney, would
have been more familiar than most Icelanders with the visual juxtaposition of a field
being ploughed while a ship furrowed the sea in the distance. The stanza goes on to
state the intention that Ro˛gnvaldr and his men have of going into battle on behalf of a
prince who will pay them. Such a comparison between the bold warrior or sailor and
the stay-at-home wimp is common in Norse poetry. What is innovatory here is the
extension of it to a much more visual comparison of sailing and ploughing which is at
the same time both literary and yet derived from personal experience. In this way, a
stanza about sailing in the Mediterranean is nevertheless rooted in the poet’s own
mental imagery of his homeland(s) and home waters.
MAM 37: 21-jesch – Press

322 maritime societies of the viking and medieval world

Such literary jeux d’esprit about far-flung journeys are less common in
Ro˛gnvaldr’s poetry about sailing in the waters of the British Isles. Here, a more
realistic mode prevails, as in this stanza describing the unfortunate wreck of two
ships newly given to Ro˛gnvaldr by the king of Norway (Jesch 2009, 585):
Brast, þás bæði lesti
— bauð hro˛nn skaða mo˛nnum —
— sút fekk veðr it váta
vı́fum — Hjo˛lp ok Fı́fu.
Sék, at sjá mun þykkja
snarlyndra fo˛r jarla
— sveit gat vás at vı́su
vinna — ho˛fð at minnum.
There was a loud noise when both Hjo˛lp and Fı́fa were damaged; the wave caused
men harm; the wet weather gave women sorrow. I see that that voyage of bold-
hearted jarls will be kept in memory; the crew got drenching work for sure.
This is no joyous triumphal voyage to battle but a rather sorry end on the rocks
of Shetland. The stanza is introduced in the prose of the saga by a detailed and visual
account of the voyage, giving information about the weather, the visibility and even
the days of the week (Guðmundsson 1965, 195; my translation):
Þat var þriðjudagskveld, er jarlar létu ı́ haf, ok sigldu allgóðan byr um náttina.
Miðvikudag var stormr mikill, en um náttina urðu þeir við land varir; þá var myrkr
mikit; þeir sá boðaslóðir o˛llum megin hjá sér. Þeir ho˛fðu áðr samfloti haldit. Þá var
engi kostr annarr en sigla til brots báðum skipunum, ok svá gerðu þeir. Þar var urð
fyrir, en lı́tit forlendi, en hamrar it efra. Þar heldusk menn allir, en týndu fé miklu;
sumt rak upp om náttina.
It was a Tuesday evening when the earls put out to sea, and they sailed through the
night with an excellent wind. On Wednesday there was a great storm and during
the night they became aware of land nearby; it was very dark then; they saw
breakers around them on all sides. They had previously sailed close together. Then
there was no choice but to wreck both ships, and they did so. There was a rocky
beach ahead, with little foreshore, and cliffs above. All were saved, but much cargo
was lost; some of it came ashore during the night.
The poetry and prose together provide a detailed description of the scene and the
wreck, with all its noise and drench, that is surely based on direct experience.
Subsequent events in the saga suggest it all happened somewhere near Gullberwick in
Shetland, and this was confirmed by an interdisciplinary research project in the 1970s
which failed to locate the wreck but did establish its rough location with a high
degree of plausibility (Collings et al. 1974–77; Morrison 1973).
The stanza itself encapsulates an experience of living in the Northern Isles that
goes beyond the moment of the shipwreck. There are no kennings in this stanza, no
complex literary tricks aimed at a rootless elite. Instead, there is reference to com-
munity, memory and the physical environment. At the head of the community are the
two earls, Ro˛gnvaldr and Haraldr Maddaðarson, returning to Orkney from Norway.
Their shipwreck causes sorrow to women — ostensibly the two ships, both of which
MAM 37: 21-jesch – Press

the threatening wave 323

have grammatically feminine names, but shipwrecks generally cause sorrow to

human women. Although no one died in this particular shipwreck, it cannot have
been a pleasant experience for the crews, who also get a mention, with commiser-
ation both for the damage caused to them and for the hard and wet work they
endured. The stanza refers to the event as something to be kept in memory, thus
alluding to the significant role of poetry in the community, preserving knowledge of
past events. In this stanza we also see how maritime events are closely tied to local
and regional identities, for although Ro˛gnvaldr and Haraldr were on their way to
Orkney, Shetland was very much in their dominion and the shipwreck ironically a
part of their coming home.
Luckily there were no fatalities in the Shetland shipwreck and much of the rest
of Ro˛gnvaldr’s poetry in this episode is jocular in tone, though still with an
undercurrent of danger, as in this stanza in which he bemoans the state of his clothing
after the shipwreck (Jesch 2009, 586–587):
Skekk hér skinnfeld hrokkinn;
skrauts mér afar lı́tit;
stórrs, sás stendr of órum,
stafnvo˛llr, yfirho˛fnum.
Nærgis enn af úrgum
álvangs mari go˛ngum
— brim rak hest við hamra
húns — skrautligar búnir.
I shake out here a wrinkled leather garment; it provides me with very little finery;
the prow-field [SEA] which surrounds our outerwear is big. Someday we’ll go more
finely dressed from a spray-swept horse of the eel-plain [SEA > SHIP]; surf drove
the stallion of the mast-head [SHIP] onto cliffs.
This stanza reiterates the message of the previous one that the surf drove the
ship onto the rocks and again uses the traditional image of the ship as a horse of
the sea. The image of the sea surrounding the sailors (and their garments) picks up on
the prose description of the shipwreck, in which the mariners and their ships were
surrounded by breakers. The ironic reference to Ro˛gnvaldr’s damaged clothing and
the contrastive anticipation of his future finery alludes to his identity as earl, at the
head of the community, and the necessity for such a leader to have appropriate
clothes as well as working ships. The shipwreck is a low point, both physically and
politically, from which things can only get better.

icelanders in the hebrides

While there are a number of other stanzas by Ro˛gnvaldr and his associates that
describe sailing, they are about sailing in southern waters, on their cruise to and from
the Holy Land, and will not be considered here. However, there are further insights
into sailing in Scottish waters in the stanzas composed by some very lost Icelanders
in ad 1202. In that year the Icelandic bishop-elect Guðmundr Arason invited the
physician Hrafn Sveinbjarnarson to join him on his consecration voyage to Norway.
MAM 37: 21-jesch – Press

324 maritime societies of the viking and medieval world

After some difficulties in clearing Iceland, and driven off course towards Ireland,
their ship eventually made Scotland at Stour Head in Sutherland. Attempting to
round Cape Wrath, they experienced a strong southerly wind and a rogue wave
which threatened to swamp them from the rear.2 They were miraculously rescued
when the bishop’s blessing caused the ship to swing around to meet the wave head-
on. This event is recorded in a stanza by Grı́mr Hjaltason (Helgadóttir 1987, 19,
102–103):
Eisandi veðr undan
uðr, nú er hvasst ór suðr,
stœrir sterkar bárur,
starf erat smátt fyrir Hvarfi.
Kløkkr verðr kjo˛lr, en rakkan
kemr hregg ı́ stað seggjum.
Nu eru fjo˛ll á sjá sollin.
Súð gengr æ sem prúðast.
The wave goes rushing on; now it blows sharp from the south; strong waves grow
big; there is no small labour (for us) off Cape Wrath. The keel becomes yielding;
the storm drives men into a steep place — now the mountains on the sea are
swollen. The ship goes all the time as finely as can be.
The travellers continued to sail south through the storm until they reached some
narrow waters where they could see land, and breakers on both sides, which they
recognized as the Hebrides. The bishop asked Hrafn to navigate, which he did after
some modest demurring. The saga cites a number of stanzas by three poets describing
the voyage as Hrafn successfully navigates them through the islands and skerries at
night.
Hrafn eventually piloted them to a safe harbour which the saga calls Sandey and
is most likely, of all the similarly named islands in Scotland, to have been the small
island of Sanday that is attached by a land-spit to Canna in the Small Isles of the Inner
Hebrides (Helgadóttir 1987, 76; Power 2005, 41–43). Here, the saga tells how an
official of the local ruler (a certain King Óláfr) attempted to collect a landing-tax
from them, which they did not wish to pay. When the travellers went to church, the
king himself invited the bishop-elect to dine with him and then refused to let him and
Hrafn go until they had paid the tax. The ship’s crew took up weapons to rescue the
hostages, but eventually they agreed to pay a tax, though less than the Hebrideans
had demanded, and sailed away for Norway. This was clearly a memorable journey
for the Icelanders, both for the difficult sea voyage itself, and for the contrast between
their relief at finally arriving at a safe harbour and the rather grasping reception they
were given by the local authorities there. Their personal experiences of the voyage are
captured in the stanzas cited in the saga.
Some of these stanzas are most likely to have been composed retrospectively.
Two stanzas attributed to Guðmundr Svertingsson come from a drápa, a long praise-
poem that he composed about Hrafn, which was probably posthumous, as is sug-
gested by the fact that the verbs in both stanzas are predominantly in the past tense
(Helgadóttir 1987, 20–21, 103–104):3
MAM 37: 21-jesch – Press

the threatening wave 325

Farar vanda hygg ek fundusk,

fyrirsagnir þraut bragna.
Þjóð var þá við leiði
þreytin segls at neyta.
Blátt var um borð at lı́ta:
brekafall vega alla,
bræddr þar er bárum ruddi
barðjór, ok skergarðar.
I think travelling perils were met; men ran out of words of command. People were
hard pressed then to make use of the sail in the following wind. Blackness was to
be seen over the side: fall of breakers and skerry reefs in all directions, there where
the tarred prow-horse [ship] cleared a way through the waves.
Gekk á grœðis blakki
greppr um nótt við ótta,
Hrafn varð, sem ek get, gumnum
gagnsmaðr, til leiðsagnar.
Heggr fekk hváru tveggja
hauktorgs vita borgit,
sýlði segl ok bulði
svo˛l hro˛nn, skipi ok mo˛nnum.
The man (Hrafn) undertook the piloting on the sea’s horse [SHIP] at night in the
face of fear — as I say, Hrafn turned out to be a beneficial man to men. The
hawthorn-tree of fire of hawk-place [OF HAND OR WRIST > GOLD >
WEALTHY MAN (HRAFN)] succeeded in saving both, ship and men — the sail
was stiff with rime and the chill wave boomed.
Hrafns saga preserves 11 stanzas of this poem in all, covering different episodes
in its hero’s life, and it is therefore unlikely that these two were composed at the time
of the voyage. Indeed, the saga introduces them with a formula (Þessa getr . . . ‘This is
told in . . .’) which indicates that they are taken from a pre-existing poem with a
retrospective account of events. Moreover, there is no indication in the saga-text that
Guðmundr actually went on the voyage.
Eyjólfr’s half-stanza is also in the past tense (Helgadóttir 1987, 21, 104):
Báru austr frá Íra
ættlandi skæ branda
hregg, áðr himna tyggi
heitbyr firum veitti.
Storms carried the horse of prow-planks [SHIP] eastward from the homeland of
the Irish before the lord of the heavens granted men a desired wind.
In this case, however, the saga does not record that it is part of a longer poem,
but presents Eyjólfr as declaiming the verse on the spot. Thus, the past tense is pre-
sented as the immediate past tense of their difficult voyage rather than a more distant
past viewed from a few years later. Eyjólfr appears nowhere else in the saga (unless he
is identical to another minor character in the saga with the same common name).
This stanza is thus more likely to have been composed and performed in the
Hebrides, during the voyage.
MAM 37: 21-jesch – Press

326 maritime societies of the viking and medieval world

Two-and-a-half stanzas by Grı́mr are by contrast predominantly in the present

tense. The first of these has already been presented above, the others follow on from
Eyjólfr’s (Helgadóttir 1987, 21–22, 104–105):
Sér á sigling vára
suðreysk kona, þuðri
súð gørvask nú nauðir,
námgjo˛rn, er hrýðr stjo˛rnum.
The Hebridean woman, curious, watches our sailing, as the stars are cleared (from
the sky = as the dawn breaks); hardships are now created for the thin-built ship.
Hér hefir beitt at brattri
Bótólfr skipi fljótu,
áðr fell sær um súðir,
Sandeyju, skæ branda.
Reisti sjálfr, ok sýsti,
snarr félagi harra
hafnarmark, fyrir hrefnis
happsverk gota sterkan.
Bótólfr has tacked the swift ship here to steep Sandey. Earlier the sea fell over the
horse of prow-planks [SHIP]’s sides. The active partner of the king himself raised
up a harbour-mark and did a good deed for strong-built horse of the ship’s strake
[CRAFT].
Both of Grı́mr’s full stanzas are presented in the saga-narrative as his spon-
taneous, extempore compositions at the time of the voyage. One could of course
question whether Grı́mr really had the leisure to compose and recite a stanza about
the weather when all hands were needed on deck to get the ship safely around Cape
Wrath. Even more questionable is whether anyone on board ship would have
bothered to listen to him when they had more important things on their minds. It is
more convincing to imagine him composing the first stanza at leisure when the ship
had reached its safe harbour and the crew were recovering. Together with his other
two stanzas (and possibly further ones which no longer survive), it would have made a
dramatic performance at an after-dinner entertainment. In this context the present
tense could be explained as a stylistic use of the historic present to give immediacy to
the story in performance and to help the listeners imagine the recent situation. There is
some further evidence for this in Grı́mr’s use of the deictic adverb hér ‘here’ in his
third stanza, given added prominence by its initial position. This usage indicates the
location of the performance which, in this context, was probably Sandey. It could of
course have applied because Grı́mr composed and performed his poem on the island
to an audience of his Icelandic fellow-travellers. However, he also makes reference, in
his previous stanza, to a Hebridean woman watching the progress of their sailing; such
a reference would have had added meaning if the poem was composed for a wider
audience including at least some locals. Moreover, the use of adverbial áðr ‘previ-
ously, earlier’ in Grı́mr’s third stanza suggests a performance time-frame not long
after the events described. While Eyjólfr’s stanza does not have the same immediacy as
those by Grı́mr, it is linked to them in that both poets use the kenning skær branda
‘horse of prow-planks’ for ‘ship’. From all this one could speculate that a performance
MAM 37: 21-jesch – Press

the threatening wave 327

on Sandey had something of the competitive about it, with several poets giving their
versions of the dramatic events, possibly as a form of after-dinner entertainment in the
local chieftain’s hall.
The saga then goes on to quote one further stanza by Grı́mr which it claims he
recited after their safe arrival in Norway, though this is far from clear from the stanza
itself (Helgadóttir 1987, 22–23, 105):
Hér náðum val vı́ðis,
vı́glundr, með Guðmundi,
sterkr, at sto˛ðva merki
stefna biskupsefni.
Frágum áðr á Eiði
einni nótt fyrir dróttins,
trauðr mun glaum at gœða
grams herr, bana Sverris.
Here, battle-tree [STRONG, WARLIKE MAN], we succeeded in steering the horse
of ocean [SHIP] with Bishop-elect Guðmundr to the harbour-mark. Earlier at Eið,
one night before the Lord’s (night), we learnt of the death of Sverrir; the king’s
people will be reluctant to make merry.
It is not inconceivable that the Icelanders heard of the Norwegian king’s death in
the Hebrides, and other aspects of the stanza link to the previous stanzas and their
supporting anecdote. Places called Eið (‘isthmus’) abound in the Hebrides as well as
in Norway (eg Stahl 2000, 102). It would indeed be a good name for the land-spit that
links Canna and Sanday. The stanza also echoes Grı́mr’s previous stanzas in empha-
sizing the successful steering of the ship and in the reference to a harbour-mark. The
reference to the Lord’s night also links to their church-going in the anecdote.4 Even if
the first half-stanza refers to their arrival in Norway, the retrospective second half
could still refer to the Hebrides. Neither Grı́mr nor Eyjólfr play any further part in
Hrafns saga, making the Sandey anecdote (including Grı́mr’s final stanza, wherever it
relates to) self-contained within the narrative.
Overall, the internal evidence of the stanzas by Grı́mr and Eyjólfr suggests that
they were composed during the journey, probably for recitation in the Hebrides, and
this is supported by indications in the surrounding prose anecdote. The anecdote
itself seems to represent a narrative framework provided for the poems when they
were performed subsequently in Iceland, as they undoubtedly were, being a good
travellers’ tale, though by whom is not clear. This in its turn provided the infor-
mation that was recycled by Guðmundr Svertingsson in his posthumous poem in
praise of Hrafn, and this poem was subsequently extracted into the anecdote when
the saga was put together. This postulated history of the poems leads to later 13th-
century Iceland, but shows how they originated in an early 13th-century Icelandic
confrontation with Hebridean maritime and cultural identities.

the threatening wave

If the hypothesis that Grı́mr’s first stanza (along with the others) was in fact
composed and performed on Sandey, to a Hebridean-Norse audience, is correct, it
MAM 37: 21-jesch – Press

328 maritime societies of the viking and medieval world

highlights the Icelander’s poetical interpretation of local seafaring conditions. Of

particular interest is the dramatic description of the wave which nearly swamped
the ship and the poet’s comparison of it to mountains. In at least one of the manu-
scripts in which this poem is preserved, a saga about Bishop Guðmundr, the episode
has a chapter-heading which refers to this phenomenon as hafgerðingar (Karlsson
1983, 141).
According to a 13th-century Norwegian manual of instruction, hafgerðingar are
a strange maritime phenomenon in the waters of Greenland (Holm-Olsen 1983,
27–28, my normalization; Larson 1917, 137–138):
Nú er þat enn eitt undr ı́ Grœnalandshafi er ek em eigi fróðastr um með hverjum
hætti er þat er. Þat kalla menn ‘hafgerðingar’ en þat er þvı́ lı́kast sem allr hafstormr
ok bárur allar þær sem ı́ þvı́ hafi eru samnisk saman ı́ þrjá staði ok gerask af þvı́
þrjár bárur. Þær þrjár gerða allt haf svá at menn vitu hvergi hlı́ð á vera ok eru þær
stórum fjo˛llum hærri, lı́kar brattum gnı́pum. Ok vitu menn fá dœmi til at þeir
menn hafi ór hafum komisk er þar hafa ı́ verit staddir, þá er þessi atburðr hefir
orðit. En þvı́ munu so˛gur vera af go˛rvar at guð man æ no˛kkura þaðan hafa frjálsat,
þá sem þar hafa verit staddir, ok mun þeira rœða sı́ðan dreifsk hafa ok fluttisk
manna ı́ millum. Hvart sem nú er svá fra sagt sem þeir hafa helzt um rœtt eða er
no˛kkut þeira rœða aukin eða vanat, ok munu vér firi þvı́ varliga um þá hluti rœða
at vér ho˛fum fá þá hitta nýliga er þaðan hafi komisk ok oss kynni þessi tı́ðendi at
segja.
Now there is still another marvel in the seas of Greenland, the facts of which I do
not know precisely. It is called ‘sea-hedges’ [hafgerðingar], and it has the appear-
ance as if all the waves and tempests of the ocean have been collected into three
heaps, out of which three billows are formed. These hedge in the entire sea, so that
no opening can be seen anywhere; they are higher than lofty mountains and
resemble steep, overhanging cliffs. In a few cases only have the men been known
to escape who were upon the seas when such a thing occurred. But the stories of
these happenings must have arisen from the fact that God has always preserved
some of those who have been placed in these perils, and their accounts have after-
wards spread abroad, passing from man to man. It may be that the tales are told as
the first ones related them, or that stories may have grown larger or shrunk some-
what. Consequently, we have to speak cautiously about this matter, for of late we
have met but very few who have escaped this peril and are able to give us tidings
about it.
This description has some similarities with Grı́mr Hjaltason’s stanza. Unlike
Grı́mr, the author makes it clear that he knows this phenomenon only from hearsay
and so the description must be taken with a pinch of salt; it has been called ‘rather
improbable’ (Benediktsson 1981, 27). The passage has a comparison of waves to
mountains and a sense of extreme danger in common with Grı́mr’s stanza and its
accompanying anecdote in Hrafns saga. Though there are a number of kennings in
which waves are compared to mountains of various sizes and shapes (Meissner 1921,
99), the texts considered here go beyond the metaphorical and extend this com-
parison in a descriptive way.
There is another reference to hafgerðingar in Landnámabók, the Icelandic Book
of Settlements, which mentions an unnamed Christian Hebridean who accompanied
MAM 37: 21-jesch – Press

the threatening wave 329

the Greenland settler Herjólfr in the late 10th century. The Hebridean composed a
poem called Hafgerðingadrápa, from which we deduce that the journey referred to in
it is a sea-journey (Benediktsson 1968, 132–134; my translation):
Mı́nar biðk at munka reyni
meinalausan farar beina;
heiðis haldi hárar foldar
hallar dróttinn yfir mér stalli.
I pray to the faultless tester of monks [GOD] to aid my journey; may the lord of the
high hall of the earth [HEAVEN] hold his hawk’s rest [ARM/HAND] over me.
It is possible that the fragments by the Hebridean have been misattributed in
Landnámabók and that Hafgerðingadrápa should in fact be associated with the
Hebrides rather than Greenland, as already suggested by the fact that the poet is said
to have been a Hebridean. Landnámabók does not actually link the composition of
the poem with Greenland, it merely says that the Christian Hebridean who composed
it was on Herjólfr’s ship. Furthermore, the poem is in hrynhent, a metre which other-
wise first appears in the mid-11th century, in the poetry of Arnórr jarlaskáld, an
Icelandic poet who also worked for the earls of Orkney and spent time in these
islands (Whaley 1998, 79–80). For this reason the poem is unlikely to be as early as
the 10th century, as implied by Landnámabók. The confusion could be explained by
the fact that the sailing routes to both Greenland and the Hebrides are dominated by
a landmark known to the Norse as Hvarf (now Capes Farewell and Wrath). Sailing
around Cape Wrath is often a difficult matter, as attested by Grı́mr’s stanza. I would
therefore suggest that Hafgerðingadrápa is about sailing to the Hebrides, that it dates
to at least the mid-11th century or later and also that it be added to the small corpus
of poetry about sailing in north British waters.5
The stanzas discussed so far are mainly from the 12th and 13th centuries. Unfor-
tunately there is little earlier evidence from skaldic verse for Norse sailings in north
British waters, with the exception of a stanza by Arnórr jarlaskáld in the 11th cen-
tury. Apart from a clear Orcadian connection, this stanza is hard to place in a histo-
rical context. The first half (not cited here) refers to the enmity between Ro˛gnvaldr
Brúsason and his rival Þorfinnr and, as Arnórr composed poems in praise of both, it is
unclear who the ‘mighty ruler’ mentioned in the second half-stanza is. But here again
the imagery is that of a large wave threatening to engulf a rather battered ship
(Whaley 2009, 251–252):
Sleit fyr eyjar útan
allvaldr bláu tjaldi;
hafði hreggsvo˛l dúfa
hrı́mi fezk of lı́ma.
The mighty ruler wore to shreds the dark awnings out beyond the islands; the
snow-cold billow had fastened itself in frost about the mast.
The ‘islands’ mentioned here could be any nameless islands, but they are more
likely to be the Eyjar, a term commonly used for Orkney, or the Northern Isles more
generally (though rarely or never for the Suðreyjar or Hebrides). In connection with
MAM 37: 21-jesch – Press

330 maritime societies of the viking and medieval world

the term allvaldr ‘mighty ruler’, the term suggests the dominions of the earls of
Orkney, which they rule and defend through their conquest of the rime-cold sea and
its threatening waves.

epilogue
Though it is not the purpose of this paper to explore Norse-Gaelic literary
contacts, it is worth noting a few parallels between the poetry discussed here and the
Gaelic poem usually known as ‘The MacSween Poem’. This is a set-piece ‘describing
the preparations for, and the outcome of, a sea-borne expedition to take possession
of Castle Sween’ in Argyll (Meek 1997, 6). As well as identifying the followers of John
MacSween as Lochlannaigh ‘Norsemen’, the poem contains a number of Norse loan-
words, particularly of nautical terms, and some motifs and structural elements for
which Meek finds closer analogues in Scandinavian than in Celtic culture. As a result,
he would place it in a ‘Norse-Gaelic context of the late thirteenth or early fourteenth
century’ (Meek 1997, 27). The poem has some interesting points of contact with the
Norse verses discussed here.6 Thus, the very first stanza, describing the assembling
of the fleet, refers to the ships ‘being cleansed for horsemen who would travel the
waves’ (st 1). There are frequent references to turbulent seas rising high against the
ships (sts 12, 13, 16) and an explicit comparison with mountains in ‘the wave will test
them in an ocean of summits’ (st 15). If the MacSween poem does not show more
specific parallels with those discussed here, it is in part because only a minority of its
stanzas are actually about sailing. Another factor may have been that its author was
attempting to strike a balance between his patrons’ Norse ancestry and conventions,
and their Gaelic/Irish pedigree (Meek 1997, 19, and pers comm).

conclusion
The poetry discussed in this paper presents a range of responses to sailing
around the northern parts of the British Isles, by poets more or less familiar with
these routes but also with Norway and Iceland and, in Ro˛gnvaldr’s case, with sea-
ways much further afield. These poets use traditional forms and imagery to express
the pan-Norse identity of north Britain. At the same time they also give voice to what
is distinctive about this region — its communities, its rulers, its language, its land-
scapes and seascapes, and, most particularly, the special challenges of sailing in these
waters. I would go so far as to suggest that there is also a special subgenre of poetry
about difficult sailing in British waters, and that this was often conceptualized as a
large wave which threatens to overwhelm the ship, as seen in several of the examples
cited above. Furthermore, an important task of poetry was to show how the success-
ful rulers of the region conquered this maritime threat. The Norse poetry considered
here expresses the specific maritime identities of Scandinavian Scotland, using the
cultural frameworks of the broader Viking diaspora, while the MacSween poem
hints at some of the ways in which these identities persisted into the late 13th or early
14th century and crossed into a new language.
MAM 37: 21-jesch – Press

the threatening wave 331

notes
1 This comparison works, I think, both with a approaching wave. I thank Donald Meek for
mould-board plough, which turns the soil over, explaining this to me.
and the simpler ard, which just cuts through the 3 In presenting Guðrún P Helgadóttir’s English
soil (I thank Niall Sharples for explaining such versions, I have incorporated her explications of
basic practical matters to me). Ro˛gnvaldr uses the kennings into her translations.
newer noun plóg in the second line, while the verb 4 Canna had several medieval churches and ecclesi-
erja in the third line is cognate with the older astical sites, see NMRS nos NG20NE 1, 2 and 13,
noun arðr. The verb also occurs, again referring and NG20SW 2 in the CANMORE database of
to a ship, in the runic inscription on the Djulefors The Royal Commission on the Ancient and
stone (Jesch 2001, 177), a century or so earlier Historical Monuments of Scotland.
than this stanza. 5 The poem is similarly dated to the 11th or even
2 The episode illustrates the dangers of what sailors 12th century by Jakob Benediktsson (1981),
call ‘pooping’, when a great wall of water wells though for very different reasons, and he would
up astern and, at worst, comes crashing down on still link it to stories about Greenland.
the ship. At best, the ship could ride up stern- 6 I cite the English translation (Meek 1997, 36–39)
wards and the sea would pass below. A vessel that by stanza number to facilitate comparison with
was very fine-lined aft would not have the buoy- the Gaelic texts of the 15th-century manuscript
ancy to ride upwards, while one whose lines and Meek’s (1997, 31–36) proposed restoration.
were too sharp forward would not rise to the

references
Benediktsson, J (ed), 1968 Íslendingabók. Landnámabók, Hið ı́slenzka fornritafélag, Reykjavı́k
Benediktsson, J, 1981 ‘Hafgerðingadrápa’, in U Dronke, G P Helgadóttir, G W Weber and
H Bekker-Nielsen (ed), Speculum norroenum. Norse Studies in Memory of Gabriel Turville-
Petre, Odense University Press, Odense, 27–32
Collings, L, Farrell, R and Morrison, I, 1974–77 ‘Earl Rögnvald’s Shipwreck’, Saga-Book 19,
293–310
Guðmundsson, F (ed), 1965 Orkneyinga saga, Hið ı́slenska fornritafélag, Reykjavı́k
Helgadóttir, G P (ed), 1987 Hrafns saga Sveinbjarnarsonar, Clarendon Press, Oxford
Holm-Olsen, L (ed), 1983 Konungs skuggsiá, Norsk historisk kjeldeskrift-institutt, Oslo
Jesch, J, 2001 Ships and Men in the Late Viking Age: The Vocabulary of Runic Inscriptions and
Skaldic Verse, The Boydell Press, Woodbridge
Jesch, J, 2005 ‘The literature of medieval Orkney’, in O Owen (ed), The World of Orkneyinga
Saga, The Orcadian, Kirkwall, 11–24
Jesch, J, 2009 ‘Ro˛gnvaldr jarl Kali Kolsson’, in K E Gade (ed), Poetry from the Kings’ Sagas 2:
From c. 1035 to c. 1300, Brepols, Turnhout, 575–609
Jesch, J, 2013 ‘Earl Rögnvaldr of Orkney, a poet of the Viking diaspora’, in A Jennings and
A Sanmark (ed), Across the Sólundarhaf: Connections between Scotland and the Nordic
World. Selected Papers from the Inaugural St. Magnus Conference 2011, Journal of the North
Atlantic Special Volume 4, 154–160
Joseph, J E, 2004 Language and Identity: National, Ethnic, Religious, Palgrave Macmillan,
Basingstoke
Karlsson, S (ed), 1983 Guðmundar sögur biskups, C A Reitzel, Copenhagen
Larson, L M (trans), 1917 The King’s Mirror, The American-Scandinavian Foundation, New York
Meek, D E, 1997 ‘‘‘Norsemen and noble stewards’’: the MacSween poem in the Book of the Dean
of Lismore’, Cambrian Medieval Celtic Studies 34, 1–49
Meissner, R, 1921 Die Kenningar der Skalden, Kurt Schroeder, Bonn
Morrison, I, 1973 The North Sea Earls: The Shetland/Viking Archaeological Expedition, Gentry
Books, London
MAM 37: 21-jesch – Press

332 maritime societies of the viking and medieval world

Power, R, 2005 ‘Meeting in Norway: Norse-Gaelic relations in the Kingdom of Man and the Isles,
1090–1270’, Saga-Book 29, 5–66
The Royal Commission on the Ancient and Historical Monuments of Scotland, CANMORE
Database, http://www.rcahms.gov.uk/search.html
Stahl, A-B, 2000 ‘Norse in the place-names of Barra’, Northern Studies 35, 95–112
Whaley, D, 2009, ‘Arnórr jarlaskáld Þórðarson’, in K E Gade (ed), Poetry from the Kings’ Sagas 2:
From c. 1035 to c. 1300, Brepols, Turnhout, 177–281
MAM 37: 22-mcdonald – Press

chapter 22
SEA KINGS, MARITIME KINGDOMS AND
THE TIDES OF CHANGE:
MAN AND THE ISLES AND MEDIEVAL
EUROPEAN CHANGE, ad c1100 – 1265

By R Andrew McDonald

Three islands lie off the shores of Britain, in different directions but all of about the
same size: the Isle of Wight to the south, Anglesey to the west and the Isle of Man
to the north-west. The first two are near the coast, being divided from it by narrow
straits of sea-water. The third, the Isle of Man, is much further away, lying mid-
way between Ulster in Ireland and Galloway in Scotland. (Thorpe 1978, 186–187)
Accounts and descriptions of the Isle of Man before the early modern period are
few and far between. Those that do exist, including the remarks of Gerald of Wales
(c1200) in his Journey Through Wales, cited above, almost always comment upon the
Island’s unique nodal location in the centre of the Irish Sea. One of the more eloquent
of these expressions came from the pen of the 16th-century English antiquarian John
Leland (d 1552), who observed that the Isle of Man is ‘situated in a navel of the sea
between Ireland and Britain’ (Leland 1770, iii, 223). It is an apt description. On a
clear day, the view from Snaefell, the Isle of Man’s highest peak, encompasses not
only the whole island itself, but also the surrounding landmasses of Ireland to the
west, Scotland to the north, England to the east and Wales to the south. There is no
evidence that Gerald of Wales ever visited the Isle of Man, but its intervisibility with
the surrounding landmasses would not have come as a surprise to him. He was well
acquainted with the geography, climate and sailing conditions of the Irish Sea, and
commented in several places about how landmasses on one side were visible in
favourable conditions from the other (O’Meara 1982, 33). Some further insight into
the situation of Man may be gained from John Speed’s map of ad 1605/10, which
depicts a greatly exaggerated Isle of Man nearly filling the Irish Sea and ringed by the
coasts of Ireland, Scotland, England and Wales (Cubbon 1974, 10–12, 24–25)
(although, in fairness to Speed, Man was drawn to a different scale than the sur-
rounding coastlines, a fact that is often overlooked).
Given its intervisibility from surrounding landmasses and strategic location on
the western seaways, as well as its good natural harbours and abundance of fertile
MAM 37: 22-mcdonald – Press

334 maritime societies of the viking and medieval world

land, it is hardly surprising that despite its small size, Man has been described as a
‘cultural crossroads’, open to influences from the surrounding lands (Davey 2002).
The interaction of different cultures thus represents a dominant theme in Manx
prehistory and history alike (see Davey 2004). Particular attention has been reserved
for the Viking Age, a period dominated by Scandinavian raiding and then settlement,
principally in the 9th and 10th centuries (Fell et al 1983; Kermode 1994; Wilson 2008).
Indeed, Man provides some of the most interesting and compelling archaeological,
sculptural and epigraphic evidence for a high degree of toleration between Gaelic
Christian natives and Norse pagan incomers. As David Freke succinctly remarked of
its early history, ‘The Isle of Man . . . seems to have benefited from a fruitful inter-
action of indigenous and alien cultures’ (Freke 1990, 113).
This paper develops the theme of Man and the Isles as a cultural crossroads
where indigenous and alien cultures fruitfully interacted, but does so from the per-
spective of what A W Moore called the ‘Second Scandinavian Period’ (Moore 1900, i,
102) and what is now locally known as the Late Norse period (Freke 2002, 132) —
that is, the era of the Kingdom of Man and the Isles, from roughly the late 11th to
the mid-13th century. Too late for students of the Viking Age, and too marginal for
mainstream histories of England, Scotland, Ireland and Scandinavia this is a period
that has long been neglected by scholars, but deserves to be better known (see
McDonald 2007a; Power 2005). Since this era coincides with a time of tremendous
economic growth, territorial expansion, and social and cultural change in Europe
generally, constituting, in the words of Robert Bartlett (1993, 3), ‘the foundation of
an expansive and increasingly homogeneous society — ‘‘the Making of Europe’’’, my
principal objective is to establish the attitudes of the kings of Man and the Isles
towards, and their engagement with, contemporary European trends. In short, the
Isle of Man, as a peripheral region of Europe, becomes a significant testing ground for
Bartlett’s notions of the ‘Europeanization of Europe’, since, as he put it, ‘the expan-
sionary power of this civilization sprang from its centres, even if it may be seen most
starkly at its edges’ (Bartlett 1993, 3).
We may begin with some basic geographic, economic and historical context.
The Isle of Man is a small island of 572 square kilometres lying, as noted above, in the
northern Irish Sea basin, roughly equidistant between Strangford Lough in Ulster to
the west and St Bees head in England to the east. Burrow Head in south-west
Scotland lies a mere 30km to the north, while Anglesey in Wales lies some 80km to
the south. In clear weather all of these surrounding lands are visible from the Isle of
Man, leading one modern historian to describe it as the ‘Midway Island’ of the Irish
Sea (Davies 1999, 7, 9). By virtue of its position in the northern Irish Sea basin at the
southern outlet of the North Channel, the narrow and often dangerous passage that
links the Irish Sea to the North Atlantic, Man was joined by the sea-road to the
Hebrides, the chain of more than five hundred topographically diverse islands arcing
through some 400km off the west coast of Scotland. Further north, round the stormy
seas off Cape Wrath (from Old Norse hvarf, turning point), lay the Orkney and
Shetland islands — themselves stepping-stones to Norway, Iceland and Greenland
(Crawford 1987, 11–27). Parts of the Hebrides, including the Isle of Skye and some of
the outer Hebridean isles fell under the control of the Manx sea kings; hence the title
MAM 37: 22-mcdonald – Press

sea kings, maritime kingdoms and the tides of change 335

‘king of Man and the Isles’ (McDonald 2007a, 161–167). Control of these islands by
the Manx kings meant that their discontiguous maritime kingdom straddled impor-
tant and well-established maritime routes that constituted what has been called the
‘Celtic sea route of the Vikings’ (Ridel 2007). Control of (or an attempt to control)
the sea routes is suggested by an episode of about ad 1200 mentioned in Hrafns saga
Sveinbjarnarsonar (see Chapter 21). When bishop-elect Guðmundr Arason and the
hero of the saga were blown off course in a ship en route from Iceland to Norway,
they arrived at the isle of Sanday in the Hebrides where a prince named Óláfr
attempted to levy a hefty landing fee upon them (Anderson 1922, ii, 358–60;
Vigfusson 1878, ii, 290–292). This Óláfr was quite likely Óláfr Guðrøðarson (d 1237),
brother of the then-ruling Rögnvaldr Guðrøðarson and eventually king in his own
right. The traffic flowing along the sea-road is further exemplified by the spectacular
find of 78 chess pieces (with 14 plain disks and one belt buckle) at Uig Bay on the
island of Lewis in ad 1831: the so-called Lewis Chessmen. Made of walrus ivory
(probably of Greenlandic origin) and crafted, in all likelihood, in Norway, the chess
pieces were prestigious items and may have been buried by a merchant travelling the
well-established route from Scandinavia to Ireland via the Hebrides (Caldwell et al
2009; Robinson 2004; Stratford 1997). Thus, ‘the strategic value of the island must
have been considerable’, never more so, perhaps, than in the period under consider-
ation here (Megaw and Megaw 1950, 148).
The Isle of Man offered more than simply a strategic location within the Irish
Sea and upon the sea routes to those who controlled it. It also represented a sig-
nificant economic prize (Davey 2002; Hudson 1999). Medieval writers ranging from
Icelandic saga authors to the Irish bard who sang King Rögnvaldr’s praises to the
author of the Manx chronicle all celebrated the fertility of the Island. Despite a high
proportion of uplands, Man contains good farmland well suited to the grazing of
cattle and sheep. Óláfr Guðrøðarson (r ad 1226–37), for example, granted the monks
of St Bees priory the right to purchase 60 head of cattle in his lands, or their equiv-
alent value in swine or sheep. Cereal crops were also grown (Cheney 1984a, 51–53;
1984b 76–78; Gelling 1960–61; 1964; Wilson 1915, no 45). Archaeological excava-
tions reveal that even upland areas were intensively exploited in the period of the Late
Norse kings, suggesting the Island was ‘fairly densely populated at the time, and that
all the good, and even medium-quality, farmland was occupied’ (Gelling 1970, 81).
Fishing was another important activity, attested in the charters of several of the
Late Norse Manx kings that granted fishing rights to religious houses, and in the
archaeological record. Excavations at Peel Castle in the 1980s showed that fish played
a major role in the medieval economy, with an expansion of fishing activity detected
in the 11th or 12th century (Freke 2002, 258–261; cf Barrett 2007). The Island was
also well endowed with significant mineral resources. King Haraldr Óláfsson (d 1248)
granted the monks of Furness abbey ‘the use of all kinds of mines which may be
found within my kingdom, both beneath the soil and above’, as well as a depot at
‘Bakenaldwath’ (probably Ronaldsway) (Loyd and Stenton 1950, 298–299; Megaw
1960–61; Oliver 1860–62, ii, 79–80). Mercantile activity is more difficult to trace in
the written sources, but a letter of King Óláfr of about ad 1228 refers specifically
to his merchants (Simpson and Galbraith 1986, 136), while the Lewis chessmen
MAM 37: 22-mcdonald – Press

336 maritime societies of the viking and medieval world

(mentioned above) point to long-distance mercantile activity along the sea-road. All
of these resources appear to have been effectively exploited by the Manx kings, who,
like other contemporary rulers, regulated commerce in their realm and benefited from
customs duties, tolls and taxes (eg Grainger and Collingwood 1929, nos 265, 266). It
is testimony to the wealth of the island and its efficient exploitation by its rulers that
despite its small size it was home to three major power centres in the Late Norse
period (Curphey nd): St Patrick’s Island on the west coast, guarding the best natural
harbour in the Island; Castle Rushen overlooking Castletown Bay in the south-east, a
late 12th- or early 13th-century tower keep in contemporary style which, by the
mid-13th century, had become the headquarters of the Manx kings; and Rushen
Abbey, a Savignac, later Cistercian, foundation from Furness Abbey established in
ad 1134 by King Óláfr, the premier monastic foundation on the Island and the burial
place of several of the later Manx kings. Against this evidence of prosperity must be
set the lack of urban centres in the Island until the 16th century, and there is a rather
puzzling lack of coinage on the Island between the late 11th and early 14th centuries.
Yet clearly the Isle of Man proved to be a considerable economic, as well as strategic,
asset to its medieval rulers.
For nearly two centuries between ad 1079 and ad 1265 — a period roughly com-
parable to the tenure of the Canmore kings in Scotland (ad 1058–1286), and one that
exceeds the famous Norman dynasty in England — Man was ruled by an often
formidable line of sea kings. The dynasty’s founder was Guðrøðr Crovan — a strong
candidate for the King Orry of Manx legend — a survivor of the battle of Stamford
Bridge (ad 1066) who in ad 1079 established himself in Man and also, a little later
(and briefly), in Dublin (Duffy 1992, 106–108). Guðrøðr and his descendants (conve-
niently known to modern scholars, if not to contemporaries, as the Crovan dynasty)
assembled and ruled a far-flung sea kingdom that embraced Dublin, Man and the
Hebrides, although control of the former and parts of the latter was intermittent and
hotly contested. The kings descended from Guðrøðr are for the most part fairly well
known in the history of the Irish Sea and the Hebrides, though they remain largely
ignored in broader British or Scandinavian contexts.1
The dynasty is particularly noteworthy for a series of long-reigning kings
through the 12th and into the 13th century: Óláfr Guðrøðarson (r ad c1113–53)
was followed by his son Guðrøðr (ad 1154–87) who was followed in turn by his sons
Rögnvaldr (Reginald/Raghnall; ad 1187–1226, d 1229) and Óláfr (ad 1226–37). Then,
from ad 1237 until ad 1265, Óláfr’s three sons ruled in turn: Haraldr (ad 1237–48),
Rögnvaldr (Reginald) (May ad 1249) and Magnús (ad 1252–65). The zenith of the
dynasty might have come under the long-reigning Rögnvaldr Guðrøðarson (d 1229).
Depicted in a Gaelic praise poem as an old-fashioned Viking warlord, and described
by the author of Orkneyinga saga as ‘the greatest fighting man in all the western
lands’ (Pálsson and Edwards 1978, 221), Rögnvaldr was also a Christian prince who
corresponded with the pope, patronized the church in his lands, and enjoyed political
and diplomatic dealings with John and Henry III of England, as well as John de
Courcy, the conqueror of Ulster (McDonald 2007a). Eventually, however, the Manx
kings succumbed to the overwhelming might of more powerful neighbours. Even
before the death of Magnús Óláfsson in ad 1265, Man had already passed under
MAM 37: 22-mcdonald – Press

sea kings, maritime kingdoms and the tides of change 337

Scottish suzerainty, but within 25 years its strategic location astride the western
seaways ensured that it became a bone of contention in the Anglo-Scottish wars until
it finally passed under English control in ad 1333 (McDonald 1997, 127–199;
McNamee 1997). The Manx kings endured periodic bouts of kin strife, challenges
from the Hebridean chieftain Sumarliði (hereafter Somerled) (d 1164) and his
descendants, the MacSorleys, as well as the waxing of English, Norwegian, and
Scottish naval power in the western seaways. But for the most part the history of the
Crovan dynasty is remarkable for its stability and for the manner in which the Manx
kings successfully amplified their power, prestige and status within the intensely
competitive arena of the Irish Sea basin, all the while modernizing their kingship
along contemporary lines (McDonald 2007a). Much of the success of these rulers
rested upon their military might, principally the galley fleets that frequently appear in
a variety of 11th-, 12th-, and 13th-century sources (McDonald 2007b; Rixson 1998).
The doings of the Manx kings descended from Guðrøðr Crovan are revealed to
us in the first instance through the medieval Latin text known from its opening rubric
as Cronica regum mannie et Insularum (BL Cotton Julius Avii) which provides
unique Manx information and narrates their activities from ad 1066 down to ad
1265, with a continuation into the early 14th century, as well as information on the
bishops of the Isles and the boundaries of Rushen Abbey in the Island (Broderick
1996; Cubbon 1924). Although it is riddled with chronological inaccuracies
(particularly in the early entries up to the mid-12th century) and other problems,
the Chronicle (or Manx Chronicle, as it is commonly known) was probably
composed at Rushen Abbey in the mid-13th century, giving it invaluable status as
virtually our only piece of indigenous contemporary historical writing from Man in
this period. Its most recent editor, George Broderick, has remarked that, ‘the main
chronicler has made every effort to be as accurate in detail as he possibly can, and has
taken seriously the preparation of his account’ (Broderick 1996, xiv). The Chronicle
sheds significant light upon the Crovan kings, and its account can be supplemented
with a patchy and diverse variety of chronicles, annals, letters, diplomatic documents,
sagas, poems and other documents of English, Scottish, Irish, Welsh, Norwegian,
Icelandic and Continental provenance, as well as a small handful of charters issued by
the Manx kings themselves. If it is safe to say that the Manx sea kings are much less
well documented than many of their neighbours and that many aspects of the Late
Norse Isle of Man remain imperfectly understood, it is no longer possible to use lack
of surviving primary sources as an excuse for their neglect; archaeology, meantime,
also continues to add to our knowledge of the medieval Isle of Man.
A brief examination of the connections of Man with the surrounding landmasses
highlights the position of the island as the ‘navel of the sea between Britain and
Ireland’. As might be expected given the close links that existed between Dublin
and Man in the 11th century, Irish connections are particularly significant in the
1st century of the Crovan dynasty (Duffy 1992; 2002; McDonald 2008). A Manx–
Norwegian axis also existed (Power 2005), however, becoming more prevalent as
Norwegian royal authority was consolidated by the early 13th century and as the
Dublin axis was severed by the English conquest of the late ad 1160s and its sub-
sequent consolidation. Contacts with England were no less important, originating
MAM 37: 22-mcdonald – Press

338 maritime societies of the viking and medieval world

perhaps with the political exile of Óláfr son of Guðrøðr Crovan at the court of
Henry I in the early 12th century and intensifying by the early 13th when King
Rögnvaldr adroitly utilized English support from John and Henry III in his struggles
with his brother, Óláfr. Scotland, too, was never far from the horizons of these rulers.
Rivalry between the Manx kings and the expansive kindred of Somerled in the
Hebrides may have made the former tenuous allies of the Scottish kings from time to
time, as circumstances dictated: following two defeats at the hands of Somerled in the
late ad 1150s, Guðrøðr Óláfsson (d 1187) appeared briefly in Scotland, likely en route
to political exile in Norway (Barrow 1960, no 131). Also significant were ties with
other political entities ringing the Irish Sea, including Galloway, the Hebrides,
Kintyre, Ulster and Wales and, even further afield, in Orkney and Ross. The kaleido-
scopic patterns of Manx orientations vis-à-vis these entities cannot be explored at
length here (see further McDonald 2007a; Power 2005); suffice to say that various
influences waxed and waned over the history of the Crovan dynasty, a process often
driven as much by external factors as by the dynamics of the Irish Sea. Thus, for
example, the weakening of the Dublin–Manx axis owes much to the English con-
quest of the late ad 1160s and early ad 1170s, although this very event also opened
new possibilities — the Anglo-Norman adventurer John de Courcy married the sister
of Rögnvaldr Guðrøðarson, thereby gaining access to the formidable fleet of galleys
commanded by the Manx ruler (Broderick 1996, f 41r). Bearing all these geopolitical
orientations in mind, what can be said about the receptivity of the Manx rulers
descended from Guðrøðr Crovan to the various influences to which they were subject
in the 12th and 13th centuries?
A good place to begin may be with an evaluation of the utility of the commonly
used designation ‘Norse’ or ‘Scandinavian’ for these Manx kings. Despite the aura of
mystery surrounding the origins and ancestry of Guðrøðr Crovan, his background is
not in question — probably he was a member of the Hiberno-Norse dynasty of
Dublin (Duffy 2002, 55–56; Hudson 2005, 171–172). The evidence of personal names
certainly supports a Scandinavian ancestry, and the preference of the dynasty for
Scandinavian names is clear. Indeed, the pool of royal names was relatively small:
Rögnvaldr, Guðrøðr, Haraldr, Óláfr, Lögmaðr and Magnús. Even younger sons or
sons with Gaelic mothers were given Scandinavian names (although a hitherto un-
noticed brother of Rögnvaldr was styled in Latin Rotherico, probably a rendering of
the Gaelic name Ruaidrı́ (Atkinson and Brownbill 1886–1919, ii, iii, 711)). This
practice contrasts with that of other contemporary Scandinavian or Gaelic-Norse
dynasties in the Irish Sea, Hebrides and North Atlantic. Thus, for example, Christian
names like Paul appear among the Orkney jarls (Crawford 1987, 80), while Scottish
names like Alexander became prevalent in the 13th century for eldest sons in some of
the Hebridean dynasties descended from Somerled of Argyll (McDonald 1999,
185–187). Considering the significance of names and patterns of naming as indicators
of cultural affiliations and orientations (Bartlett 1993, 270–278), the onomastic
evidence suggests that the dynasty of Guðrøðr Crovan was self-consciously Scandi-
navian, although Gaelic bynames were sometimes used, as, for example, in the case of
King Rögnvaldr’s son Guðrøðr ‘Don’ (donn, ‘brown’ or ‘brown-haired’) (Megaw
MAM 37: 22-mcdonald – Press

sea kings, maritime kingdoms and the tides of change 339

1978, 276–277). Interestingly, however, it is among female members of the dynasty

that there is some evidence for non-Scandinavian appellations: thus Guðrøðr
Óláfsson (d 1187) had a daughter named Affrica (Broderick 1996, f 41r), and the very
short-reigning Rögnvaldr Óláfsson (d 1249) seems to have had a daughter named
Mary (Chaloner 1864, appendix C; Oliver 1860–62, ii, 135–136).
Despite this seemingly self-conscious identification as Norse rulers, there is also
clear evidence of substantial Gaelic influence within the dynasty. Apart from the
political affiliation with Dublin in the 11th and 12th centuries, this shows through
most clearly in matrimonial politics. Of the Manx kings whose spouses are known,
the majority married into Gaelic or Gaelic-Norse families in the Irish Sea basin or the
Hebrides. To take but one telling example from many, Óláfr Guðrøðarson (d 1237)
married, first, a daughter of a noble of Kintyre, and, second, the daughter of Earl
Ferchar of Ross in northern Scotland (Broderick 1996, f 42v; McDonald 2003, 39–40).
Marriage into other Scandinavian dynasties seems, in contrast, more rare. Haraldr
Óláfsson married a daughter of the Norwegian king in ad 1248 but, sadly, both
Haraldr and his bride were drowned in a shipwreck off Shetland en route to Man
from Norway later that year (Broderick 1996, ff 46v–47r). The Norwegian text
known as Ágrip af Nóregskonunga so˛gum (Summary of the Norwegian kings’
history), composed in the late 12th century, includes genealogical information stating
that Ingibjörg, a daughter of Jarl Hákon Pálsson of Orkney, ‘was married to Óláfr,
the king of the Hebrides’, though the Manx chronicle says that he married a daughter
of Fergus of Galloway (Broderick 1996, f 35v; Einarsson 1985, 373).
Two further Gaelic connections are interesting and noteworthy. First, the
founder of the dynasty, Guðrøðr Crovan, has gone down in history with the sobri-
quet Crovan. The derivation of this is either ON kruppin (cripple) or else Irish crobh-
bhan ‘of the white hand’ or crúbach (claw). Irish sources designate him meranach,
probably derived from finger, and it has been suggested that he suffered from some
deformity of the hands (Anderson 1922, ii, 43–44; Hudson 2005, 173; Megaw 1978,
276). Another particularly interesting piece of evidence for Gaelic influence in the
dynasty is to be found in a 12th-century praise poem composed in Irish sometime
during the reign of Rögnvaldr Guðrøðarson (d 1229) (Ó Cuı́v 1956–57; Megaw 1978,
278). The fact that a Norse ruler of Man patronized Gaelic poets and the significance
of this for the broader question of the courtly culture of these rulers remains largely
unexplored in modern writing on the subject (but see Megaw 1978, 278–279; Power
2005, 6).
On the one hand, then, it may be best to describe these kings, as George
Broderick has done, as ‘Norse-Gaelic’ (Broderick 1996, vii) — a designation which
seems entirely appropriate in light of the matrimonial politics and dynastic connec-
tions of the family, and which highlights the interactions between the two cultures
under the Manx sea kings (but cf Gelling 1978). Yet despite the obvious strength of
the Norse and Gaelic influences on the Manx rulers, they certainly did not remain
aloof from wider contemporary cultural developments of the 11th to 13th centuries.
In what space remains, I propose to explore only two of the ways in which the rulers
of the Crovan dynasty engaged with contemporary European culture. I will examine
MAM 37: 22-mcdonald – Press

340 maritime societies of the viking and medieval world

first the evidence for the adoption of knighthood by the Manx kings, and then will
explore their receptivity to what Giles Constable (1996) described as ‘the Reforma-
tion of the Twelfth Century’, that is, the church and monastic reform that character-
ized the period ad 1100–1300.
We may begin with knighthood, since in many respects the knight was the glit-
tering symbol of the international or trans-national society of the central Middle
Ages. By the 12th century knighthood, which had been imported into Britain in the
wake of the Norman Conquest of ad 1066, was becoming known outside of England
in Scotland, Wales, and even the Hebrides (Crouch 1992, 120–163; Gillingham 1981).
The Scottish kings David I (d 1153), Malcolm IV (d 1165), William I (d 1214) and the
two Alexanders (d 1249 and 1286) were all knighted, and Malcolm and William in
particular proved themselves to be ardent devotees of the cult of knighthood (Neville
and McDonald 2007). In Wales at about the same time, as David Crouch has
remarked, ‘there is overwhelming evidence of the Welsh not just adopting Anglo-
French military technology and dress but having been thoroughly exposed to the
knightly ethos at an early date’ (Crouch 1992, 158). In the Hebrides, the descendants
of Somerled, the king of the Isles and Kintyre, were depicting and describing them-
selves as knights from the first half of the 13th century (McDonald 1999, 191–195).
Since the era of the Crovan dynasty coincides with the great age of knighthood,
and since the matrimonial, political and diplomatic connections of these sea kings
criss-crossed the Irish Sea, it is worthwhile asking whether and to what extent the
Manx kings themselves were influenced by contemporary trends in knighthood and
chivalry.
The 13th-century Manx chronicle provides some illumination of this problem. It
relates that in ad 1247 King Haraldr ‘was made a knight by Lord Henry king of
England, as was his father, and with much honour and precious gifts from the king he
was allowed to return home’ (Broderick 1996, f 46r). Similarly, nine years later in
ad 1256 ‘Magnus king of Man and the Isles went to the court of the Lord king of
England. The . . . king of England received him with grace and honour, and made him
a knight’ (Broderick 1996, ff 49r–v). The Chronicle is at this juncture both contem-
porary and well informed and there is little reason to doubt it, but these statements
may also be corroborated from English documents. A letter of safe conduct issued for
Magnús in ad 1256, for example, describes him as having been decorated with the
belt of knighthood (Rymer 1739–45, i, ii, 12). There is no similar description of
Haraldr in ad 1247, but safe conducts exist for both ad 1246 and ad 1249, which
corroborate at the very least the contacts with England (Rymer 1739–45, I, i, 264),
and the well-informed and contemporary English chronicler Matthew Paris records
the knighting of both Haraldr and Magnús (Luard 1872–83, iv: 551, v: 549).
Thus, three of the 13th-century Manx rulers appear to have been knighted. But
earlier Manx rulers might also have been knighted, or at the very least, have been
acquainted with the knightly ethos. The Manx chronicle relates, for example, that
before Óláfr son of Guðrøðr Crovan came to power after a lengthy period of civil
war, ‘he was at that time living at the court of Henry, King of England. . .’ (Broderick
1996, f 35r). As a political exile at the court of the Norman king of England, Óláfr
MAM 37: 22-mcdonald – Press

sea kings, maritime kingdoms and the tides of change 341

calls to mind the parallel example of David I of Scotland, who was, of course, raised
at the same court – and incidentally knighted there as well (Anderson 1908, 155–156;
Barrow 1985). Whether Óláfr was indoctrinated into knighthood while living in
England must remain an open question, but it seems likely, at least. Also interesting is
the fact that Óláfr’s son, Guðrøðr, received arms and equipment from the English
king in the late ad 1150s, whilst in exile there (Hunter 1844, 155, 168). Moreover,
King Rögnvaldr (d 1229) maintained close diplomatic and political relations with the
English kings John and Henry III, as well as John de Courcy, in the first quarter of
the 13th century. In addition to performing homage to John, Rögnvaldr received a
grant of Carlingford in Ulster in ad 1213 along with a hundred measures of corn, to
be held for the service of one knight (Hardy 1837, 186; McDonald 2007a, 137–141).
While none of this proves that Rögnvaldr or his kinsmen were actually knighted,
these examples suggest at the very least a fundamental acquaintance with some of the
contemporary conventions of knighthood and chivalry.
The adoption of knighthood by what might be called ‘native’ kindreds in the
British Isles is particularly important in cultural terms because, as Rees Davies
eloquently reminds us, it ‘opened the door onto an exhilarating international world
of aristocratic fellowship and customs’ – thereby promoting accommodation and
acculturation (Davies 1990, 51). As powerful sea kings in a Viking mould in the Irish
Sea basin who were also knights, the members of the dynasty of Guðrøðr Crovan
straddled two worlds, old and new, Gaelic-Norse and contemporary European or
‘Frankish’. This aspect is neatly encapsulated on a seal utilized by King Haraldr in the
ad 1240s, which displayed an armorial bearing a lion on one side and a galley with
sails furled on the other, while King Rögnvaldr is said to have utilized a seal with an
equestrian motif on one side and a galley on the other (Loyd and Stenton 1950,
298–299, 302; Oliver 1860–62, i, 118). As I have argued elsewhere (McDonald 2007a,
216), however, it was not just their engagement with contemporary European culture
in the form of knighthood that mattered. There was also the fact that as non-English
or Scottish monarchs in the British Isles who were also knights, the Manx kings
joined an exclusive club into which very few native Welsh or Irish rulers were
admitted (Davies 1990, 51). Knighthood among the Manx sea kings therefore needs
to be viewed within the broader context of an increasingly sophisticated brand of
kingship in Man and the Isles.
In religious terms, we could perhaps be forgiven if for a moment we lost sight of
the fact that with the dynasty of Guðrøðr Crovan we are dealing with convention-
ally pious kings in a contemporary Christian mould. After all, Guðrøðr himself
behaved in seemingly typical Viking style when, following his conquest of Man in
ad 1079, he offered his followers the choice of sharing out the land on the island or of
plundering it:
It pleased them more to lay waste the whole island and to enrich themselves with
its valuables, and thus to return to their homes. Godred [Guðrøðr] on the other
hand granted the southern part of the island to the few who had stayed with him,
and the northern part to the remainder of the Manxmen . . .
(Broderick 1996, ff 32v–33r)
MAM 37: 22-mcdonald – Press

342 maritime societies of the viking and medieval world

Similarly, the Gaelic praise poem on Guðrøðr’s great-grandson Rögnvaldr deliber-

ately portrays this ruler engaged in the heroic Viking pastime of plundering:
You will deal death from smooth-plained Man, slaughter by a great ship’s host . . .
You are the man of brindled barques, the shore you will reach is cursed . . . You’ll
seek harbour behind Aran while probing Ireland’s cold shores. (Clancy 1998, 238,
239; cf Ó Cuı́v 1956–57)
Yet by the time the Gaelic poet composed these verses, the Manx kings had, to
borrow a phrase from the title of Benjamin T Hudson’s seminal book, completed the
transition from Viking pirates to Christian princes (Hudson 2005); indeed, they had
never really been ‘Vikings’ at all, if by that is meant pagans. One of Guðrøðr Crovan’s
sons, Lögmaðr, for example, embarked on pilgrimage to Jerusalem in the ad 1090s,
the timing hinting at participation in the first Crusade (Broderick 1996, f 33v). Another
son, Óláfr, who came to power in the early 12th century, was described as ‘devout and
enthusiastic in matters of religion and was welcome both to God and men’ (though
his addiction to ‘the domestic vice of kings’ (concubines) has been noted above)
(Broderick 1996, f 35v). Moreover, the same Rögnvaldr whom the Gaelic bard
depicted as a fierce pillager of Ireland corresponded with Popes Innocent III and
Honorius IV and became a vassal of the papacy in ad 1219 (Theiner 1864, 11). His
supplication was greeted a few years later with glowing praise from the pope, who
described the ‘ready faith and earnest devotion which you [Rögnvaldr] desire in a
laudable manner to exhibit towards the Roman church’ (Oliver 1860–62, ii, 64–65;
Theiner 1864, 21–22). Even if Rögnvaldr’s devout kissing of the papal feet had as
much to do with the uncertain political circumstances in which he found himself at
the time as with genuine piety, these Latin letters to and from the papacy reveal a
different milieu from that of the Gaelic praise poem.
Papal connections with Man and its rulers were nothing new by Rögnvaldr’s
time. In ad 1176 Vivian, Cardinal Legate of the Apostolic See, took time out from
other business in Britain and Ireland to visit Man, where (quoting the Chronicle) he
‘caused king Godred [Guðrøðr] to be lawfully betrothed to his wife called Fionnula
. . .’ (Broderick 1996, ff 39v–40r). At one stroke we see here the more centralized
papacy of the 12th century asserting itself in Man, and ironing out the rather dubious
(from the point of view of the Papal Curia, anyway) marriage customs that seemed to
persist in the Gaelic-Norse parts of the British Isles. Exactly what it was the Cardinal
Legate felt necessary to sort out is illuminated by some earlier remarks in the
Chronicle, when the matrimonial politics of Guðrøðr’s father, Óláfr, were briefly dis-
cussed. Óláfr, we are told, ‘took a wife called Affrica, a daughter of Fergus of
Galloway, by whom he had Godred [Guðrøðr]; and he had many concubines from
whom he begat three sons’ (Broderick 1996, f 35v). Concubinage in fact continued to
be practised into the 13th century, since as late as about ad 1220–23 Guðrøðr’s son
Óláfr is said to have kept a concubine (Broderick 1996, ff 42r–v; see McDonald
2007a, 70–77).
It was, however, the first Óláfr, son of Guðrøðr Crovan, who wrought further
religious transformation in his kingdom, earning a reputation as perhaps the most
important figure in the medieval ecclesiastical history of Man. In ad 1134 King Óláfr
MAM 37: 22-mcdonald – Press

sea kings, maritime kingdoms and the tides of change 343

simultaneously established a territorial diocese for Man and the Isles, granting the
right of election of the bishop to the monks of Furness in Cumbria, and also founded
a monastery at Rushen, colonized with monks from the Savignac establishment at
Furness (Atkinson and Brownbill 1886–1919, ii, iii, 708–709; Oliver 1860–62, ii, 1–3).
Rushen abbey (which later became Cistercian when the Savignacs merged with that
order in the ad 1140s) became the largest, wealthiest, and most significant monastic
establishment in the Island until it was dissolved by Henry VIII in June ad 1540, and
it became a dynastic burial place in the second quarter of the 13th century, serving as
the final resting place for three of the four Manx kings who ruled between ad 1226
and ad 1265 (Broderick 1996, ff 44v, 47r, 49v). Forty years after the foundation of
Rushen, Óláfr’s son, Guðrøðr, established a second monastery in the Isle of Man at
Myroscough in the north in ad 1176. This was a daughter of the Cistercian house
of Rievaulx in north Yorkshire, but it was modestly endowed and was eventually
assimilated by Rushen (Broderick 1996, f 40r; Cubbon 1927). Later and more suspect
evidence credits King Rögnvaldr with the establishment of a house of Cistercian nuns
at Douglas (McDonald 2007a, 200); if accurate, this is further evidence of engagement
with contemporary monastic tendencies.
The religious patronage of the Manx kings was not confined to the shores of the
Isle of Man. A number of monasteries ringing the Irish Sea basin also benefited from
their largesse. In addition to founding Rushen and Myroscough in the Isle of Man
itself, the Manx kings also acted as patrons to no fewer than five monastic houses
in England (Holm Cultram, St Bees Priory, Furness, Conishead and St Werburgh in
Chester), one in Scotland (Whithorn Priory), and several in Ireland (St Mary’s,
Dublin, Bangor and Sabal), representing Benedictine, Savignac, Cistercian, Augustin-
ian and Premonstratensian orders (McDonald 2007a, 192–201). It is, moreover,
entirely possible that the paucity of surviving source materials has deprived us of
knowledge of further religious patronage on the part of the Manx kings. The loss of
materials relating to Iona abbey in the Hebrides and to the monasteries of Galloway
has almost certainly obscured Manx patronage of these houses as well: King
Guðrøðr, for example, was buried on Iona in ad 1187 (Broderick 1996, f 40r), sug-
gesting a connection with this ancient and important monastery. Whatever the case
may be, even such a cursory overview as this demonstrates that the Manx kings, by
virtue of their wide-ranging religious patronage both onshore and offshore the Isle of
Man, played an important role in developing the reformed Church in the Irish Sea
basin (Stringer 2003, 2). W C Cubbon perceived in all of this what he called the
‘gradual tightening of the grasp of the Church upon the kings of Mann’ (1932–42,
105), but of course that is only half the story. The Manx rulers themselves were also
actively embracing the precepts of not merely Christianity but of the particular brand
of reformed Christianity that permeated the 11th to 13th centuries. As rulers on the
margins of Europe who reorganized the church in their territories, the Manx kings
might be compared to their contemporaries and counterparts in Wales, Ireland,
Scotland, Galloway and the Hebrides (Barrow 2003; McDonald 1995b; Stringer
2000). What is especially remarkable about this process in Man, however, is the
apparent enthusiasm of the Manx princes for the process and the fact that it occurred
as early in Man as it did in Scotland under David I. Indeed, the foundation of Rushen
MAM 37: 22-mcdonald – Press

344 maritime societies of the viking and medieval world

in ad 1134 actually predates the foundation of the first Cistercian house in Scotland,
Melrose, in ad 1136 (Anderson 1922, ii, 195).
Religious patronage was significant for more than just its spiritual benefits. The
religious played an important role in enhancing royal authority through their
employment as an educated staff capable of producing the written instruments that
brought the non-literate and undocumented regions of Europe into what has been
prosaically described as the ‘world of the chancery’ (Bartlett 1993, 285) in the 11th to
13th centuries. Indeed, the reformed church provided the context for the introduction
of charters into many regions of Europe, and the spread of reformed monasticism in
the central Middle Ages has been regarded as a major factor in the expanding number
of documents that survive for this period (Clanchy 1993, 60). Although the surviving
number of royal charters from the Manx sea kings is slight at about 20 from the
period of the Crovan dynasty, the number is sufficient to demonstrate that the Manx
kings were quick to participate in the so-called ‘charter revolution’ of the period and
to engage with contemporary models of literacy, documentary thinking and bureau-
cratic organization that marked an important aspect of the culture of the period
(Bartlett 1993, 285). It is significant that among the witnesses to these Manx charters
it is possible to discern a substratum of clerics who may have played a role in the pro-
duction and dissemination of written instruments within the kingdom of Man and the
Isles. Three individuals are identified in the witness lists as chaplains, four as clerks,
and six are designated as masters. This suggests that the Manx kings had access to
men with the talent and education for scribal work, and it is tempting to speculate that
there may have existed, by the second quarter of the 13th century, what has been
described as a ‘rudimentary writing office’ or an ‘embryonic Manx chancery’
(McDonald 2007a, 203). It may be overly ambitious and inaccurate to talk in terms
of the development of ‘bureaucratic kingship’ in a Manx context in this period, but
then again it may be that the paucity of the surviving source materials once more
obfuscates evidence pointing to sophisticated trends in contemporary kingship in
Man and the Isles. Certainly the use of charters was something that the Hebridean
cousins of the Manx sea kings were embracing at precisely the same time (McDonald
1999, 95–97).
Finally, an important concomitant to the Latin charter was the use of waxen
seals for purposes of authentication, and these seals also possessed considerable sym-
bolic value in denoting status and power (Clanchy 1993, 51, 317). No seal from a
Manx king of the Crovan dynasty survives (so far as is known) today, but there is
considerable evidence to show that they were known and utilized from the second
half of the 12th century (McDonald 2007a, 204–206; Megaw 1959–60; see also
McDonald 1995a). References to, as well as descriptions and illustrations of Manx
seals are known for kings Guðrøðr (d 1187), Rögnvaldr (d 1229) and Haraldr
(d 1248): two charters of Haraldr that once formed part of the famous Cottonian
library were unfortunately destroyed in the fire of ad 1731 that devastated the col-
lection, but not before they had been depicted by the 17th-century antiquarian Sir
Christopher Hatton (d 1670) in his Book of Seals. The double-sided seals show a
galley with sails furled on one side and what appears to be a lion on the other (Loyd
and Stenton 1950, 298–299; Oswald 1860, frontispiece), while an early 18th-century
MAM 37: 22-mcdonald – Press

sea kings, maritime kingdoms and the tides of change 345

description of Rögnvaldr’s seal stated it to have exhibited an equestrian figure on one

side and a ship under sail on the other (Oliver 1860–62, i, 118).
As the lords of a small island ‘not far distant by sea’ from neighbouring lands (as
King Óláfr put it in a letter to the Archbishop of York in ad 1134) (Atkinson and
Brownbill 1886–1919, ii, iii, 709; Oliver 1860–62, ii, 4–5), it is ultimately unsurprising
that the Manx sea kings of the 11th to 13th centuries were receptive to foreign influ-
ences. The seaways formed a cultural arena rather than a political barrier (Carver
1995). It would, however, prove dangerous to emphasize innovation at the expense
of tradition, and it is important to recognize that, despite the reception of foreign
influences, Manx and Hebridean kingship of the period in fact represented what has
been called in another context ‘an intriguing mix of old and new’ (Frame 1990, 115)
or, elsewhere, a ‘hybrid’ society and kingship (Grant 1988, 119; see also Caldwell
2004). There appear to have been few of the tensions between perceived ‘civility’ and
‘barbarism’ that characterized other parts of Britain in this period (Davies 2000).
Thus, for example, while the role of the Manx kings as enthusiastic patrons of the
reformed religious orders and the knightly ethos can be highlighted, this does not
mean they embraced these European trends to the exclusion of all else. In fact, the
relationship to and the impact of the new monastic foundations on the existing
monastic communities in the Isle of Man, for example, has hardly been explored and
remains obfuscated by paucity of documentation. But the existence of earlier mon-
astic communities at sites like Maughold in the north of the Isle of Man is well
attested, and this site in particular seems to have retained its importance into the
11th and 12th centuries, at least (Broderick 1996, ff 38r–39r, 50v; Megaw 1946–50,
169–181). Moreover, the relationship of an early Christian cemetery to the south of
the main monastic complex at Rushen to the Savignac/Cistercian abbey there remains
unclear (Davey 1999, 38–43). Thus, ‘[h]ow the favouritism (if favouritism it was)
displayed by the kings of the Crovan dynasty to the continental religious orders
impacted such preexisting monastic communities . . . remains an important problem
in the ecclesiastical history of the Isle of Man’ (McDonald 2007a, 194). Similarly, the
degree to which European knighthood and chivalry was taken up by the medieval
Manx ruling elite is also an open question. Certainly, despite their appearance in
Manx and Hebridean contexts and upon the seals of the Manx and Hebridean rulers
of the period, knights can never have formed a significant component of the formid-
able armies that the Manx kings commanded; these multi-ethnic armies, which were
raised from throughout the lordship of the sea kings, included Gaelic troops from
the Hebrides and Ireland and may have included the predecessors of the famous
galloglass (Duffy 2007; McDonald 2007b; see also Chapter 23). Examination of the
receptivity of medieval Manx and Hebridean rulers to contemporary European influ-
ences therefore reinforces the notion of the islands between Britain and Ireland not as
cultural backwaters, but as occupying a significant location at Peter Davey’s ‘cross-
roads of power and cultural influence’ (Davey 2002). The study also demonstrates
how old and new could fruitfully interact in a suitable, and in this case, maritime,
environment. Indeed, it is an intriguing thought that the sea-girt nature of these
maritime kingdoms may have been a significant factor in promoting accommodation,
the insular environment providing just the right balance between easy proximity to
MAM 37: 22-mcdonald – Press

346 maritime societies of the viking and medieval world

foreign influences and yet also offering sufficient remoteness to temper the potentially
overwhelming impact of those influences (see Frame 1990, 105). The tides of Euro-
pean change lapped on the shores of these islands far distant from the traditional
‘centres’ of European society, but they were not inundated.

note
1 The orthography of personal names presents a Norse names in Old Norse-Icelandic form, except
significant challenge in a Manx and Hebridean for Sumarliði, which is so commonly rendered as
context; both Norse and Gaelic were spoken in Somerled that I retain this form. On the chal-
the Isles in the Middle Ages and a variety of forms lenges of finding a suitable orthography for per-
can be, and have been, utilized by scholars. I give sonal names see McDonald 2007a, 13.

references
Anderson, A O (ed), 1908 Scottish Annals from English Chroniclers ad 500 to 1286, David Nutt,
London
Anderson, A O (ed), 1922 Early Sources of Scottish History ad 500–1286, 2 vols, Oliver and Boyd,
Edinburgh
Atkinson, J C and Brownbill, J (ed), 1886–1919, The Coucher Book of Furness Abbey, 2 vols in
6 parts, Chetham Society, Manchester
Barrett, J H, 2007 ‘The pirate fishermen: the political economy of a medieval maritime society’, in
B Ballin Smith, S Taylor and G Williams (ed), West Over Sea: Studies in Scandinavian Sea-
borne Expansion and Settlement Before 1300, Brill, Leiden, 299–340
Barrow, G W S (ed), 1960 The Acts of Malcolm IV King of Scots, 1153–1165, Edinburgh University
Press, Edinburgh
Barrow, G W S, 1985 David I of Scotland: The Balance of New and Old, Stenton Lecture 1984,
University of Reading, Reading
Barrow, G W S, 2003 ‘The royal house and the religious orders’, in G W S Barrow (ed), The
Kingdom of the Scots, Government, Church and Society from the Eleventh to the Fourteenth
Century, Edinburgh University Press, Edinburgh, 151–169
Bartlett, R, 1993 The Making of Europe: Conquest, Colonization and Cultural Change, 950–1350,
Princeton University Press, Princeton
Broderick, G (ed), 1996 Cronica regum Mannie & Insularum: Chronicles of the Kings of Man and
the Isles BL Cotton Julius Avii, Manx National Heritage, Douglas
Caldwell, D, 2004 ‘The Scandinavian heritage of the Lordship of the Isles’, in J Adams and
K Holman, Scandinavia and Europe 800–1350: Contact, Conflict, and Coexistence, Brepols,
Turnhout, 69–86
Caldwell, D H, Hall, M A and Wilkinson, C M, 2009 ‘The Lewis Hoard of gaming pieces: a re-
examination of their context, meanings, discovery and manufacture’, Medieval Archaeology
53, 155–203
Carver, M, 1995 ‘The North Sea highway: cultural arena or political barrier’, in J M Fladmark (ed),
Sharing the Earth: Local Identity in Global Culture, Donhead Publishing, London, 177–190
Chaloner, J, 1864 A Short Treatise of the Isle of Man Digested into Six Chapters, Manx Society,
Douglas
Cheney, C R, 1984a ‘Manx synodal statutes, ad 1230(?)–1351, Part II: translation of Latin texts’,
Cambridge Medieval Celtic Studies 7, 51–63
Cheney, C R, 1984b ‘Manx synodal statutes, ad 1230(?)–1351, Part I: introduction and Latin texts’,
Cambridge Medieval Celtic Studies 7, 63–89
Clanchy, M T, 1993 From Memory to Written Record: England 1066–1307, Blackwell, Oxford
MAM 37: 22-mcdonald – Press

sea kings, maritime kingdoms and the tides of change 347

Clancy, T O (ed), 1998 The Triumph Tree: Scotland’s Earliest Poetry ad 550–1350, Canongate,
Edinburgh
Constable, G, 1996 The Reformation of the Twelfth Century, Cambridge University Press,
Cambridge
Crawford, B E, 1987 Scandinavian Scotland, Leicester University Press, Leicester
Crouch, D, 1992 The Image of Aristocracy in Britain, 1000–1300, Routledge, London and New
York
Cubbon, A M, 1974 Early Maps of the Isle of Man, Manx Museum and National Trust, Douglas
Cubbon, W C (ed), 1924 The Chronicle of Man and the Isles: A Facsimile of the Manuscript Codex
Julius A.VII in the British Museum, Manx Museum, Douglas
Cubbon, W C, 1927 ‘The lands of the monks of Myrosco’, Proceedings of the Isle of Man Natural
History and Antiquarian Society 3(2), 147–51
Cubbon, W C, 1932–42 ‘History of Rushen abbey, Ballasalla, Isle of Man — 1134 ad’, Proceedings
of the Isle of Man Natural History and Antiquarian Society 4, 98–115
Curphey, R A, nd Ancient Centres of Government of the Isle of Man, Government Property
Trustees, Braddan
Davey, P J (ed), 1999 Rushen Abbey, Ballasalla, Isle of Man: First Archaeological Report, Centre
for Manx Studies, Douglas
Davey, P J, 2002 ‘At the crossroads of power and cultural influence: Manx archaeology in the high
Middle Ages’, in P J Davey and D Finlayson (ed), Mannin Revisited: Twelve Essays on Manx
Culture and Environment, Scottish Society for Northern Studies, Edinburgh, 81–105
Davey, P J, 2004 ‘The Isle of Man: central or marginal in the Neolithic of the northern Irish Sea?’,
in V Cummings and C Fowler (ed), The Neolithic of the Irish Sea: Materiality and Traditions
of Practice, Oxbow Books, Oxford, 129–144
Davies, N, 1999 The Isles: A History, Oxford University Press, Oxford
Davies, R R, 1990 Domination and Conquest: The Experience of Ireland, Scotland and Wales,
1100–1300, Cambridge University Press, Cambridge
Davies, R R, 2000 The First English Empire: Power and Identities in the British Isles, 1093–1343,
Oxford University Press, Oxford
Duffy, S, 1992 ‘Irishmen and Islesmen in the kingdoms of Dublin and Man, 1052–1171’, Ériu 43,
93–133
Duffy, S, 2002 ‘Emerging from the mist: Ireland and Man in the eleventh century’, in P J Davey and
D Finlayson (ed), Mannin Revisited: Twelve Essays on Manx Culture and Environment,
Scottish Society for Northern Studies, Edinburgh, 53–62
Duffy, S (ed), 2007 The World of the Galloglass, Four Courts Press, Dublin
Einarsson, B (ed), 1985 Ágrip af Nóregskonunga s?gum, Hið Íslenzka Fornritafélag, Reykjavı́k
Fell, C, Foote, P, Graham-Campbell, J and Thomson, R (ed), 1983 The Viking Age in the Isle of
Man: Select Papers from the Ninth Viking Congress, Isle of Man, 4–14 July 1981, Viking
Society for Northern Research, London
Frame, R, 1990 The Political Development of the British Isles 1100–1400, Oxford University Press,
Oxford
Freke, D, 1990 ‘History’, in V Robinson and D McCarroll (ed), The Isle of Man: Celebrating a
Sense of Place, Liverpool University Press, Liverpool, 103–122
Freke, D (ed), 2002 Excavations on St Patrick’s Isle, Peel, Isle of Man, 1982–1988: Prehistoric,
Viking, Medieval and Later, Liverpool University Press, Liverpool
Gelling, P S, 1960–61 ‘Shielings in the Isle of Man’, Journal of the Manx Museum 6(77), 123–125
Gelling, P S, 1964 ‘Medieval shielings in the Isle of Man’, Medieval Archaeology 6–7, 156–172
Gelling, P S, 1970 ‘A Norse homestead near Doarlish Cashen, Kirk Patrick, Isle of Man’, Medieval
Archaeology 14, 74–82
MAM 37: 22-mcdonald – Press

348 maritime societies of the viking and medieval world

Gelling, M, 1978 ‘Norse and Gaelic in medieval Man: the place-name evidence’, in P J Davey (ed),
Man and Environment in the Isle of Man, Archaeopress, Oxford, 251–264
Gillingham, J, 1981 ‘The introduction of knight service into England’, Anglo-Norman Studies 4,
53–64, 181–187
Grainger, F and Collingwood, W G (ed), 1929 The Register and Records of Holm Cultram,
Cumberland and Westmorland Antiquarian and Archaeological Society, Kendal
Grant, A, 1988 ‘Scotland’s ‘‘Celtic fringe’’ in the late Middle Ages: the MacDonald Lords of the
Isles and the kingdom of Scotland’, in R R Davies (ed), The British Isles 1100–1500: Com-
parisons, Contrasts and Connections, John Donald, Edinburgh, 118–141
Hardy, T D (ed), 1837 Rotuli chartarum in turri Londinensi asservati vol I pars I ab anno
MCXCIX ad annum MCCXVI, Record Commission, London
Hudson, B T, 1999 ‘The changing economy of the Irish Sea province’, in B Smith (ed), Britain and
Ireland 900–1300: Insular Responses to Medieval European Change, Cambridge University
Press, Cambridge, 39–66
Hudson, B T, 2005 Viking Pirates and Christian Princes: Dynasty, Religion, and Empire in the
North Atlantic, Oxford University Press, Oxford
Hunter, J (ed), 1844 The Great Rolls of the Pipe for the Second, Third, and Fourth Years of the
Reign of King Henry II, ad 1155, 1156, 1157, 1158, Eyre & Spottiswoode, London
Kermode, P M C, 1994 Manx Crosses, Pinkfoot Press, Balgavies (Angus)
Leland, J, 1770 Joannis Lelandi antiquarii de rebus Britannicis collectanea, 6 vols, editio altera,
Richardson, London
Loyd, L and Stenton, D M (ed), 1950 Sir Christopher Hatton’s Book of Seals, Oxford University
Press, Oxford
Luard, H R (ed), 1872–83 Matthew Paris, Chronica majora, 7 vols, Longman & Co, London
McDonald, R A, 1995a ‘Images of Hebridean lordship in the twelfth and early thirteenth centuries:
the seal of Raonall MacSorley’, Scottish Historical Review 74, 129–143
McDonald, R A, 1995b ‘Scoto-Norse Kings and the reformed religious orders: patterns of monastic
patronage in twelfth-century Galloway and Argyll’, Albion 27(2), 187–220
McDonald, R A, 1997 The Kingdom of the Isles: Scotland’s Western Seaboard, c1100–c1336,
Tuckwell Press, East Linton
McDonald, R A, 1999 ‘Coming in from the margins: the descendants of Somerled and cultural
accommodation in the Hebrides, 1164–1317’, in B Smith (ed), Britain and Ireland, 900–1300:
Insular Responses to Medieval European Change, Cambridge University Press, Cambridge,
179–198
McDonald, R A, 2003 ‘Ferchar Maccintsacairt and the early earls of Ross, c 1200–1274’, in
S Boardman and A Ross (ed), The Exercise of Power in Medieval Scotland c 1200–1500, Four
Courts Press, Dublin, 23–45
McDonald, R A, 2007a Manx Kingship in its Irish Sea Setting, 1187–1229: King R?gnvaldr and the
Crovan Dynasty, Four Courts Press, Dublin
McDonald, R A, 2007b ‘Dealing death from Man: Manx sea power in and around the Irish Sea,
1079–1265’, in S Duffy (ed), The World of the Galloglass, Four Courts Press, Dublin, 45–76
McDonald, R A, 2008 ‘Man, Ireland, and England: the English conquest of Ireland and Hiberno-
Manx relations’, in S Duffy (ed), Medieval Dublin, Volume VIII, Four Courts Press, Dublin,
131–149
McNamee, C, 1997 The Wars of the Bruces: Scotland, England and Ireland, 1306–1328, Tuckwell
Press, East Linton
Megaw, B R S, 1946–50 ‘The monastery of St Maughold’, Proceedings of the Isle of Man Natural
History and Antiquarian Society 5(2), 169–181
MAM 37: 22-mcdonald – Press

sea kings, maritime kingdoms and the tides of change 349

Megaw, B R S, 1959–60 ‘The ship seals of the kings of Man’, Journal of the Manx Museum 6(76),
78–80
Megaw, B R S, 1960–61 ‘‘‘Bakenaldwath’’ and the medieval lead mines’, Journal of the Manx
Museum 6(77), 105–107
Megaw, B R S, 1978 ‘Norseman and native in the kingdom of the Isles: a re-assessment of the
Manx evidence’, in P J Davey (ed), Man and Environment in the Isle of Man, Archaeopress,
Oxford, 265–314
Megaw, B R S and Megaw, E, 1950 ‘The Norse heritage in the Isle of Man’, in Sir C Fox and
B Dickins (ed), The Early Cultures of North-west Europe, Cambridge University Press,
Cambridge, 141–170
Moore, A W, 1900 History of the Isle of Man, 2 vols, T Fisher Unwin, London
Neville, C J and McDonald, R A, 2007 ‘Knights, knighthood and chivalric culture in Gaelic
Scotland, c 1050–1300’, Studies in Medieval and Renaissance History (3rd series), 4, 57–106
Oliver, J R (ed), 1860–62 Monumenta de Insula Manniae, or a Collection of National Documents
Relating to the Isle of Man, 3 vols, Manx Society, Douglas
Ó Cuı́v, B, 1956–57 ‘A poem in praise of Raghnall, king of Man’, Éigse 8, 283–301
O’Meara, J J (ed), 1982 Gerald of Wales, the History and Topography of Ireland, Penguin, London
Oswald, H R, 1860 Vestigia Insulae Manniae Antiquiora, or a dissertation on the armorial
bearings of the Isle of Man, the regalities and prerogatives of its ancient kings . . ., Manx
Society, Douglas
Pálsson, H and Edwards, P (ed), 1978 Orkneyinga Saga: The History of the Earls of Orkney,
Penguin, London
Power, R, 2005 ‘Meeting in Norway: Norse-Gaelic relations in the kingdom of Man and the Isles,
1090–1270’, Saga Book 29, 5–66
Ridel, É, 2007 ‘From Scotland to Normandy: the Celtic sea route of the Vikings’, in B Ballin Smith,
S Taylor and G Williams (ed), West over Sea: Studies in Scandinavian Sea-borne Expansion
and Settlement Before 1300, Brill, Leiden, 81–94
Rixson, D, 1998 The West Highland Galley, Birlinn, Edinburgh
Robinson, J, 2004 The Lewis Chessmen, British Museum Press, London
Rymer, T (ed), 1739–45 Foedera, conventiones, literae et cujuscunque generis acta publica, inter
reges Angliae et alios quosvis imperators, reges, pontifices, principes, vel communitates habita
aut tractate, 10 vols, 3rd edn, The Hague
Simpson, G G and Galbraith, J (ed), 1986 Calendar of Documents Relating to Scotland Preserved
in the Public Record Office and the British Library, V (Supplementary) a.d. 1108–1516,
Scottish Record Office, Edinburgh
Stratford, N, 1997 The Lewis Chessmen and the Enigma of the Hoard, British Museum Press,
London
Stringer, K J, 2000 ‘Reform monasticism and Celtic Scotland: Galloway, c1140–c1240’, in E J
Cowan and R A McDonald (ed), Alba: Celtic Scotland in the Medieval Era, Tuckwell Press,
East Linton, 127–165
Stringer, K J, 2003 The Reformed Church in Medieval Galloway and Cumbria: Contrasts, Connec-
tions and Continuities, Friends of the Whithorn Trust, Whithorn
Theiner, A (ed), 1864 Vetera monumenta Hibernorum et Scotorum historiam illustrantia, Typis
Vaticania, Rome
Thorpe, L (ed), 1978 Gerald of Wales, the Journey Through Wales/The Description of Wales,
Penguin, London
Vigfusson, G (ed), 1878 Sturlunga Saga Including the Islendinga Saga of Lawman Sturla Thordsson
and Other Works, 2 vols, Clarendon Press, Oxford
Wilson, D M, 2008 The Vikings in the Isle of Man, Aarhus University Press, Aarhus
Wilson, J (ed), 1915 The Register of the Priory of St. Bees, Surtees Society, Durham
MAM 37: 23-caldwell – Press

chapter 23
THE SEA POWER OF THE WESTERN
ISLES OF SCOTLAND IN THE LATE
MEDIEVAL PERIOD

By David H Caldwell

From the early 14th century to the end of the 15th century the Western Isles of
Scotland (Figure 23.1) formed part of a vast lordship held by the MacDonalds
(Caldwell 2008a, 49–75). Their lordship represented a revival of the Kingdom of the
Isles which was ruled by a dynasty of kings based on the Isle of Man. Somerled
(Sumarliði), the ancestor of the MacDonalds, had annexed much of the kingdom in
ad 1156, and from then until ad 1266 there were two rival dynasties in the Isles, both
recognizing the kings of Norway as over-kings. By the Treaty of Perth in ad 1266
Man and the other isles were sold by the king of Norway to the king of Scots (Lustig
1979) and local leaders who might have regarded themselves as kings were reduced in
rank to magnates of Scotland. The MacDonalds, the most successful and powerful of
the kindreds descended from Somerled to have come through the Wars of Indepen-
dence, would have been seen in Gaelic society as kings, and they were certainly pre-
pared in the 14th and 15th centuries to challenge the authority of the Stewart dynasty,
not without success. This threat to the kings of Scots was finally removed by the
forfeiture of John II Lord of the Isles in ad 1494.
The Lords of the Isles stressed their Irish roots and were important patrons of
Gaelic culture, but in reality a good deal of Scandinavian blood flowed in their veins
and much of their way of life derived from a Viking past. The very first MacDonald,
Angus Mor (Áengus mac Domnall) of Islay, was one of the commanders of the inva-
sion fleet of King Hákon IV Hákonarson of Norway in ad 1263, and a contemporary
Irish praise poem describes how:
You have come round Ireland; rare is the strand which
you have not taken cattle from: graceful longships are sailed
by you, scion of Tara, you are like an otter.1

galleys and birlings

Ships underpinned the power of Angus Mor’s descendants. Indeed, the Lordship
of the Isles was a thalassocracy controlling the west coast of Scotland and the North
MAM 37: 23-caldwell – Press

sea power of the western isles 351

Figure 23.1 Map

showing the islands listed
in the 1596 report and the
extent of the Lordship of
the Isles (D H Caldwell
and Vicki Herring)

Channel between Scotland and Ireland. Its (war) ships were probably sailed and
carried up the Great Glen to the Moray Firth. Despite the forfeiture of ad 1494, island
clans retained considerable sea power throughout the 16th century. This was espe-
cially the case with Clann Iain Mhoir, a leading branch of the MacDonalds led by the
MAM 37: 23-caldwell – Press

352 maritime societies of the viking and medieval world

MacDonalds of Dunyvaig and the Glynns, based in Antrim, Islay and Kintyre, and the
MacLeans of Duart with their main base in the island of Mull (Caldwell 2008a,
77–103).
Considerable numbers of these ships could be amassed in fleets for major cam-
paigns, like the 160 Somerled is said to have sailed up the Clyde in ad 1164 (Anderson
1990, 255), and the 180 Donald Dubh, the pretender to the Lordship of the Isles,
sailed to Ireland in ad 1545 (State Papers Henry VIII 1834, 529). There is no overall
picture for any time in the medieval period of how many ships of war were available
in the Western Isles, where they were based and who owned them. No substantial
pieces of any of them survive, enough to make a convincing reconstruction. They are
known primarily from documentary sources and from images of them on medieval
West Highland sculpture (Caldwell 2007, 144–151; McDonald 2007a; Rixson 1998).
This paper is not the place to rehearse all that has been supposed or argued
about them. Some salient points do, however, have to be made. Firstly, there is
general agreement that they are descendants of the Viking longships that first plied
these waters in the late 8th century. Representations on West Highland sculpture
dateable from the 14th to the 16th century show clinker-built ships with high prows
and sterns, a mast with a square sail and several oar ports aside (Figure 23.2). In this
writer’s opinion much confusion has been caused by a desire to see these carvings as
scale models rather than images shortened and heightened to give them more pres-
ence, a process in stone carving that can be traced back, for instance, to the repre-
sentations of longships on the 11th- or 12th-century Heðinn stone at Maughold in the
Isle of Man (Wilson 2008, 69, fig 32) and a cross of similar date, possibly also carved
in Man, but in the Abbey Museum in Iona (Fisher 2001, 135, no 95).
The late 13th-century seal of Angus Mor shows another shortened and height-
ened ship but with the addition of a rudder (Caldwell 2007, 145, fig 1). Later repre-
sentations on West Highland sculpture usually have rudders, and that arguably is the
main development in the design of these ships since the Viking Age. Lack of other
solid information for changes in ship design should not be taken as evidence that these
ships did not change throughout the medieval period in shape, size and performance,
and in how they were used. Particular problems of comprehension relate to the twin
issues of their overall size and the number of men they could carry. Their actual length
and breadth will probably not be known unless appropriate wrecks can be recovered.
Early documents were concerned with the number of oars each had as a measure of
their size, and in this respect a key piece of evidence is the report to the Scottish Privy
Council in ad 1615 on ships in the Western Isles. It tells us that there were galleys with
between 18 and 24 oars (ie between 9 and 12 oars aside), and birlings with between 12
and 18 oars. There were also smaller boats with eight oars (Masson 1891, 347–348).
These figures for Western Isles’ ships in ad 1615 can be compared with the
vessels required for military service specified in a number of surviving land grants of
the late medieval period. For instance:
O Charter of ad 1315 by King Robert I to Colin son of Neil Campbell of the lands of
Lochawe and Ardskeodnish in Argyll for one ship with 40 oars, etc (Duncan 1988,
no 46);
MAM 37: 23-caldwell – Press

sea power of the western isles 353

Figure 23.2 The cross shaft of Ranald of Islay,

late 14th century, formerly on the island of Texa
but now in the Museum of Scotland. It is
decorated with a galley under sail. Note the oar
ports and rudder (D H Caldwell)

O Charter of ad 1343 by King David II to Torquil MacLeod of Lewis of the land of

Assynt for a ship of 20 oars (Webster 1982, 507);
O Charter of ad 1463 by John Earl of Ross (John II Lord of the Isles) to his brother
Celestine of the land of Sleat in Skye for one ship of 18 oars, etc (Munro and
Munro 1986, no 80);
O Charter by Angus Og (son of John II Lord of the Isles) in ad 1488 to Hector
MacLean of land (Torloisk) in Mull for the service of a ship of 22 oars (Munro and
Munro 1986, no 121);
O Charter of ad 1498 by King James IV to Alexander MacLeod of Dunvegan, of
lands in Skye formerly held of the Lord of the Isles: Harris, Duirnish, Minginish,
Bracadale, Lyndale and Trotternish, for a ship of 26 oars and two ships of 16 oars,
etc (Balfour Paul 1882, no 3284).
MAM 37: 23-caldwell – Press

354 maritime societies of the viking and medieval world

The ad 1615 report (Masson 1891, 347–348) also says that each oar in these ships was
pulled by three men, and hence we can deduce that a galley of 20 oars would have had
60 rowers and a birlinn of 12 oars would have had 36 rowers. These ships must have
been relatively broad-beamed to seat six men abreast, and there could hardly have
been space for any but a handful of men that were not at the oars. The muscle power
that propelled these crafts was provided by the warriors who used them for transport.
That this formula of three men to an oar was relevant in the 15th century is sug-
gested, or at least not denied, by a report giving the number of men and ships in the
fleet mustered in ad 1452(?) by Donald Balloch, the leader of Clann Iain Mhoir, to
raid in the Firth of Clyde. It was said by the ‘Auchinleck Chronicler’ to have had 5000
or 6000 men in 100 galleys, giving an average of 50 or more per ship (Craigie 1923,
221–222). Further back in time than that we cannot meaningfully push documentary
evidence for the size or capacity of Scottish ships, but King Hákon’s ship, in which he
sailed to Scotland in ad 1263, is said in Hákonar saga Hákonarsonar to have had
37 benches (and so 74 oars) and, most usually, four men in each ‘half-room’ (the
space occupied by a rowing bench) (Anderson 1990, 610 no 13, 614; Dasent 1894, 342,
345).2 Clearly, this was an exceptional, large ship with proportions that may have
differed considerably from others in his invasion fleet, but it does suggest that there
could have been three or even four men pulling on each oar.
We should be aware that evidence from the Northern Isles and Scandinavia in
the preceding centuries indicates ships with only one man per oar. That is the con-
clusion to be drawn from the sections on the leiðangr (system of coastal defence) in
early Norwegian laws. The 11th- to 12th-century Gulating Law required the oar of
each man who failed to appear for a levy to be raised as a sign of his absence, while
the Frostating Law, dating in its present form to about ad 1260, assumed two men to
a bench, and therefore one to each oar, in making provision for bows for the ships
(Larson 1935, 191, 319). The two warships excavated in Roskilde Fjord in Denmark,
‘Skuldelev 2’, built in Ireland about ad 1042, and ‘Skuldelev 5’, built in Denmark
about ad 1030, have proportions only really suitable for one man at each oar. The
former would probably have been propelled by 60 oars and had an overall length of
between about 29.2m and 31.2m and a maximum width of 3.76m. The latter was
originally 17.3m long and up to 2.47m wide. It was fitted for 26 oars and has been
identified by its excavators as a ship specifically for coastal defence (Crumlin-Pedersen
2002, 174–194, 246–278, 314). Accounts regarding Orkney of Earl Þorfinnr’s two ships
of 20 benches each with over 100 men (Pálsson and Edwards 19 81, 74) and of Earl
Haraldr Maddadarson’s ship in ad 1151, described as having 30 benches and 80 men
(Anderson 1990, 215), indicate only one or two men per oar.
The leiðangr in Norway and the leding known from early Danish sources (Lund
2003), relied on the local provision of ships rowed by men armed with their own
weapons. Both the Norwegian Gulating and Frostating Laws made the provision of a
bow and arrows for each thwart (rowing bench) obligatory (Larson 1935, 196, 319),
apparently envisaging that the ships would engage in battle at sea. Sea battles did take
place in Hebridean waters, as between Somerled and King Guðrøðr Óláfsson of
Man in ad 1156 (Broderick 1991, f 37v) and between John II Lord of the Isles and his
MAM 37: 23-caldwell – Press

sea power of the western isles 355

son Angus Og at Bloody Bay, Mull, about ad 1484 (Cameron 1894, 163; MacPhail
1914–34, 1, 47–50) but they may have been the exception, if not accidental. The
prime function of West Highland galleys and birlings was to carry troops for fighting
on land. Those with three men to an oar envisaged in the Privy Council report of ad
1615 may have evolved to carry the maximum number of warriors speedily on rela-
tively short sea crossings. In the latter part of the 16th century they had to be fast
enough to avoid, or even out-run, patrols of English ships guarding the approaches to
Ireland (Hayes-McCoy 1937, 102, 127, 130, 154, 251–252, 298).

a report of ad 1596
These galleys and birlings, manned by warriors, were the basis of the power of
the Lords of the Isles. Just how impressive that power was appears to be explained by
a late 16th-century report on the Western Isles, long known from one of its versions,
but the full significance of which has not been fully appreciated.
The version generally quoted was first published by William Skene in his Celtic
Scotland in ad 1880 (Skene 1880, 428–440). Skene believed it was an official report
written for King James VI that could be dated between ad 1577 and ad 1595. He
copied it from a manuscript (Adv MS 31.2.6) in the National Library of Scotland that
was formerly in the collection of Sir Robert Sibbald (1641–1722), Geographer Royal
to King Charles II. The report is in a slim volume that also contains Dean Monro’s
ad 1549 account of the Western Isles. Both are written in the same hand, the report as
if it were a continuation of the ad 1549 account. The hand has not been identified,
but appears to date to the late 17th century (Smith 1880, 436) and might be that of
a copyist paid by Sibbald. He had distributed an advertisement in ad 1682 with a
questionnaire, looking for information to aid him in producing a geography of
Scotland (Maidment 1833, 28), and it is possible that it was through this that he got
access to these two documents, perhaps both from the same informant. It has not
been possible to identify this source, but a likely candidate would have been
Archibald Graham, Bishop of the Isles. He is known to have been one of Sibbald’s
respondents, and had Jo Fraser, Dean of the Isles, write Sibbald a description of Iona
in ad 1693, although that does say the registers and records of Iona had already been
destroyed (Mitchell 1907, 217).
The author of the report and the specific circumstances of its production can be
identified.3 In early ad 15964 the English agent in Edinburgh, Robert Bowes, sent an
Edinburgh merchant, John Cunningham, to visit Lachlan MacLean of Duart to
report back on his ability and readiness to support the English administration in
Ireland against the uprising by the Earl of Tyrone and the extent to which Tyrone
could rely on other islanders to side with him (Giuseppi 1952, nos 126, 133).
Cunningham caught up with MacLean at Breachacha Castle in the island of Coll,
which explains why the report contains a detailed description of that castle, the only
one mentioned.
The report provides much geographical information on the islands, some of it of
dubious value, like the statement that the island of Skye is almost round, and lists
MAM 37: 23-caldwell – Press

356 maritime societies of the viking and medieval world

numbers of fighting men. It was clearly unsatisfactory in many ways, not least in not
providing information on the main players in the region, other gossip and any
recommendations on future action, all features of the contemporary report by Dionise
Campbell, Dean of Limerick, who had been sent by the English to sound out the Earl
of Argyll on his views (Giuseppi 1952, 201–211, no 173). It dealt with almost all of the
inhabited islands, irrespective of their ownership and the likelihood that the occupants
would turn out to fight in Ireland, but did not include other west coast areas like
Kintyre which regularly sent warriors there. All this no doubt explains why it never
got sent to London. Two considerably revised versions did, one in an unidentified
Scottish hand (TNA, SP 52/51/589 = Cameron 1936, no 198) mistakenly calendared
under ad 1593 since it was undated, and the other (TNA, SP 52/58/41 = Giuseppi
1952, no 143), in the handwriting of Bowe’s clerk, endorsed Edinburgh 21 March
1596. It is bound in volume SP 52/58 of Scottish papers immediately after a letter from
Lachlan MacLean from Coll, 18 March 1596 with an enclosure giving MacLean’s
answers to the questions (Giuseppi 1952, 158, no 133) sent by Bowes with Cunning-
ham (TNA, SP 52/58 nos 39 and 40 = Giuseppi 1952, no 142). It would appear likely
that these documents accompanied MacLean’s secretary, John Auchinross, and
Bowes’ servant, George Nicolson, when they reported to Sir Robert Cecil in London
soon afterwards (Giuseppi 1952, nos 152, 156).
For present purposes, there are three things to be noted that all three versions of
this report on the Western Isles have in common. Firstly, there is the statement that
the Hebrides were divided into four groups based on the islands of Lewis, Skye, Mull
and Islay. Secondly, there is the extent given of each island in terms of a valuation in
marks. Thirdly, there are the figures given for the number of men each island sup-
ported as fighting men. If the document emanated from John Cunningham’s trip to
Coll in ad 1596 it is clear that he must have relied on existing documentation to
acquire this knowledge. Much of that documentation, it would seem to the author, is
likely to have been describing a much earlier situation, possibly one that went back
hundreds of years. Certainly the fourfold division of the islands had no real relevance
beyond ad 1266 though it might well have been remembered in later times. The
inclusion of land valuations in marks might suggest a date after, perhaps not long
after, ad 1266 when this assessment would have been applied by Scottish officials to
the Western Isles, and possibly before ad 1275 (Mackie 1946, 145–146).
Some further clues to the time the documentation originated are perhaps pro-
vided by the actual listing of the islands. They are ordered from north to south, start-
ing with Lewis and ending with Bute.5 That is in reverse order to three other early
lists which see things from a Manx perspective and which may therefore ultimately be
dependent on official documentation from there (Barrow 1981, 12–13). These include
the ‘Manx Bull of 1231’, now believed to date to the late 14th century (Megaw 1978,
293–295, 313–314), the list in Fordun’s 14th-century Chronicle (Skene 1872, 39–40),
and the list in Monro’s ad 1549 account of the Western Isles (Munro 1961). So the
north to south ordering in the report might be an indication that the underlying
documentation came from a Norwegian source or originated from a Scottish one
post-ad 1266.
MAM 37: 23-caldwell – Press

sea power of the western isles 357

If a Norwegian source, Hákonar saga Hákonarsonar may provide the context. It

recounts how the Norwegian king, while his fleet was off Arran in ad 1263, sent emis-
saries to King Alexander III to try and agree terms of peace:
King Hakon had caused a list to be made of all the islands that he claimed for
himself, to the west of Scotland; and the king of the Scots had named those that he
would not let go: they were Bute and Arran, and the Cumbraes; but about the rest
there was little [conflict] between the claims of the kings. And yet the agreement
did not come off. (Anderson 1990, 623; cf Dasent 1894, 352–353)
There are also, perhaps, lessons to be learned from what islands are in the list.
Included are the islands of Arran and Bute in the Firth of Clyde, but not the Isle of
Man, the Cumbraes and the island of Kerrera beside Oban. The non-appearance in
the ad 1596 report of Man, long firmly in English hands, might be explained by its
removal by John Cunningham as having no relevance for a report bound for the
English government. If the treaty said to have been made in ad 1098 between King
Magnús Berfœttr of Norway and the king of Scots literally gave the former all the
islands to the west of Scotland (Duncan and Brown 1959, 192–194), then perhaps the
underlying documentation originated in a time in the mid-13th century or later when
the Scots might already have annexed the Cumbraes and Kerrera. That does not
explain the inclusion of Arran and Bute, both of which were in Scottish hands by
ad 1230 (Duncan and Brown 1959, 203), unless their presence relates to the claims
made by King Hákon during the abortive negotiations with King Alexander III in
ad 1263, or to the discussions surrounding the Treaty of Perth. King Hákon granted
both islands to supporters before leaving Scotland in 1263 (Anderson 1990, 635).
In favour of a Manx origin for the documentation is the level of detail provided
on some of the islands. Thus, while North Uist, Benbecula and South Uist are listed
merely as the island of Uist with a single quota of fighting men, Lewis and Harris,
although one island, are listed separately, and the entry for Skye in the version pub-
lished by Skene (1880, 432–433) gives the fighting strength for each of its five main
districts. These features might have resulted from the greater familiarity of a Manx
official with Lewis, Harris and Skye, believed to have long been under the control of
the Manx kings as a result of the ad 1156 agreement between King Guðrøðr Óláfsson
and Somerled (Broderick 1991, f 37v). It could also be argued that a Manx origin
would explain why Man itself was not included in the list of islands.
All that can reasonably be concluded from these pointers is that some of the
documentation underlying the report may be of considerable age, possibly from not
long after the acquisition of the Isles by the Scots in ad 1266. If this documentation
was made available by MacLean we might further speculate that, like Monro’s
ad 1549 account, it came from the archives of the bishops of the Isles. The MacLeans
of Duart had long had a close association with the bishopric and at the time of
Cunningham’s visit held many of the bishopric lands, including Iona and its monastic
buildings (Thomson 1888, no 1491). It would not be surprising if Lachlan had
acquired, or had access to bishopric documents.
Whatever the form, date or source of the underlying documentation, the likely
conclusion to be drawn about the quotas of fighting men is that they originated in the
MAM 37: 23-caldwell – Press

358 maritime societies of the viking and medieval world

Kingdom of the Isles, perhaps prior to the annexation of much of it by Somerled in

ad 1156, and probably prior to ad 1266. That is the logical deduction to be drawn
from the fact that the system of quotas is known only for the Isles and not for other
parts of the mainland.
Table 23.1 lists the quotas of fighting men that each island raised, a total of 7071
including Arran and Bute. There is no discernible consistent relationship between
the numbers of these, the extents in marks or the sizes of the islands, often noted as
so many miles in length and breadth. To paraphrase the information in the report,
using primarily the version published by Skene, these men, or gentlemen, made up a
professional force of warriors 6000 strong that was available at any time of the year
to go and fight elsewhere, for as long as was necessary, even for a whole year at a time
as was sometimes the case when their participation was required in the wars in
Ireland. They did not work the land, but those who did were specifically required to
stay at home.
Two additional pieces of information may derive directly from Lachlan
MacLean. Firstly (Skene version only), it is stated that Islay, in addition to its quota
of 800 men, supplied a gentleman from each markland (hence 360) to be household
men to their master (Caldwell 2008a, 143; Skene 1880, 438). This would appear to be
the bodyguard (Gaelic, luchd-taighe) of the Lords of the Isles, which according to a
late 17th-century source kept guard on the loch side adjacent to the Lords’ residence
at Finlaggan (Martin 1994, 273). Secondly, it is noted that only a third of the force of
6000 men were fully armoured with aketons (quilted coats), habergeons (mail coats)
and ‘knapshal bannetts’ (helmets). The revised versions add that the rest were
bowmen, many of them then with firearms.

the warriors
These warriors were not a feudal host, obliged to turn out to fight by virtue of
holding land, nor was their maintenance and readiness to fight anything to do with a
general obligation on men to provide ‘Scottish service’ in the royal army. They were a
professional fighting force, known in later medieval Lowland sources as caterans
(Gaelic, ceatharn). Nevertheless, when the Lords of the Isles and other chiefs in the
Isles turned out to fight for the kings of Scots it would have been these men who
would have made up their following. They may have been indistinguishable in
appearance and method of fighting from the warriors from Kintyre and other main-
land territories of the Lords of the Isles, or the warriors of Argyll and west coast clans
like the Campbells, but the implication of the ad 1596 report is that it was only in the
Isles that they were raised as a standing army. The versions of the report that reached
London also gave an overview of the Gaelic-speaking areas of the mainland and their
chief men, but provided no information on their fighting strength (Cameron 1936,
no 198; Giuseppi 1952, no 143).
One of the earliest specific identifications of the contribution of Isles men to a
Scottish royal army relates to ad 1307. Robert Bruce, his campaign to be recognized
as king all but dead, had just spent the winter as a fugitive in the Western Isles. Early
MAM 37: 23-caldwell – Press

sea power of the western isles 359

Table 23.1 The islands included in the 1596 report with their extents in pounds Scots or
marks, and the number of fighting men each was to produce

The Islands Extents Men Variations Comments

Lewis & Rona £40 700 North Rona, to N of Lewis, Skene version
only
Harris 20 marks 140
Uist 140 marks 600
Barra with other isles £20 200
Rona 80 marks 60 50 men Berneray? To W of Lewis
Pabbay 2 marks 40 between N Uist and Harris
Heisker 1 mark 20 or 24 men in Monach Isles, to W of N Uist
‘Colsmon’ - - ‘not inhabited’, not in Skene version
St Kilda - 0
Trotternish 80 marks 500 district of Skye
Sleat 30 marks 700 district of Skye
Strath (Shuardail) - - 160 men district of Skye
Waternish 20 marks 200 district of Skye
Duirinish 28 marks 240 district of Skye
Bracadale 16 marks 140 500 men district of Skye
Minginish - - district of Skye, not in Skene version
Raasay 8 marks 80
Eigg 30 marks 60 50 men
Rum 10 marks 6 or 7 men, 10 men
in TNA versions
Canna 6 marks 20
Muck 4 marks 16
Scalpay 4 marks 20 Scalpay? Between Skye and Raasay
Mull 300 marks 900
Lismore 80 marks 100 200 men in TNA
versions
Shuna & Shuna 8 & 4 marks 60 50 men Shuna, in L Linnhe, and Shuna beside
Luing
Ulva 12 marks 60 50 men
Gometra 4 marks 16 or 20 men
Inch Kenneth 4 marks 16 or 20 men off Mull
Iona 30 marks 0
Coll 30 marks 140
Tiree 140 marks 300
Islay 360 marks 800
Jura & Scarba 30 marks 100 Scarba only in Skene version
Colonsay & Oronsay 30 & 4 marks 100
Seil 60 marks 120 140 men in TNA
versions
Luing 40 marks 0 20 men in TNA
versions
Scarba 4 marks 17 Skene version only, also included with
Jura
Gigha 30 marks 100
Rathlin 30 marks 100
Arran 300 marks 100
Bute - 300
MAM 37: 23-caldwell – Press

360 maritime societies of the viking and medieval world

in the year he returned to his earldom of Carrick in Ayrshire with a force of Isles men,
and Irish contingents as well (Rothwell 1957, 370, 377–378). Scottish sources identify
a key supporter as Christina MacRuari, Lady of Garmoran, a lordship including the
Uists, Barra, Rum and Eigg as well as the mainland districts of Moidart, Arisaig,
Morar and Knoydart (Skene 1872, 335). These men would have been a significant
element in the army that Bruce led to victory at Loudoun Hill a few weeks later.
In ad 1369 they were the enemies and rebels that King David II had required
John I Lord of the Isles to expel when the latter submitted to royal authority (Munro
and Munro 1986, no 6). They were the standing force of men which a 17th-century
source says John I had under the command of Hector More Macillechoan
(MacCowan) for the defence of Lochaber and the frontiers of his land (Macphail
1914–34, 1, 25). When not at home in the Isles they could be quartered in surrounding
areas as an effective way of raising blackmail. The Earl of Moray agreed in ad 1394
to pay Alexander, the brother of Donald Lord of the Isles, a yearly sum of 80 marks
as protection money. Part of the deal was that caterans should not be a burden on the
lands of his earldom or the bishopric of Moray (Macphail 1914–34, no 14).
They were the men who went to fight in Ireland. Some of those who stayed there
founded families of galloglass – that is mercenary kindreds – like the MacSweeneys
and the MacDonnells, in some cases serving the same chiefly families for generations
(Nicholls 2007). There is no evidence for comparable mercenary kindreds forming in
Scotland. Service as a cateran or professional warrior may have been an opportunity
for leading young men to gain experience and boost their status prior to settling
down. It may have been a more permanent career for younger sons and others with
no prospects of acquiring land.
A number of 15th- and 16th-century effigies of warriors in churches and burial
grounds in the Western Isles and neighbouring parts of the mainland are shown as if
fastening on their sword belts, one hand grasping the scabbard and the other tugging
on the end of the belt that has been passed through its buckle (Figure 23.3). This pose
appears to be unique to this part of Europe, and it has recently been suggested that it
might be a deliberate piece of symbolism indicating group membership, in which
case of a warrior caste (Caldwell et al 2010, 41–42). Although swords also feature
prominently on many other medieval grave-slabs in the region, documentary sources
indicate that these warriors often fought as axe-men or bowmen (Caldwell 2007,
159–168).
They formed the army which Donald Lord of the Isles took to battle in ad 1411
in support of his claim to be recognized as Earl of Ross. It clashed with an army
drawn from the north-eastern Lowlands, led on behalf of the Crown by the Earl of
Mar, at Harlaw in Aberdeenshire. There was no agreed, clear-cut victory but a sense
of how alien the Islesmen were perceived to be was retained long afterwards in
Aberdeen. Leastwise, the historian, John Major, wrote in the early 16th century how
schoolboys re-enacted in their games this clash between the ‘wild’ and ‘civilized’
Scots (Major 1892, 348–349). It is perhaps not without relevance that a 17th-century
MacDonald source remembered that Donald’s army was 6000 strong (Macphail
1914–34, 1, 28–31). As noted above, the figure for the forces of the lordship mustered
MAM 37: 23-caldwell – Press

sea power of the western isles 361

Figure 23.3 Effigy of a warrior at the priory on the island of Oronsay, dating to the late 15th
or early 16th century. He is dressed in an aketon, mail collar and basinet, and is fastening on his
sword-belt. Two small figures are buckling his spurs (D H Caldwell)

by Donald Balloch about ad 1452 are said to have been 5000 or 6000 men (Craigie
1923, 221–222).
The implication of the ad 1596 report is that this system for maintaining a
professional force survived the demise of the Lordship of the Isles at the end of the
15th century. In July 1595 Lachlan MacLean of Duart had written to Robert Bowes
that he had sustained 600 men in garrison for three months with the aim of dis-
couraging Donald Gorm and Dunyvaig (MacDonald chiefs) from leaving their
Scottish lands undefended while they went off to Ireland in support of Tyrone, but as
this was an expensive proceeding, and as he was not encouraged by the English to do
so, he had recently disbanded the force (Cameron 1936, no 581). Lachlan was clearly
hoping, by impressing the English administration with his efforts, to receive payment.
Hence the mention of expenses. It is remarkable that he had been prepared and able
to keep such a force together for such a length of time with no payment and no
obvious prospects that the men could win booty. This is striking corroboration that
the quotas for warriors outlined in the ad 1596 report were no flight of fancy but a
functioning system.
At no point in the 16th century was there a leader in the Isles with sufficient
authority to muster the whole force, but clan chiefs like the MacDonalds of Dunyvaig
and the MacLeans of Duart clearly ensured the continuance of the social order which
allowed for the selection, training and maintenance of these warriors. They were thus
able to contract to supply large forces of men to fight in Ireland. They were known as
redshanks in Anglo-Irish sources and buannachan in Gaelic (Macinnes 1996, 57–59;
cf Hayes-McCoy 1937, 12–13, 72–76). For instance, in ad 1532 it was reported in
Edinburgh that Alexander MacDonald of Dunyvaig had taken a force of 4000 men to
Ireland in support of Sir Hugh O’Donnell (Gairdner 1880, no 1246) and in March
1596 Lachlan MacLean of Duart was offering to provide the English administration
in Ireland 2000 of his own men along with 1000 of the Earl of Argyll’s men (Giuseppi
1952, no 142).
MAM 37: 23-caldwell – Press

362 maritime societies of the viking and medieval world

The system came to an end at the beginning of the 17th century with a notable
shift in the political map of the British Isles consequent upon James VI of Scotland
becoming king of England and Ireland. The flight of the Earl of Tyrone and other
Ulster leaders in ad 1607 removed the main buyers of the services of Isles mercenaries,
and Isles chiefs were under severe pressure from stronger and more intrusive govern-
ment in Edinburgh. The defining moment may well have been the court held at Iona
in ad 1609 by the bishop of the Isles as the king’s representative, with the enforced
presence of the Isles chiefs. Included in the statutes drawn up and agreed were pro-
visions for the banning of the use of firearms and limitations on the number of men
maintained for war. No man was to be suffered to remain in the Isles unless he had
revenue, an income from rents or a trade. None of the chiefs was to maintain more
household men than could be supported by his rents, a leading player like Angus
MacDonald of Dunyvaig having to agree to keep only six gentlemen, and Lachlan
MacLean of Duart eight. Hostelries were to be established in each island so that there
would be no excuse for travellers or strangers living off the hospitality of the locals.
To leave no doubt that the oppression of the poor tenants and labourers of the
ground by buannachan was not in future to be tolerated these sorners (forceful
exacters of free board and lodgings) and idle bellies, as they were called, were to be
deemed thieves and intolerable oppressors, and treated accordingly (Masson 1889,
26–30). While not the subject of this paper, the redeployment of these warriors after
ad 1609 was to involve major upheavals and changes in Isles’ society (Macinnes 1996,
67–71).

the sea power of the isles — a system of naval defence?

Demonstrating that the Western Isles had the capacity to field a strong force of
ships and warriors is one thing. Understanding how this came about and the social,
legal or political mechanisms which maintained this capability, are quite other
matters. It is noteworthy that the ad 1596 report makes no mention of ships, only
men, while the grants of land known to have been made in return for ship service
would not seem to have been a means of raising the requisite number of ships, even
making a considerable allowance for the loss of early charters. A clear example of this
is provided by the ad 1498 charter, cited above, to MacLeod of Dunvegan which
required his service with a ship of 26 oars and two ships of 16 oars (Balfour Paul 1882,
no 3284). Assuming three men to an oar that implies a total complement of 174 —
say 200 at the most if steersmen and commanders are included. The ad 1596 report,
however, indicates that the lands of Harris, Duirnish, Minginish, Bracadale and
Trotternish, which were granted by this charter, supported a total force of 1140 men
— more if Lyndale in Skye, also listed in the charter, had a separate quota.
In any case, the supply of ship service as described in the charters was a feudal
obligation whereas the maintenance of a standing force was not. It would appear
more logical if the provision of men and the ships to convey them had the same roots,
although the lack of a consistent relationship between the extents of the islands listed
in the ad 1596 report and the standing forces they provided might suggest it is futile
to search for a formula of so many men per markland.
MAM 37: 23-caldwell – Press

sea power of the western isles 363

There is also no evidence for the commutation of ships into money payments,
with one exception, in the case of the lands of the Tenandry of Lossit in Islay. A
charter of ad 1617 by the bishop of the Isles required the grantee, Mr Thomas
Rollock, to provide annually a boat (cymba in the Latin text) with 14 oars for the two
and a half marklands (ie quarterland) of Lossit included in the tenandry, or else a
payment of £10 (Smith 1895, 353–358). The author has argued elsewhere that the
ship service in this charter may be of very early date, certainly medieval, but that it is
not like other charters requiring ships for military purposes (Caldwell 2008b). A
payment of £10 Scots from an Islay quarterland in ad 1617 appears to have been a
very heavy burden when compared with more typical sums of 10 shillings (= £0.50)
Scots in the charter of the island to John Campbell of Cawdor in ad 1614 (Thomson
1892, no 1137). It would thus not seem wise to extrapolate evidence for the service of
other ships in Islay from this data.
Some scholars, most recently Andrew McDonald (2007b, 206–207), have sup-
posed that much of the naval power of the Kingdom of the Isles stemmed from a
Scandinavian type of naval defence like the leiðangr described in the early Norwegian
Laws and the leding known from early Danish sources (Larson 1935, 188–200,
314–323; Lund 2003). The provision and maintenance of ships for defensive purposes
would thus have been tied to particular units of land, or perhaps islands as with the
quotas of men in the ad 1596 report.
While not unlikely, the evidence for a leding system in the Kingdom of the Isles is
distinctly lacking. In the case of the Isle of Man, support for its existence centres on
the equation of its six districts, known as sheadings, with skipreiður, each of which
would have been responsible for providing, maintaining and equipping at least one
vessel. This identification was aided by the supposed derivation of the word sheading
from Old Norse skeiðarþing, meaning an administrative district which provided a
warship, but Carl Marstrander, the Norwegian scholar who advanced this etymol-
ogy, later retracted it (Marwick 1949). Manx land units, forming sub-divisions of
the sheadings, have been supposed to be the equivalent of manngjerd (or lide),
Norwegian units that were responsible for providing an appropriately equipped man
for the leiðangr, but there has not always been agreement as to whether the equa-
tion should be made between Manx treens and manngjerd, or Manx quarterlands
(of which there were four to a treen) and manngjerd (Marwick 1949), and there has
been considerable doubt as to whether sheadings and treens on Man originated in
a Scandinavian milieu or are earlier in date (Megaw 1978, 279–288; Reilly 1988,
38–41). If already in existence prior to the arrival of the Norse, they may still, of
course, have been used by the latter as the basis for making naval or military levies.
Our knowledge of early land divisions in the Western Isles (for recent overviews,
see Connor and Simpson 2004, 645–653; Easson 1987; cf for Islay, Caldwell 2008a,
137–149) suggests a complex picture of different systems introduced at different times
in different places and, suffice it to say here, it has proven difficult to tease out sig-
nificant pieces of evidence for a leiðangr. There are a few documents that have to be
reviewed. The first is a letter by the pro-English Earl of Atholl to King Edward I of
England in ad 1304 which provides evidence for a relationship between ships and
MAM 37: 23-caldwell – Press

364 maritime societies of the viking and medieval world

units of land known as davachs, each of which equated to five or six marklands, or
20 pennylands (Connor and Simpson 2004, 658). The earl had received intelligence
from the earl and bishop of Ross that ‘Lochlan’ (the patriotic Lachlan MacRuairi)
and his friends had ordered that each davach of land should furnish a galley of 20 oars
(Bain 1881–88, 2, no 1633). Where this order was to be enforced is not specified, but it
presumably applied to the MacRuari lordship of Garmoran, including the Uists,
Barra and the Small Isles. There is no further evidence for this system for supplying
ships, and as a war-time measure it may have been of short duration. It may be
unwise to use this document, as others have done (Easson 1987, 7), as confirmation
that davachs, at least those on the coast, were routinely expected to supply a ship,
and each of a davach’s 20 pennyland divisions an oarsman.
Indeed, there is a lack of specific evidence from the Isles for davachs, or other
units of lands like ouncelands, pennylands and quarterlands, having been a basis for
the levying of men for warfare. That military service on the Scottish mainland as late
as the 14th century could relate to units of land is evident from charters (Barrow
2003, 247–248; Duncan 1978, 380–381; Easson 1987, 7–8). There is also the statement
by Fordun that King Alexander III had in ad 1264–65 planned to raise an army to
aid King Henry III of England on the basis of levying three men from every hide of
land (Skene 1872, 297). Duncan (1978, 380–381) has reasonably suggested that hides
in this context should be equated with units of land including davachs.
Jane Dawson has brought to the writer’s attention a letter of 5 May 1570 from
Archibald Campbell, 5th Earl of Argyll, to Colin Campbell of Glenorchy, advising
him that he has stented (selected on the basis of an assessment) a host for a muster at
Glenfalloch — one man from each markland, for a period of 8 or 10 days (Dawson
1977, no 93). The linkage of fighting men with marklands is clearly of interest in this
relatively late West Highland context, and is deserving of further study. The men do
not appear to have been a warrior cadre but a contingent of fencibles taken away
from their lands and jobs to do their duty on behalf of their landlord and clan chief.
None of this brings us any nearer to the basis of the quota system for men in the
Isles outlined in the ad 1596 report. Moreover, it is worth pointing out that surviv-
ing legislation relating to the defence of the realm, from the ad 1318 arming act of
Robert I (RPS 1318/29) onwards until ad 1603, never involved any territorial quota
system.
The last potential piece of evidence to be considered here is an act of the Scottish
Parliament in ad 1430 (RPS 1430/21) which required all barons and lords near the sea
in the west and north, especially opposite the Isles, to have galleys, one for each four
marks worth of land. Those that were required by their grants of land to have galleys
were to continue to provide them. The galleys were to be supplied by May 1431. This
act was to apply to all lands and lordships on the coast, and lands up to six miles
inland were to contribute to the upkeep and supply of the galleys. The act can be
taken as an admission that, at least as far as mainland lordships were concerned,
there was then no regular or thorough system for supplying ships already in place,
only a patchy or haphazard provision of certain ships as reddenda for holding lands.
The important thing to note is that the act applied to the mainland and was clearly
MAM 37: 23-caldwell – Press

sea power of the western isles 365

intended to counter the perceived threat from the Lordship of the Isles. While it was a
type of coastal defence like the leiðangr, there is nothing in the way it was drafted to
suggest that it had any antecedents in Scotland.
The lack of evidence for a leiðangr in the Isles may not result from the loss or
destruction of records but quite simply because there was no such thing. If Isles’
society was militarized from an early date there may have been no need for a leiðangr.
Providing the ships may have been what gave men status and power. They may often
have been gifts from kings or clan chiefs to their clients in return for making them
available for their own campaigns. A recent study of the 12th-century Irish Lebor na
Cert (Book of Rights) has concluded that it was by that means that royal fleets were
raised in ‘Viking Ireland’, and suggests the process was like an Irish version of the
Norse leiðangr (Swift 2004, 205). If this happened in 12th-century Ireland, it would
not be too difficult to believe it might have happened in the Western Isles.
The origins and exact nature of the sea power of the Western Isles in late medi-
eval times will always, through lack of hard facts, be the subject of speculation. What
cannot be doubted is that the ships and men of the Isles were a force to be reckoned
with. Until the end of the 16th century, local society was geared up for warfare and
could ship armies up to 6000 strong to fight in mainland Scotland and in Ireland.
That this naval and military capability was rooted in the Kingdom of the Isles
appears highly probable.6

notes
1 This translation has been provided by Hugh 5 Rathlin, now seen politically and geographically
Cheape (cf Bergin 1935, 65). as part of Ireland, is included.
2 A variant of the Frisbók manuscript version of the 6 The author is grateful to Jane Dawson for advice
saga cited in Anderson (1990, 610, no 13) says this on the Breadalbane Papers, and Hugh Cheape,
ship had only 27 benches. Alan Macniven and Andrew McDonald for read-
3 The account that follows is based on a detailed ing and commenting on a draft of this paper.
study of contemporary documentation mostly None of them bear any responsibility for the re-
now held in The National Archive at Kew. It maining imperfections. The section on ‘a report
provides an important insight into what was of ad 1596’ draws on a previously published note
happening in the Isles and Ireland at that time. (Caldwell 2013); the author and editor of this
Consideration of this material here is limited in volume are grateful to Nicholas Maclean-Bristol,
order to focus on military power and how the editor of West Highland Notes & Queries, for
report and related documents might be relevant allowing it to be substantially reproduced here.
to earlier times.
4 Dates have been modernized to reflect a year
beginning on 1 January.

references
Anderson, A O (ed), 1990 Early Sources of Scottish History, 500 to 1286, Volume Two, Paul
Watkins, Stamford
Bain, J (ed), 1881–88 Calendar of Documents Relating to Scotland, H M General Register House,
Edinburgh
Balfour Paul, J (ed), 1882 The Register of the Great Seal of Scotland ad 1424–1513, H M General
Register House, Edinburgh
Barrow, G W S, 1981 ‘The sources for the history of the Highlands in the Middle Ages’, in
L MacLean (ed), The Middle Ages in the Highlands, Inverness Field Club, Inverness, 11–22
MAM 37: 23-caldwell – Press

366 maritime societies of the viking and medieval world

Barrow, G W S, 2003 The Kingdom of the Scots, Edinburgh University Press, Edinburgh
Bergin, O, 1935 ‘An address to Aonghus of Islay’, Scottish Gaelic Studies 4, 57–65
Broderick, G (ed), 1991 Chronicles of the Kings of Man and the Isles, Manx National Heritage,
Douglas
Caldwell, D H, 2007 ‘Having the right kit: West Highlanders fighting in Ireland’, in S Duffy (ed),
The World of the Galloglass: Kings, Warlords and Warriors in Ireland and Scotland, 1200–
1600, Four Courts Press, Dublin, 144–168
Caldwell, D H, 2008a Islay: The Land of the Lordship, Birlinn, Edinburgh
Caldwell, D H, 2008b ‘The Lossit Cymba’, West Highland Notes & Queries series 3(12), 3–11
Caldwell, D H, 2013 ‘An intelligence report on the Hebrides, 1595/6’, West Highland Notes &
Queries series 3(23), 3–11.
Caldwell, D H, McGibbon, F, Miller, S and Ruckley, N A, 2010 ‘The image of a Celtic society —
medieval West Highland sculpture’, in P O’Neill (ed), Celts in Legend and Reality: Papers
from the Sixth Australian Conference of Celtic Studies, University of Sydney, Sydney, 13–59
Cameron, A, 1894 Reliquiae Celticae, Texts, Papers and Studies in Gaelic Literature and Philology,
Volume II: Poetry, History and Philology, ed A MacBain and Rev J Kennedy, The Northern
Counties Newspaper and Printing and Publishing Company Limited, Inverness
Cameron, A I (ed), 1936 Calendar of State Papers Relating to Scotland and Mary Queen of Scots
1547–1603, Volume 11 (1593–1595), H M General Register House, Edinburgh
Connor, R D and Simpson, A D C, 2004 Weights and Measures in Scotland: A European Perspec-
tive, National Museums of Scotland and Tuckwell Press, Edinburgh
Craigie, W A (ed), 1923 The Asloan Manuscript, Volume 1, Scottish Text Society, Edinburgh
Crumlin-Pedersen, O, 2002 The Skuldelev Ships I, The Viking Ship Museum, Roskilde
Dasent, Sir G W (ed), 1894 Icelandic Sagas, Volume 4: The Saga of Hacon and a Fragment of the
Saga of Magnus, HMSO, London
Dawson, J (ed), 1977 Clan Campbell Letters 1559–1583, 5th Series, Volume 10, Scottish History
Society, Edinburgh
Duncan, A A M, 1978 Scotland: The Making of the Kingdom, Oliver and Boyd, Edinburgh
Duncan, A A M (ed), 1988 Regesta Regum Scottorum V, Robert I 1306–1329, Edinburgh
University Press, Edinburgh
Duncan, A A M and Brown, A L, 1959 ‘Argyll and the Isles in the earlier Middle Ages’, Proceedings
of the Society of Antiquaries of Scotland 90, 192–220
Easson, A R, 1987 ‘Ouncelands and Pennylands in the West Highlands and Islands of Scotland’, in
L J Macgregor and B E Crawford (ed), Ouncelands and Pennylands, University of St Andrews,
St Andrews, 1–11
Fisher, I, 2001 Early Medieval Sculpture in the West Highlands and Islands, Monograph Series 1,
Royal Commission on the Ancient and Historic Monuments of Scotland, Edinburgh
Gairdner, J (ed), 1880 Letters and Papers Foreign and Domestic of the Reign of Henry VIII,
Volume 5, HMSO, London
Giuseppi, M S (ed), 1952 Calendar of State Papers relating to Scotland and Mary Queen of Scots
1547–1603, Volume 12 (1595–1597), HMSO, Edinburgh
Hayes-McCoy, G A, 1937 Scots Mercenary Forces in Ireland (1565–1603), Burns, Oates and Wash-
bourne, Dublin
Larson, L M, 1935 The Earliest Norwegian Laws: Being the Gulathing Law and the Frostathing
Law, Columbia University Press, New York
Lund, N, 2003 ‘Naval force in the Viking age and in high medieval Denmark’, in J B Hattendorf
and R W Unger (ed), War at Sea in the Middle Ages and the Renaissance, Boydell, Wood-
bridge, 25–34
MAM 37: 23-caldwell – Press

sea power of the western isles 367

Lustig, R L, 1979 ‘The Treaty of Perth: a re-examination’, Scottish Historical Review 58, 35–57
Macinnes, A I, 1996 Clanship, Commerce and the House of Stuart, 1603–1788, Tuckwell Press,
East Linton
Mackie, J D (ed), 1946 ‘Thomas Thomson’s memorial on old extent’, Stair Society 1946, 67–331
Macphail, J R N (ed), 1914–34 Highland Papers, Scottish History Society, Edinburgh
Maidment, J (ed), 1833 The Autobiography of Sir Robert Sibbald, Knt, M.D. to which is prefixed
Some Account of his MSS, Thomas Stevenson and John Wilson, Edinburgh and London
Major, J, 1892 A History of Greater Britain, Scottish History Society, Edinburgh
Martin, M, 1994 A Description of the Western Islands of Scotland circa 1695, Birlinn, Edinburgh
Marwick, H, 1949 ‘Naval defence in Norse Scotland’, Scottish Historical Review 28, 1–11
Masson, D (ed), 1889 The Register of the Privy Council of Scotland, Volume ix ad 1610–1613,
H M General Register House, Edinburgh
Masson, D (ed), 1891 The Register of the Privy Council of Scotland, Volume x ad 1613–1616,
H M General Register House, Edinburgh
McDonald, R A, 2007a ‘Dealing death from Man: Manx sea power in and around the Irish sea,
1079–1265’, in S Duffy (ed), The World of the Galloglass: Kings, Warlords and Warriors in
Ireland and Scotland, 1200–1600, Four Courts Press, Dublin, 45–76
McDonald, R A, 2007b Manx Kingship in its Irish Sea Setting 1187–1229, Four Courts Press,
Dublin
Megaw, B, 1978 ‘Norseman and native in the Kingdom of the Isles: a re-assessment of the Manx
evidence’, in P Davey (ed), Man and Environment in the Isle of Man, Part 2, Archaeopress,
Oxford, 265–314
Mitchell, Sir A (ed), 1907 Geographical Collections Relating to Scotland made by Walter Mac-
farlane, Volume 2, Scottish History Society, Edinburgh
Munro, J and Munro, R W (ed), 1986 Acts of the Lords of the Isles 1336–1493, Scottish History
Society, Edinburgh
Munro, R W (ed), 1961 Monro’s Western Isles of Scotland, Oliver and Boyd, Edinburgh
Nicholls, K, 2007 ‘Scottish mercenary kindreds in Ireland, 1250–1600’, in S Duffy (ed), The World
of the Galloglass: Kings, Warlords and Warriors in Ireland and Scotland, 1200–1600, Four
Courts Press, Dublin, 86–105
Pálsson, H and Edwards, P (ed), 1981 Orkneyinga Saga: The History of the Earls of Orkney,
Penguin, London
Reilly, P, 1988 Computer Analysis of an Archaeological Landscape: Medieval Land Divisions in
the Isle of Man, Archaeopress, Oxford
Rixson, D, 1998 The West Highland Galley, Birlinn, Edinburgh
Rothwell, H (ed), 1957 The Chronicle of Walter of Guisborough, Royal Historical Society,
London
RPS, The Records of the Parliaments of Scotland, http://www.rps.ac.uk
Skene, W F (ed), 1872 John of Fordun’s Chronicle of the Scottish Nation, Volume 2, Edmonston
and Douglas, Edinburgh
Skene, W F, 1880 Celtic Scotland: A History of Ancient Alban, Volume 3, David Douglas,
Edinburgh
Smith, G G, 1895 The Book of Islay, Privately printed
Smith, J A, 1880 ‘Additional notes to — notice of the remains of the great auk (Alca impennis,
Linn.) found in Caithness’, Proceedings of the Society of Antiquaries of Scotland 14, 436–444
State Papers Henry VIII, 1834 State Papers of King Henry the Eighth, Part III Continued: Corre-
spondence Between the Government of England and Ireland 1538–1546, His Majesty’s Record
Commission, London
MAM 37: 23-caldwell – Press

368 maritime societies of the viking and medieval world

Swift, C, 2004 ‘Royal fleets in Viking Ireland: the evidence of Lebor na Cert, a.d. 1050–1150’, in
J Hines, A Lane and M Redknap (ed), Land, Sea and Home: Proceedings of a Conference on
Viking-period Settlement, Maney/Society for Medieval Archaeology, Leeds, 189–206
Thomson, J M (ed), 1888 Register of the Great Seal of Scotland, Volume 5, ad 1580–1593,
H M General Register House, Edinburgh
Thomson, J M (ed), 1892 Register of the Great Seal of Scotland, Volume 7, ad 1609–1620,
H M General Register House, Edinburgh
Webster, B (ed), 1982 Regesta Regum Scottorum VI: The Acts of David II King of Scots 1329–1371,
Edinburgh University Press, Edinburgh
Wilson, D M, 2008 The Vikings in the Isle of Man, Aarhus University Press, Aarhus
MAM 37: 24-downham – Press

chapter 24
COASTAL COMMUNITIES AND
DIASPORA IDENTITIES IN
VIKING AGE IRELAND

By Clare Downham

The nature of cultural integration between Gaels and Scandinavians during the
Viking Age has been a much discussed topic in the fields of archaeology, history,
literature and linguistics over many years. This article is based on the premise that
comparative studies of cultural interactions in other contexts can shed light on issues
of integration and identity in the Insular Viking Age. In making these comparisons I
will dwell on two key features of Viking settlements in Ireland, namely their focus at
coastal locations and their role in overseas trade. It will be argued that merchant
communities dwelling on the Irish coast consciously fostered a hybrid Hiberno-
Scandinavian identity to facilitate their role as go-betweens in trade between Gaelic-
speaking peoples and the network of Viking traders which reached across medieval
Christendom and beyond. Over time, this hybrid Hiberno-Scandinavian identity
represented a prestigious level of social and economic status. It became desirable to
express elements of this identity in new areas as a result of secondary migration or
emulation. As the Vikings in Ireland can be seen to excel in particular economic roles,
comparisons may be drawn between their experiences and other coastal communities
which developed within historical trade diasporas.

the trade diaspora model

The theory of ‘trade diasporas’ was first put forward by Abner Cohen (1971) and
this theory was much developed in Philip Curtin’s monograph Cross-cultural Trade
in World History (Curtin 1984). Curtin’s study analysed communities of merchants
who settled in foreign lands, forming a network of commercial settlements as a trans-
national phenomenon which flourished in the pre-industrial age (Curtin 1984, 3). A
common feature of these settler-communities was the development of new hybrid
identities. The communities of migrants retained elements of their original culture,
but also integrated in varying degrees with their host society to serve as intermedi-
aries in overseas trade. This model provides some interesting parallels for the analysis
MAM 37: 24-downham – Press

370 maritime societies of the viking and medieval world

of Viking coastal communities in Ireland which developed their own distinctive

Hiberno-Scandinavian identity (Downham 2007a, 34–38).
Perhaps it is necessary first to analyse the terminology of the ‘trade diaspora’.
The noun diaspora is derived from Greek diasporá meaning ‘sown widely’
(Thompson 1998, 374, 372). Until the mid-20th century this term was used nearly
exclusively to describe the historical dispersion of the Jews. Since then the term’s
meaning has expanded to include other groups dispersed from a real or imagined
homeland. The usefulness of the term ‘diaspora’ has recently come under attack from
several quarters, either because its meaning has been over-stretched and applied to a
bewildering range of real-life situations, or because the theories of diaspora have
become ahistorical conceptualizations divorced from reality (Brubaker 2005, 3;
Gourgouris 2005; Levi 2001, 87–88).1 Some scholars have used the term ‘merchant
network’ in particular situations in preference to ‘trade diaspora’ (Markovits 2002,
22). ‘Network’ is a more flexible term than ‘diaspora’, as it may allow for the transfer
of elements of culture rather than perceived identities. This could fit better with what
we observe when aspects of Scandinavian culture are disseminated beyond Viking
communities in a colonial context. Nevertheless, in an Insular Viking Age context
Scandinavian identity, whether in claims of ancestry, language or material expression
seems to have played a part in the lives of those we would describe as ‘Vikings’
(Downham 2009, 140, n 4).2 There is therefore a risk that the term ‘network’ could
sometimes separate us too radically from the concept of a homeland or an imagined
homeland which is embedded in theories of diaspora (Brubaker 2005, 5;
Tsagarousianou 2004, 56–59).3
A further comment on these labels is that ‘trade’ can be applied more broadly
than that of ‘merchant’. It would not be contentious to argue that most exchange in
the Viking Age operated beyond the hands of those we could identify as merchants.
Hiberno-Scandinavian culture clearly functioned beyond merchant networks.
However, it can be argued that this hybrid culture did evolve prestigious conno-
tations of mercantile and political power. Beyond the hubs of Hiberno-Scandinavian
contact individuals utilized or imitated features of this culture, perhaps deliberately
associating themselves with a broader ‘merchant network’ or a Viking ‘trade
diaspora’. This issue will be further explored at a later stage in this article.

hiberno-scandinavian culture and some parallels

A hybrid identity seems to have evolved from a fairly early stage among Viking
groups in Ireland. In the late ad 850s the emergence of a group called Gall-goı́dil
(‘foreigner-Gaels’) is attested in Irish chronicles, where they are associated both with
the south-west and the north of Ireland.4 Intermarriages between Vikings and Gaels
are also recorded from the 9th century, in particular between the Viking leaders of
Dublin and the Northern Uı́ Néill dynasty (Downham 2012a, 26–27; see Figure 24.1).
These early strategies of integration may have been necessary for the political and
economic success of Viking settlements in Ireland, particularly after attempts to
establish wide-ranging territorial power or influence in the ad 840s were beaten down
MAM 37: 24-downham – Press

coastal communities and diaspora identities 371

Figure 24.1 The northern Uı́ Néill and the Vikings of Dublin: royal interactions

in a series of Irish military victories.5 News of the Irish success against Vikings spread
as far afield as the Carolingian court (Nelson 1991, 66, sa 848).
In the central decades of the 9th century a string of Viking bases were established
around the Irish coasts (Downham 2010; see Figure 24.2). Some of these bases devel-
oped into major ports including Dublin, Waterford and Limerick. While these sites
may have begun as raiding bases, their success over time would depend more on a
regular supply of goods traded from the surrounding hinterland in exchange for
foreign imports (Wallace 1987). To achieve this, Viking leaders can be seen to culti-
vate good relations with some groups within Irish society. These alliances may have
been more than military ventures as they could also facilitate the development of a
trading network between Viking camps and the Irish interior. Irish kings stood to
benefit from the economic advantages of commerce with Vikings, as well as access
to prestige goods from overseas imported through Viking trading networks. The
foundation of Viking bases in the north-east of Ireland and possibly at Rosnaree in
the Boyne valley may be linked to recorded Viking alliances with kings Aed mac Néill
and Cinaed mac Conaing in the central years of the 9th century (Downham 2009). It
has long been recognized by economic historians that individuals will seek to secure
the stability and predictability of markets by embedding economic ties within a nexus
of social commitments (Oka and Kusimba 2008, 365; see Chapter 12). Ties of
reciprocity between Viking and Irish populations through alliance and kinship could
therefore be seen as aids to the development of cross-cultural trade (cf Sindbæk 2008,
154–155).
Some Viking ports developed an urban character in the course of the 10th cen-
tury. Nevertheless, these continued to maintain a strong military dimension. Viking
settlements were commonly locked in war with one or more of the Irish kings and
often groups of Vikings and leaders of individual bases acted in competition with
each other for power or influence (Downham 2007b). As such the Vikings in Ireland
may represent what Curtin has termed a ‘militarised trade diaspora’ (Curtin 1984, 5,
32–35, 236). The development of Hiberno-Scandinavian culture can be seen to facil-
itate cross-cultural exchange. Nevertheless, leaders and communities who bore this
MAM 37: 24-downham – Press

372 maritime societies of the viking and medieval world

Figure 24.2 Possible

locations of 9th-century
Viking camps (Vicki
Herring after Downham
2010)

culture competed side by side with each other. It has sometimes been assumed that
common culture and co-operation go together, and indeed intercultural and intra-
cultural divisions and conflict seem to be an under-researched area in diaspora studies
(Brubaker 2005, 13). Even in contemporary commercial networks, international
business people often adopt common cultural practices such as wearing Western-style
suits and speaking English, but they are nonetheless working in a competitive
environment. Their assumed identity may announce status rather than implicit
harmony.
In seeking to interpret the Hiberno-Scandinavian nexus in the context of other
trade diasporas, I will draw on a broad range of comparisons in the remainder of this
article. The reader may notice that particular attention has been given to Swahili and
Peranakan cultures as a source for comparison. Along the eastern Africa coast the
interaction of Arab traders from the 9th century led to the development of Swahili
MAM 37: 24-downham – Press

coastal communities and diaspora identities 373

culture. In Malaysia the settlement of Chinese traders from the late Middle Ages led
to development of Peranakan (sometimes called ‘Baba’ or ‘Straits Chinese’) identity.
These cultures were chosen for more in-depth research in part because I had found
them to be interesting before starting this project. I do not see particular affinities
between Hiberno-Scandinavian cultures and these groups over other trade diasporas.
Rather, by focusing on particular case studies I hope to show the potential use of
comparisons between trade diasporas for interpreting aspects of Viking culture in
Ireland.
The analysis of trade diasporas highlights identity as a complex phenomenon
(including its material, spiritual or intellectual properties). A wide variety of under-
lying factors appear to influence the level of cultural borrowing and hybridization in
the spheres of language, religion, and custom when two groups come into contact.
For example, one could ask why did Insular Vikings adopt Christianity rather than
follow the traditional religion of Scandinavia and why did not significant numbers of
Irish people adopt the belief systems of the Scandinavians? In other diaspora situ-
ations the imported religion was sometimes maintained and it could even displace the
belief system of indigenous communities. For example, bearers of Peranakan culture
are characterized by their observation of Chinese traditional religious practices rather
than the Islamic faith of the neighbouring Malays. In eastern Africa, Islamic beliefs
were followed in Swahili coastal ports and this led to the conversion of some inland
communities. The reasons for Viking conversion in Ireland could include the intoler-
ance of Christianity to polytheism which was perceived as a barrier in cross-cultural
relations. Scandinavian traditional religion was abandoned as a hybrid Hiberno-
Scandinavian culture developed (Downham 2012, 30–31). Furthermore, traditional
Scandinavian religion was primarily observed in an agrarian setting at ‘home’, but
Christianity was in the 9th century the religion of European towns as well as the
Irish interior. It was therefore beneficial to adopt Christianity in terms of creating a
trading network among co-religionists. It could also be fostered as part of an urban
‘cultural package’. In other words, Christianity may have been regarded as part of the
culture of transnational trade, networking and economic success (cf McPherson 1993,
133). In comparing the Vikings in Ireland to other trade diasporas many differences
emerge which may invite particular angles for analysis. Furthermore, despite the
bewildering variety of cultural responses which result from contact between two
different societies, there are also interesting points for comparison.
Curtin has made the point that physical barriers in the landscape as dividing
lines between diverse environments tend to promote cross-cultural trade (Curtin
1984, 15). Many trade diasporas have a maritime dimension. The Vikings who settled
in Ireland were largely restricted to coastal settlements and denied the possibility of
large-scale land-taking, partly due to the politically fragmented character of the Irish
landscape. The technological superiority of Viking ships and the Vikings’ trans-
marine contacts were, however, well suited to generating wealth through trade and
piracy, and people of Hiberno-Scandinavian culture can be seen to have successfully
exploited this economic niche.6 The sea had a strong influence on Viking culture in
general and it served as a uniting feature between disparate communities of Norse
speakers. In a similar way, the sea also looms large in the diaspora of other ‘aquatic
MAM 37: 24-downham – Press

374 maritime societies of the viking and medieval world

cultures’ (Curtin 1984, 20–27). Viking Age burials along the Irish littoral (at Rathlin
Island and Larne, Co Antrim; Ballyholme, Co Down; Eyrephort, Co Galway;
between Three Mile Water and Arklow, and possibly at Morragh Co Wicklow) may
be interpreted through a comparative lens (Ó Floinn 1998, 144–147; see Chapter 20).
Mortuary rituals were enacted on the shore in some other coastal societies where
lived experience was dominated by the interplay of sea and land (for example, among
the ‘Saltwater people’ of the Torres Strait in Australia, McNiven 2003). The coast
could be seen to represent a liminal space which was appropriate in symbolizing the
passage between life and death, this world and the next.7 The Vikings, like other
coastal societies, were a sea-orientated people for whom water represented a route
way rather than a boundary, perhaps the inverse of modern territorially based per-
ceptions (Shnukal 2004, 2). For Vikings in Ireland, connection with lands across the
sea helped them to maintain an identity which was hybrid and yet distinct from their
Gaelic neighbours.
The place of Vikings in Ireland as partially integrated but separate was also
consciously maintained in the development of origin legends. These legends can be
compared with other diaspora myths. According to Gerald of Wales, writing in the
late 12th century, the Viking towns of Dublin, Waterford, and Limerick were each
founded by one of three brothers from Norway. The names of the brothers — Olaf,
Sitric and Ivar — occur frequently in the royal dynasties of these ports. The brothers
are said to have arrived in the guise of peaceful traders, but soon turned against
their Irish neighbours (O’Meara 1948–50, 175; 1982, 122). Another legend in the
12th century Latin Historia Gruffud ap Cynan credits the foundation of Dublin to the
famous Norwegian king Harald Finehair, whose brother is said to have founded
Waterford. Harald is further identified as the great-grandfather of Olaf Cuarán who
ruled Dublin in the 10th century (Russell 2006, 54–56).
In East Africa Swahili speakers emphasize their difference from their neighbours
by stressing their descent from immigrants (Nurse and Spear 1985, 4). Different
myths circulate in individual towns and families, but a link to the Shirazi of Persia is
common. In the town of Kilwa a tradition is that Sultan Ali bin Selimani came from
Persia in a ship with his children and trade goods. The Sultan plied the local elder,
Mrimba with gifts, marrying his daughter and obtaining the right to settle. In another
version of the legend Sultan Ali sailed down the coast with his father and six brothers
in seven ships, each stopping to found a different town along the way. Other legends
link the Swahili with Arabia (Nurse and Spear 1985, 70–79; Wynne-Jones 2007, 370).
In the Peranakan society of Malaysia a prominent tradition is that a Chinese princess
Hang Li-Po married Sultan Mansur Shah of Melaka in the 15th century and that her
handmaids also married with local men (Tan 1993, 61). Common points in these
legends are that visitors from overseas of royal or noble stock unite with indigenous
people and generate lines of illustrious descendants. These myths therefore high-
light the distinctiveness of trader communities, endowing them with an exotic and
prestigious ancestry, but one which is also bicultural. The myths can also be seen to
express statements about the nature of the relationship between the ‘outsiders’ and
‘indigenous’ peoples.
MAM 37: 24-downham – Press

coastal communities and diaspora identities 375

It can be argued that trade diaspora communities remained deliberately distinc-

tive in their host societies, treading a careful line between integration and oblivion. In
addition to a diaspora mentality of perceiving links with people overseas (which are
reinforced by origin legends) this sense of being different was expressed by Vikings in
Ireland through language, material culture and custom. In Dublin bilingualism or
multilingualism characterized life in the port into the 12th century as attested by
inscriptions showing familiarity with Gaelic, Norse and English (Downham 2007a,
36–37). At one stage the inhabitants of the Viking ports may have spoken a type of
pidgin or heavily accented Gaelic (Johnson 2004, 85). A number of Middle Irish texts
refer to the speech of Vikings in Ireland as faltering or stammering (Quin 1983, sv got).
Similarly in Airec Menman Uraird Maic Coisse, a tale dated to the late 10th century,
the speech of the Gall-goı́dil is referred to as gicgoc (probably an onomatopoeic word,
a bit like ‘gabble’ or ‘gobbledegook’) and the talk of merchants as gibgab; barely
intelligible gabbling (Bergin et al 1907–13, ii, 72, lines 6–9).8 Language is also a dis-
tinctive feature of Peranakan and Swahili cultures. According to Tan Chee-Beng ‘The
most distinct aspect of Baba [ie Peranakan] culture is the Baba dialect’ (Tan 1979, 20).
This dialect of Malay takes a number of loan words from Hokkien in particular
relating to the spheres of kinship and ritual (Tan 1993, 32, 67). The Baba dialect
currently has low prestige from the perspective of Malay and Hokkien speakers, so
that Peranakan children are often educated in Malay, Hokkien or English, rather
than their mother tongue (Tan 1993, 32, 68). The Swahili language has a higher
prestige along Africa’s central east coast, although many Swahili speakers can speak
at least one other language (Nurse and Spear 1985, 61). Swahili has a basic sound
system and grammar akin to Bantu, but it is characterized by a large number of loan
words from Arabic, as well as borrowings from neighbouring African languages,
Persian, and various Indian and European languages through trade and other contact
(Nurse and Spear 1985, 6, 14). Early traveller accounts indicate that Arabic words
were used in Swahili before the 12th century, although Derek Nurse and Thomas
Spear have argued that the main period of borrowing occurred after 1500 (Nurse and
Spear 1985, 15). Swahili can therefore be regarded in its early development as a dis-
tinctive dialect of a Bantu rather than an Arabic creole (as it has sometimes been
described). In each case, language can be seen as an expressive vehicle for defining
and maintaining identity in a diaspora context, whether through multilingualism or a
distinctive take on the host language.
I have argued previously that in terms of architecture and material culture (for
example in terms of food preparation) the inhabitants of Dublin maintained an
identity which was distinctive and somewhat eclectic in comparison to their Irish
neighbours up to and during the 12th century (Downham 2007a, 34–38; but see also
Chapter 15).9 Features of this urban Hiberno-Scandinavian identity may have also
applied in other Viking ports. Again comparison may be drawn with other trade
diaspora groups including the Swahili and Peranakan cultures. Among the Swahili,
similarities as well as contrasts can be observed in the material culture of their
medieval towns and their rural hinterlands. Locally produced pottery has been found
in both urban and rural zones at Kilwa Kisiwana dating from the 9th to the
MAM 37: 24-downham – Press

376 maritime societies of the viking and medieval world

14th century, but certain types of red burnished ware were more characteristic of the
town than the hinterland and some classes of imported ware from the Arabian pen-
insula and China were emphatically coastal in their distribution (Wynne-Jones 2007,
376). The architecture of Swahili towns was also distinctive from the African interior,
being in style (and in the use of coral as a building material) superficially similar to
Arabian models (Nurse and Spear 1985, 16). Similarly in modern Peranakan culture,
distinctive forms of identity are expressed in architecture and household goods which
combine Chinese and Malay styles (Tan 1993, 65–67). This cultural eclecticism
in coastal trading communities may result from individuals’ desires to display access
to exotic goods or fashions as well as concerns to maintain a distinctive cultural
identity.
In addition to material culture, public events and customs may be regarded as
important venues for exercising communal identity. Diaspora communities provide
interesting case studies for exploring this issue. Among the Yunnanese Chinese in
Chiang Mei (north Thailand), traditional Chinese rituals are employed at funerals as
conscious expressions of Chinese heritage, but some mortuary traditions are adopted
from the indigenous populations (Maxwell 1998, 133–135). In Peranakan society,
customs which combine elements of Chinese and Malay culture may also be observed.
Marriage rituals are, however, seen to follow the pattern of ‘old fashioned’ Chinese
weddings, thus preserving the more archaic customs of the homeland (Tan 1993, 65).
Public or customary observances in a diaspora context may therefore be seen as either
conservative and nostalgic, or forward looking and creative in bringing a distinctive
or innovative cultural element to traditional observances. Roza Tsagarousianou has
argued that:
Diasporic experiences, even their apparently more traditionalist variants, should
not be dismissed simplistically as backward-looking, as they are almost invariably
constituting new transnational spaces of experience that are complexly interfacing
with the experiential frameworks that both countries of settlement and purported
countries of origin represent. (Tsagarousianou 2004, 57)
This observation complements David Griffiths’ interpretation of mid-10th-
century Viking mortuary monuments around the British coasts of the Irish Sea as
‘involving ‘‘conservative’’ elements (the mound burials) and forward-looking elements
more in tune with the contemporary tradition of the host country’ (Griffiths 2004,
133). A similar interpretation of conservatism and innovation in mortuary com-
memorations may be applied to Viking burials in Ireland, as part of this ‘Irish Sea
theme’ (Griffiths 2004, 131; see also Chapter 20).
The importance of public rituals in maintaining a diaspora identity may be
relevant when considering the role of the thing mound which stood in Viking Age
Dublin, and the regalia (the sword of Carlus and the ring of Tomar) which is associ-
ated with the Viking kings of that port in the 10th century (Downham 2007c, 7–9).10
Sadly the nature of public rituals in the Viking ports of Ireland is a matter largely
beyond historical view. Nevertheless, it is possible to envisage that such events re-
enacted archaic patterns of behaviour from the Viking homelands. They might also
embody social or cultural change through the creative interpretation of traditional
MAM 37: 24-downham – Press

coastal communities and diaspora identities 377

rituals or the addition of indigenous practices.11 Custom and ritual could be sig-
nificant media for expressing transitions of identity in a new cultural environment.

status and dissemination

The Vikings’ ability to dominate in the sphere of Irish overseas trade is wit-
nessed by the comments of medieval observers (James 1978). Gerald of Wales out-
rageously claimed that the Irish were too lazy to trade and so ‘they thought it useful
indeed that a people who would bring them the products of other regions . . . should
be admitted into the kingdom’ (O’Meara 1948–50, 175; 1982, 122). Mary Valante has
also drawn attention to the lack of status or protection afforded to foreign or Irish
traders in early Irish law tracts (Kelly 1988, 7; Valante 2008, 35). Valante has argued
that prior to the Viking Age, exports and imports were channelled through royally
sponsored centres such as Dalkey Island in Dublin Bay or Dunnyneill Island on
Strangford Lough (Valante 2008, 34). This evidence could imply that traders were
heavily dependent on elite patronage and protection, for they were assigned little
status in their own right. Alternatively, as trading was of low regard or significance, it
may have been largely shunned as a primary occupation (that is people conducted
trade alongside other activities). The low social status of trade is a common feature in
pre-industrial hierarchical societies where there is a marked division between manual
labourers and aristocrats (Bonacich 1973, 583; Curtin 1984, 5–6, 18). More broadly,
trading is seen in many societies as inherently untrustworthy or immoral (Curtin
1984, 5–6; Oka and Kusimba 2008, 359). As trading was an undesirable profession in
Irish society this created a niche which could be exploited by foreign entrepreneurs.
Thus we may see in Ireland a situation comparative to the preconditions of other
‘trade diasporas’ and foreign ‘middleman minorities’ which have been investigated by
economic historians (eg Bonacich 1973; Oka and Kusimba 2008, 345, 349–350, 359).
The Vikings’ military prowess and rapid generation of wealth in Ireland may
mean that they were feared and distrusted by their neighbours. Nevertheless, while
their economic and political power remained strong, Vikings could neither be
assigned a uniformly low status nor ignored. I have already suggested that one moti-
vation for Irish kings to ally with Vikings was to gain access to commercial networks,
whereby they might accrue wealth through exports, or acquire high-status exotica to
redistribute along networks of patronage or gift exchange. Over time therefore
aspects of the foreign traders’ civilization could become attractive and useful to local
elites. This point has been argued by Kenneth McPherson (1993, 134) in relation to
the spread of Swahili culture along Africa’s eastern coast in the Middle Ages.
In Ireland we find goods of Scandinavian or Viking cultural identity distributed
beyond the hubs of Gaelic–Scandinavian interaction. Sometimes this is assumed to
reflect the movement of Viking raiders and traders. However, these diffusions may
also occur because Irish people consciously adopted elements of Scandinavian
culture. John Bradley has commented that:
The close similarity between the material culture of Dublin and that of the rural
Irish settlement sites such as Knowth, Ballinderry 1, and Beal Boru, shows a high
MAM 37: 24-downham – Press

378 maritime societies of the viking and medieval world

degree of interaction between the culture of the countryside and that of the port
towns. The material culture is neither purely Scandinavian nor purely Irish, but is
rather a common culture which one may term Hiberno-Scandinavian or Hiberno-
Norse. The existence of this common culture makes it virtually impossible to
identify archaeologically a rural ‘Scandinavian’ rather than a rural ‘Irish’ settle-
ment of tenth to twelfth century date. Such a distinction is in fact a political one.
(Bradley 1988, 60)
Knowth, Ballinderry Crannog and Beal Boru are all high-status sites under Irish
rule, located away from the centres of Viking power. In architectural terms they are
of Irish construction, but nonetheless all contain a significant proportion of Viking
and Hiberno-Scandinavian style artefacts ranging in date across the 10th and 11th
centuries. For example, at Ballinderry objects of high status include a Viking sword,
the famous Hiberno-Scandinavian wooden gaming board and a silver kite brooch
which may have been manufactured in Dublin (Johnson 1999, 68). The inhabitants of
these sites seem to have associated themselves with lucrative trading networks and
the transnational mercantile culture of the powerful Hiberno-Scandinavian ports.
At sites of lower status such as Truska and Beginish, on Ireland’s western littoral,
coastal dwellers may have consciously adopted elements of Viking mercantile culture
which bore prestigious connotations (Gibbons and Kelly 2003; Sheehan et al 2001).
These coastal dwellers may also have intermingled with Viking visitors who plied the
local sea lanes. In addition, Irish entrepreneurs of the 10th and 11th centuries who
wanted to enter the fast-growing field of overseas trade may have assumed aspects
of Hiberno-Scandinavian identity in order to get a footing in existing mercantile
networks.
Objects associated with Hiberno-Scandinavian identity were not only dissemi-
nated in an Irish context. This is clearly witnessed in the distribution of various
brooch and pin styles along Viking trading routes. Ringed pins were adopted from
the Irish as a form of dress-fastener in the 9th century (Johnson 2004, 83). Various
stylistic developments of this simple object which took place in a Hiberno-
Scandinavian milieu have been charted by Thomas Fanning. These pins were widely
exported and imitated across the Viking world (a famous example being discovered
at L’Anse aux Meadows in Newfoundland, Fanning 1994, 34). Other jewellery styles
associated with Ireland are the thistle and bossed penannular brooches which emerge
in the mid- to late 9th century and kite brooches which appear as a new form from
the 10th century (Edwards 1990, 145; Johnson 2004, 83–84). We see evidence of
brooch and pin styles being exported from Ireland to Viking settlements in north
Wales, Galloway, northern England and beyond the Irish Sea through the network
of the Scottish islands (where a closely associated Gaelic-Scandinavian culture
flourished), to the North Atlantic. The dissemination of Hiberno-Scandinavian style
may also be evaluated through art and artefact forms in other media including carved
stones and bone combs across the Irish Sea (Ashby 2006, 224, 226–227, 231, 235–236,
242; Chapter 19; Bailey 1984, 12–13, 28–30).
The geographical expansion in expressions of Hiberno-Scandinavian culture is
also replicated in non-artefactual media including personal names and saints cults.
MAM 37: 24-downham – Press

coastal communities and diaspora identities 379

This topic was explored in a British context by early 20th-century scholars W G

Collingwood and A H Smith, and has recently been reinvigorated and refined by the
work of Fiona Edmonds (Collingwood 1927; Edmonds 2009; Smith 1927). In Iceland,
a significant influx of Gaelic-speaking peoples and Gaelic culture can be dated back
to the Viking Age, although it is not always easy (or credible) to separate out the dif-
ferent strands of Scottish and Irish influence (Sigurðsson 2000).
Søren Sindbæk’s (2007) theory of the ‘small world’ of Viking networks may be
relevant to explain how a modest number of travellers could have spread Hiberno-
Scandinavian artefacts and styles. However, these travellers could only have made
such an impression if Hiberno-Scandinavian culture was considered prestigious
through bearing connotations of wealth or political power. The volume of Viking
silver hoards recovered from Ireland and the pan-Insular ambitions of that island’s
Viking kings may provide some explanation for a wider desire to acquire or imitate
Hiberno-Scandinavian fashions. In Ireland and across Viking networks of com-
munication it can be argued that emblems of Hiberno-Scandinavian culture were
appropriated and utilized. The individuals who did so sought to enhance their status,
perhaps by indicating their participation in overseas trade, or to demonstrate their
knowledge of current fashions, or to associate themselves with the mercantile culture
of powerful Hiberno-Scandinavian sea kings.
A series of historical factors can lead to the decline of once successful diaspora
cultures. Hiberno-Scandinavian identity did not flourish much beyond ad 1200 in
Ireland. This can be attributed to English conquest of the Viking trading ports in
the ad 1170s. Nevertheless, the Hiberno-Scandinavian communities of Dublin and
Waterford were initially given protected and preferential status by English kings
when their situation is compared to the native Irish (O’Byrne 2004, 115). In other
trade diaspora situations incoming colonial powers often associate with existing
‘foreign’ trading elites to help enforce their power. Existing cultural networks associ-
ated with a trade diaspora community could be used to their advantage. In this way
Swahili was adopted as a language of government in colonial east Africa, encourag-
ing its further spread inland, and Peranakan traders in the Straits of Malacca were
given support to extend their ventures into Singapore under English rule (Freedman
1960, 27). It may be that a decline of Hiberno-Scandinavian culture had set in before
ad 1170. This decline may be attributed to a southward shift in the prevailing
patterns of transnational trade and the growth of national kingships.12
Hiberno-Scandinavians had lost political power in England during the 10th
century and had fallen increasingly under the power of provincial Irish kings in the
11th century. This leads me to a final question as to whether Hiberno-Scandinavians
came to be regarded as something of a pariah society, a characteristic which is not
uncommon in diaspora communities (tragically exemplified in the history of the
Jews: Curtin 1984, 5, 18; Markovits 2002, 20). Despite many alliances and inter-
marriages, after four centuries of settlement in Ireland Vikings were still described as
‘foreigners’ (gaill) and so regarded as alien from a Gaelic cultural perspective. The
Vikings’ economic success as traders may have stirred up jealousy and mistrust among
the Irish with whom they interacted on a daily basis. We see in Irish literature of the
MAM 37: 24-downham – Press

380 maritime societies of the viking and medieval world

11th and 12th centuries the targeting of the Vikings as ‘whipping boys’ in the creation
of a unified national identity (eg Ó Corráin 1978, 31–32). In this literature the
depredations of historical Vikings were discussed in order to justify the subjection of
contemporary Hiberno-Scandinavians, and their status as foreigners was contrasted
with the cultural unity of the Irish. While these literary representations suggest a
decline in the prestige of Hiberno-Norse culture, ecclesiastical metalwork of the 11th
and 12th centuries tells a somewhat different story (Murray 2004, 141). The popu-
larity of Ringerike and Urnes motifs on reliquaries from this period indicates that the
eclectic fusion of Gaelic, Scandinavian and English art styles found in medieval
Dublin had a certain caché, evoking wealth, urban identity and external connections
(Lang 1987, 178; Ó Floinn 1987, 179). The power and prestige of Hiberno-Norse
culture would linger on until the Viking towns of Ireland were brought firmly under
English control.
In conclusion, it can be argued that merchant communities dwelling on the Irish
coast consciously fostered a hybrid Hiberno-Scandinavian identity in the 9th century
and beyond to facilitate their role in trade between Gaelic-speaking peoples and a
transnational ‘Viking’ network. Over time, this hybrid Hiberno-Scandinavian iden-
tity came to represent a prestigious level of social and economic status. Elements of
this identity were then adopted more widely as a result of secondary migration from
the Viking ports and because of elite emulation. The rise and decline of Hiberno-
Scandinavian culture may be interpreted in light of the economic and political
parallels which may be drawn with other historic coastal trade diasporas.

notes
1 Since this article was written several relevant 7 This is also relevant for the interpretation of
works have emerged, including Jesch (2008), Viking Age ship burials, although only one such
Abrams (2012), Barrett (2012), Downham burial has been tentatively identified in Ireland at
(2012b), Glørstad (2012), Edmonds (2014) and Ballyholme, Co Down (Ó Floinn 1998, 146–147;
McLeod (2014). cf Ballard et al 2003, 388; Knapp 2008, 16).
2 It should be noted that the term Viking is itself 8 For the dating of this text, see Carey (1997, 50).
used by different scholars to mean different 9 With regard to dress see also the comment by
things. Ruth Johnson (2004, 83) that women in Dublin,
3 One could draw on the diaspora theory of York and Lincoln wore fitted caps as ‘an impor-
‘imagined homeland’ to include those who tant part of a Viking woman’s ethnic costume,
embrace diaspora culture within the host com- helping to distinguish her from local women’.
munity, and come to regard themselves as having 10 The use of conservative and innovative cultural
distant blood-ties with the imagined homeland emblems side by side is also relevant for the
4 Mac Airt and Mac Niocaill (1983, 314–315) saa interpretations of images on Viking coins minted
855 [=856].3, 856 [=857].1; Hennessy (1866, in England and the designs on carved stones
154–157) saa [856], [857], [858]. raised in areas under Viking rule. The conscious
5 Mac Airt and Mac Niocaill (1983, 306–307) saa combination of archaic/Scandinavian and indi-
846 [=847].4, 847 [=848].5; Hennessy (1866, genous/forward looking concepts may be
146–149) saa [847], [848], [849]. implied in the configuration of heathen and
6 In northern Scotland Vikings held more terri- Christian images on opposing sides of a coin or a
tory, but the potential to develop large-scale monument.
trading posts was limited. James Barrett (2007) 11 For the transformative role of ritual see Stacey
has shown how in Orkney Viking maritime (2007, 51–52).
technology was applied to fishing and piracy to 12 This pattern may be exemplified in the decline of
generate wealth for the elite beyond land-based Chester and the growth of Bristol as ports of
activities. trade in the 11th century.
MAM 37: 24-downham – Press

coastal communities and diaspora identities 381

references
Abrams, L, 2012 ‘Diaspora and identity in the Viking Age’, Early Medieval Europe 20, 17–38
Ashby, S, 2006 Trade and identity: bone and antler combs in the early medieval North Atlantic
region. Unpublished PhD thesis, University of York, York
Bailey, R N, 1984 ‘Irish Sea contacts in the Viking Period — the sculptural evidence’, in G Fellows-
Jensen and N Lund (ed), Tredie Tvaerfaglige Vikingesymposium, Hikuin, Copenhagen, 7–36
Ballard, C, Bradley, R, Nordenborg Myhre, L and Wilson, M, 2003 ‘The ship as symbol in the
prehistory of Scandinavia and Southeast Asia’, World Archaeology 35, 385–403
Barrett, J, 2007 ‘The pirate fishermen: the political economy of a medieval maritime society’, in
B Ballin Smith, S Taylor and G Williams (ed), West Over Sea: Studies in Scandinavian Sea-
borne Expansion and Settlement Before 1300, Brill, Leiden, 299–340
Barrett, J H (ed), 2012, Being an Islander: Production and Identity at Quoygrew, Orkney,
ad 900–1600, McDonald Institute for Archaeological Research, Cambridge
Bergin, O J, Best, R I, Meyer K, and O’Keeffe, J G (ed), 1907–13 Anecdota from Irish Manuscripts,
5 vols, Hodges Figgis, Dublin
Bonacich, E, 1973 ‘A theory of middleman minorities’, American Sociological Review 38, 583–594
Bradley, J, 1988 ‘The interpretation of Scandinavian settlement in Ireland’, in J Bradley (ed),
Settlement and Society in Medieval Ireland: Studies Presented to F.X. Martin O.S.A., Boethius
Press, Kilkenny, 49–78
Brubaker, R, 2005 ‘The ‘‘diaspora’’ diaspora’, Ethnic and Racial Studies 28, 1–19
Carey, J, 1997 ‘The three things required of a poet’, Ériu 48, 41–58
Cohen, A, 1971 ‘Cultural strategies in the organization of trading diasporas’, in C Meillassoux
(ed), The Development of Trade and Indigenous Markets in West Africa, Oxford University
Press, Oxford, 266–278
Collingwood, W G, 1927 ‘Christian Vikings’, Antiquity 1, 172–180
Curtin, P D, 1984 Cross-cultural Trade in World History, Cambridge University Press, Cambridge
Downham, C, 2007a ‘Living on the edge: Scandinavian Dublin in the twelfth century’, in B Ballin
Smith, S Taylor and G Williams (ed), West Over Sea: Studies in Scandinavian Sea-borne
Expansion and Settlement Before 1300, Brill, Leiden, 33–52
Downham, C, 2007b ‘Irish chronicles as a source for inter-Viking rivalry ad 795–1014’, Northern
Scotland 26, 51–63
Downham, C, 2007c Viking Kings of Britain and Ireland: The Dynasty of Ívarr to ad 1014,
Dunedin Academic Press, Edinburgh
Downham, C, 2009 ‘‘‘Hiberno-Norwegians’’ and ‘‘Anglo-Danes’’: anachronistic ethnicities in
Viking Age England’, Mediaeval Scandinavia 19, 139–169
Downham, C, 2010 ‘Viking camps in ninth century Ireland: sources, locations and interactions’, in
S Duffy (ed), Medieval Dublin 10, Four Courts, Dublin, 93–125
Downham, C, 2012a ‘Religious and cultural boundaries between vikings and the Irish: the evidence
of conversion’, in J Nı́ Ghradaigh and E O’Byrne (ed), The March in the Medieval West, Brill,
Leiden, 15–34
Downham C, 2012b ‘Viking ethnicities: a historiographic overview’, History Compass, 10, 1–12
Edmonds, F, 2009 ‘Hiberno-Scandinavian influence east of the Pennines’, in Between the Islands:
Interaction with Vikings in Ireland and Britain, http://www.crassh.cam.ac.uk/flexshare/
papers1/BetweenIslands/FionaEdmonds_Paper.pdf
Edmonds, F, 2014 ‘Saints’ cults and Gaelic-Scandinavian influence around the Cumberland coast
and north of the Solway Firth’, in J V Sigurðssonn and T Bolton (ed), Gaelic-Norse Relation-
ships in the Irish Sea in the Middle Ages 800–1200, Brill, Leiden, 39–63
Edwards, N, 1990 The Archaeology of Early Medieval Ireland, B T Batsford, London
MAM 37: 24-downham – Press

382 maritime societies of the viking and medieval world

Fanning, T, 1994 Viking Age Ringed Pins from Dublin, Royal Irish Academy, Dublin
Freedman, M, 1960 ‘Immigrants and associations: Chinese in nineteenth-century Singapore’, Com-
parative Studies in History and Society 3, 25–48
Gibbons, E K and Kelly, E P, 2003 ‘A Viking Age farmstead in Connemara’, Archaeology Ireland
17(1), 28–32
Glørstad, Z T, 2012 ‘Sign of the times? The transfer and transformation of penannular brooches in
Viking-Age Norway’, Norwegian Archaeological Review 45, 30–51
Gourgouris, S, 2005 ‘The concept of diaspora in the contemporary world’, in I Baghdiantz
McCabe, G Harlaftis and I Pepelasis Minoglou (ed), Diaspora Entrepreneurial Networks:
Four Centuries of History, Berg, Oxford, 383–390
Griffiths, D, 2004 ‘Settlement and acculturation in the Irish Sea region’, in J Hines, A Lane and
M Redknap (ed), Land, Sea and Home: Proceedings of a Conference on Viking-period Settle-
ment, Maney/Society for Medieval Archaeology, Leeds, 125–138
Hennessy, W M (ed), 1866 Chronicum Scotorum: A Chronicle of Irish Affairs, from the Earliest
Times to ad 1135, with a Supplement Containing the Events from ad 1114 to ad 1150, Rolls
Series, London
James, D, 1978 ‘Two medieval Arabic accounts of Ireland’, Journal of the Royal Society of
Antiquaries of Ireland 108, 5–9
Jesch, J, 2008 ‘Myth and cultural memory in the Viking diaspora’, Viking and Medieval Scandi-
navia 4, 221–226
Johnson, R, 1999 ‘Ballinderry Crannóg no. 1: a reinterpretation’, Proceedings of the Royal Irish
Academy 99 C, 23–71
Johnson, R, 2004 Viking Age Dublin, Townhouse, Dublin
Kelly, F, 1988 A Guide to Early Irish Law, Dublin Institute for Advanced Studies, Dublin
Knapp, A B, 2008, Prehistoric and Protohistoric Cyprus: Identity, Insularity and Connectivity,
Oxford University Press, Oxford
Lang, J, 1987 ‘Eleventh-century style in decorated wood from Dublin’, in M Ryan (ed), Ireland and
Insular Art ad 500–1200, Royal Irish Academy, Dublin, 174–178
Levi, S C, 2001 The Indian Diaspora in Central Asia and its Trade 1550–1900, Brill, Leiden
Mac Airt, S and Mac Niocaill, G (ed), 1983 The Annals of Ulster (to ad 1131), Dublin Institute for
Advanced Studies, Dublin
Markovits, C, 2002 The Global World of Indian Merchants, 1750–1947, Cambridge University
Press, Cambridge
Marstrander, C, 1916 ‘Remarques sur les Zur keltischen Wortkunde, I–VI de Kuno Meyer’, Revue
Celtique 36, 335–410
Maxwell, A, 1998 Merchants and Migrants: Ethnicity and Trade Among Yunnanese Chinese in
Southeast Asia, Yale University Southeast Asia Studies, New Haven
McLeod, S, 2014 The Beginning of Scandinavian Settlement in England: The Viking ’Great Army’
and Early Settlers, c. 865–900, Brepols, Turnhout
McNiven, I J, 2003 ‘Saltwater people: spiritscapes, maritime rituals and the archaeology of
Australian indigenous seascapes’, World Archaeology 35, 329–349
McPherson, K, 1993 The Indian Ocean: A History of People and the Sea, Oxford University Press,
Oxford
Murray, G, 2004 ‘The arm-shaped reliquary of St Lachtin: technique, style and significance’, in
C Hourihane (ed), Irish Art Historical Studies in Honour of Peter Harbison, Four Courts
Press, Dublin, 141–164
Nelson, J L (ed), 1991 The Annals of St-Bertin: Ninth Century Histories, Volume I, Manchester
University Press, Manchester
MAM 37: 24-downham – Press

coastal communities and diaspora identities 383

Nurse, D and Spear, T, 1985 The Swahili: Reconstructing the History and Language of an African
Society, 800–1500, University of Pennsylvania Press, Philadelphia
O’Byrne, E, 2004 ‘Cultures in contact in the Leinster and Dublin marches, 1170–1400’, in S Duffy
(ed), Medieval Dublin 5, Four Courts Press, Dublin, 111–148
Ó Corráin, D, 1978 ‘Nationality and kingship in pre-Norman Ireland’, T W Moody (ed),
Nationality and the Pursuit of National Independence, Appletree Press, Belfast, 1–35
Ó Floinn, R, 1987 ‘Schools of metalworking in eleventh and twelfth century Ireland’, in M Ryan
(ed), Ireland and Insular Art ad 500–1200, Royal Irish Academy, Dublin, 179–187
Ó Floinn, R, 1998 ‘The archaeology of the early Viking Age in Ireland’, in H B Clarke, M Nı́
Mhaonaigh and R Ó Floinn (ed), Ireland and Scandinavia in the Early Viking Age, Four
Courts Press, Dublin, 131–165
Oka, R and Kusimba, C M, 2008 ‘The archaeology of trading systems, part 1: towards a new trade
synthesis’, Journal of Archaeological Research 16, 339–395
O’Meara, J J (ed), 1948–50 ‘Giraldus Cambrensis in Topographia Hibernie: text of the first
recension’, Proceedings of the Royal Irish Academy 52 C, 113–178
O’Meara, J J (ed), 1982 Gerald of Wales: the History and Topography of Ireland, Penguin,
Harmondsworth
Quin, E G (ed), 1983 Dictionary of the Irish Language. Compact Edition, Royal Irish Academy,
Dublin
Russell, P (ed), 2006 Vita Griffini filii Conani: The Medieval Latin Life of Gruffud ap Cynan,
University of Wales Press, Cardiff
Sheehan, J, Stumann Hansen, S, and Ó Corráin, D, 2001 ‘A Viking Age maritime haven: a reassess-
ment of the island settlement at Beginish, Co Kerry’, Journal of Irish Archaeology 10, 93–120
Shnukal, A, 2004 ‘Introduction’, in A Shnukal (ed), Navigating Boundaries: The Asian Diaspora in
Torres Strait, Pandanus Books, Canberra
Sigurðsson, G, 2000 Gaelic Influence in Iceland. Historical and Literary Contacts: A Survey of
Research, University of Iceland Press, Reykjavı́k
Sindbæk, S M, 2007 ‘The small world of the Vikings: networks in early medieval communication
and exchange’, Norwegian Archaeological Review 40, 59–74
Sindbæk, S M, 2008 ‘Local and long distance exchange’, in S Brink and N Price (ed), The Viking
World, Routledge, London, 150–158
Smith, A H, 1927 ‘Some aspects of Irish influence on Yorkshire’, Revue Celtique 44, 34–58
Stacey, R C, 2007 Dark Speech: The Performance of Law in Early Ireland, University of
Pennsylvania Press, Philadelphia
Tan, C-B, 1979 ‘Baba Chinese, Non-Baba Chinese and Malays: a note on ethnic interaction in
Malacca’, Southeast Asian Journal of Social Science 7, 20–29
Tan, C-B, 1993 Chinese Peranakan Heritage in Malaysia and Singapore, Penerbib Faja Bakti,
Kuala Lumpur
Thompson, D (ed), 1998 The Concise Oxford Dictionary of Current English, Oxford University
Press, Oxford
Tsagarousianou, R, 2004 ‘Rethinking the concept of diaspora: mobility, connectivity and communi-
cation in a globalised world’, Westminster Papers in Communication and Culture 1, 52–65
Valante, M, 2008 The Vikings in Ireland: Settlement, Trade and Urbanisation, Four Courts Press,
Dublin
Wallace, P F, 1987 ‘The economy and commerce of Viking Age Dublin’, in K Düwel, H Jankuhn,
H Siems and D Timpe (ed), Untersuchungen zu Handel und Verkehr der vor- und frühge-
schichtlichen Zeit in Mittel- und Nordeuropa: Teil IV: Der Handel der Karolinger- und
Wikingerzeit, Vandenhoeck & Ruprecht, Göttingen, 200–245
Wynne-Jones, S, 2007 ‘Creating urban communities at Kilwa Kisiwani, Tanzania, ad 800–1300’,
Antiquity 81, 368–380
MAM 37: 25-index – Press

INDEX
Note that majors rulers and individuals with patronymics and/or bynames
are alphabetised by first name

A’ Cheardach Bheag, 285, 288–289 Aspa Löt, 92

Aa, River, 130 assembly, 55, 64, 66, 88, 92, 131, 164, 249, 376
Aamot, 301 Assynt, 353
Aberdeen, 360 Æthelweard, 36
Abusa, 96 Atholl, Earl of, 363
Adam of Bremen, 69, 74 Auchinross, John, 356
Aed mac Néill, 371 Augustinian Order, 343
Affrica, daughter of Fergus of Galloway, 342 Aumale, 142
Affrica, daughter of Guðrøðr Óláfsson, 339 Avaldsnes, 179
Africa, 3, 374–376, 379
afterlife, 171, 315, 374 Bagge River, 74
Agder, 190 Baldwin II, Count of Flanders, 125
agriculture, 54, 74, 81, 126, 148, 167, 176, 226–227, Ballanorris, 212
245, 253–255, 291, 335 Ballaquayle Hoard, 229
Åhus, 162, 165 Ballateare, 310–311
Åland, 47 Ballinaby, 311–312
Ālande River, 51 Ballinderry, 266, 308, 378
Alexander, brother of Donald Lord of the Isles, 360 Balloch, Donald, 354, 361
Alexander II of Scotland, 340 Ballyholme, 299, 374
Alexander III of Scotland, 340, 357, 364 Balnakeil, 307, 315
Alfred the Great, 69 Baltic Sea, 25, 41–42, 44, 46
Ali bin Selimani, Sultan, 374 Baltic States, 27, 41
alliances, 76, 83, 371, 379 Baltic Ware, 28, 30, 34–35, 47, 70, 84, see also
Almindingen, 75 pottery
Alrekstad, 184, 186 Bamburgh, 146
Althing, 64, 66, see also assembly Bandlundeviken, 29
amber, 2, 51, 83, 103, 105, 114, 225, 228 Bangor, 299, 343
ancestors, 310, 374 Banks, Sir Joseph, 230
Angus Mor of Islay, 350, 352 Bantu, 375
Angus Og, son of John II, 353, 355 Barra, 360
antiquarianism, 222, 299–300, 308, 333, 344 Beachview, 227, 264, 266–267
antler, 190, 194, 225, 228, 259, 263, 272, 279–281, beads, 73, 78, 82, 225
285, 289–291 amber, 83, 225, 228
worked, 225, 281 glass, 42, 77, 162, 165, 225
Antwerp, 129 silver, 77–78
Archaeology Data Service, 210 Slavic, 78
Ardskeodnish, 352 steatite, 241–242
Argyll, Earl of, 356, 361, 364 Beal Boru, 378
Arlanda, 89 Beckside, 144
Ärlinghundra, 89, 91 Beginish, 205, 212, 378
Arnórr Jarlaskáld, 329 Benedictine Order, 343
armies, 2, 6, 207, 345, 360, 364, see also Great Army Bengtsson, F G, 14
armour, 358 Beowulf, 99, 148
Arnulfus I, Count of Flanders, 125 Bergen, 5, 8, 182–195, 264, 267, 279
Arran, 357–358 Bishops of, 188, 192, 195
art, 4, 83, 177, 210, 228–229, 291, 380 churches, 187–192
MAM 37: 25-index – Press

index 385
divisions, 183–184 high status, 44, 55, 315
trade, 191–192, 194 mass, 6
Bergues, 128–129 pagan, 228
Beverley, 144–145, 148–149, 268 prehistoric, 171, 221
Biervliet, 128 weapon, 45–46, 50, 299, 301, 303–304, 310–311,
Birka, 22, 30, 81, 102–103, 112, 162, 260, 262, 313–315, see also grave goods
264–266 Burstwick, 142–143, 148–149
Birsay, 220–221, 227, 264–267 Bute, 204, 356
Birsay-Skaill Landscape Archaeology Project, 220 Byzantium, 22, 43, 47–48, 51, 62, 321
Bjäresjo, 96
Bjärkoarätt, 30 Cairston, 230
Blackhammer, 281 Calais, 129
Black Sea, 25, 47 Campbell, Archibald, 364
Blankenberge, 129 Campbell, Dionise, Dean of Limerick, 356
Bogeviken, 29 Campbell, Neil, 352
Bolmsö, 96 Campbell of Cawdor, John, 363
bones Campbell of Glenorchy, Colin, 364
osteology, 315 Canna, 324
worked, 190, 194, 245, see also faunal remains Carisbrooke, 140
borders, 103, 162–163, 373 Carlingford, 341
Borg, Lofoten, 226 Carlus, sword of, 376
Bornais, South Uist, 201, 211, 228, 237–239, 243–244, Carolingian Empire, 2, 7, 108–109, 111, 115, 119, 163
247–248, 253–255, 264–268, 272, 286–287, 290 Carrick, 360
Bornholm, 2, 69–76, 78–85, 160 Caspian Sea, 25, 47, 63
Baltic influences, 79–82, 84–85 castles, 129, 149, 230
burials, 76, 85 Castletown Bay, 336
English contact, 82 cattle, 227, 250–252, 291, 335, 350, see also faunal
hoards, 75, 80 remains
power, 76 cattle husbandry, 288
settlement, 75, 84–85 Cecil, Sir Robert, 356
Borre, 177 Celestine, brother of John II, 353
Bostadh, Great Bernera, 201, 210, 285–287, 290 cemeteries, 27, 37, 45–46, 55, 70, 73, 77, 95, 103, 111,
Bourbourg, 125 310, 315
Bowes, Robert, 355–356, 361 central places, 52, 54–55, 88–89, 92, 96, 106, 138, 140,
Bracadale, 353, 362 145–149, 160, 164, 186, 188, 190
Bråviken, 101–102 administration of, 141, 149
Breachacha Castle, 355 Charlemagne, 31, 109
Breckness, Laird of, 230 Charles II of England, 355
Breckon, Sands of, 268 Chester, 272, 343
Bredene, 124 Chiang Mei, 376
bridges, 140–141, 145, 147 China, 373, 376
Bristol, 8, 272 chivalry see knighthood
brochs, 221, 265 Christianity, 44–45, 55, 66, 78, 106, 190, 310, 341,
Broderick, George, 337, 339 343, 373, see also conversion
bronze, 31 symbols, 18, 20
Bronze Age, 47, 54, 89, 174, 179, 281 Christian names, 338
Brouster, Scord of, 281 Church, 6, 67, 69, 74, 81, 182, 190–191, 230, 340, 342
Bruges, 126–130 churches, location of, 55
Bryant’s Gill, 208 Cille Pheadair, 201, 211, 228, 237, 248, 264–266, 270,
Buckquoy, 228, 264, 285 286–287, 290
bullion, 4, 222, see also currency Cinaed mac Conaing, 371
burial mounds, 92–93, 95, 376 Cistercian Order, 66, 336, 343–345
burial rituals, 44–45, 78, 287–288, 301, 304, 307–308, Cladh Hallan, 287
310–311, 374, 376 Clann Iain Mhoir, 351, 354
burials see also cremations climate change, 48, 124, 231, 255
boat, 15, 95, 171–172, 177–180, 305 Clontarf, Battle of, 232
bog, 15, 171–173, 175–176, 178–180 Clyde, 352, 354, 357
child, 225 Cnip, 225, 264–265, 270, 285–286, 310
Christian, 77, 313 Cnut the Great, 3, 76, 82–83, 184
female, 45, 50, 73, 77, 177–180, 301, 314 Cnut the Holy, 74, 83
MAM 37: 25-index – Press

386 maritime societies of the viking and medieval world

coastal locations, 147–148, 150 Daugava River, 44–45, 47–51, 57
coffins, 77, 118, 171, 311–312 Daugmale, 47, 49–51, 56
cog, 139 David I of Scotland, 340–341, 343
coinage, 4, 28, 32, 74, 82, 139, 163–164, 167 David II of Scotland, 353, 360
counterfeit, 29–30, 32 Deerness, Brough of, 265
Dorestad, 119 deer, 279–281, 285–291, 293, 295, 297
English, 28, 75, 82, 229 husbandry, 288
German, 28, 75 red, 227, 279–280, 285–286, 289, 291
gold, 30 rituals, 287–289, 291
hoards, 82 roe, 280, 285–286
Islamic, 28, 30, 62–63, 74–75, 166–167 defences, 30, 44, 52, 101–102, 147–148, 184, 358
Kufic, 229 see also hillforts; protection
Norwegian, 245 De Heul, 111
prehistoric, 55 Denmark, 199
sceattas, 163 Kingdom of, 69, 73–76, 81, 83–84, 99, 163
Cold War era, 41 disease, 180
Coll, 355–356 Dithmarschen, 19
Cologne, 114 DNA, 199, 202, 255, 281
Colonsay, 313 Dnieper River, 25–27, 47
comb cases, 265 documents, 337, 344
comb production, 162, 259–261, 268–269, 271, Dole, 50
290–291 dolphins, 280
combs, 42, 165, 225, 228, 239, 259–273, 289–290, 378 Domburg, 115
communication, 14, 54, 209, 320, see also interactions Domesday Book, 139, 142–143
Conishead, 343 domestic culture, 198–199, 203, 209–210, 228
conversion, 14, 44–46, 69–70, 76, 106, 229, 373 Don, River, 26, 147
Copeland, 315 Donald Dubh, 352
copper alloy, 242, 245–246, 268 Donald Gorm, 361
Cork, 205 Donald, Lord of the Isles, 360
cosmology, 20 Donets River, 26
Cottam, 139, 208 Dorestad, 2, 4–5, 108–116, 118–121, 200
Cottonian library, 344 brooch, 113
Couronia, 45, 50–51 decline, 111
craft working, 26–27, 50, 55, 141, 149, 160, 162, etymology, 120
164–167, 190–191, 194, 259, 280, 290 excavation, 111–113
creels, 115, 117 fishing, 115
cremations, 45, 50–51, 171, 178, 315 harbour, 110–111, 116, 118, 120
Cronk Moar, 311–312 house, 110
Crossreagh, 205, 212 mint, 119
Crovan dynasty, 336–340, 344–345 population, 114
Cruach Mhór, 211 trade, 114–115
crusades, 342 Douglas, 343
crystal, 66 Dovre, 195
Cubbie Roo’s Castle, 230 drengs, 76, 83
Cumbraes, 357 dress
Cumwhitton, 312 accessories, 64, 273, 378
Cunningham, John, 355–357 female, 66, 380
curraghs, 21 male, 27
currency, 5, 78, 163–164, 167, 229, see also bullion; Slavic, 82
ingots Driffield, 140
silver, 4, 31, 78, 163–164, 167–168 Drimore Machair, 201, 211
Dublin, 205–206, 212, 264, 266, 272, 299, 304–305,
Dagsberg, 105 308–311, 313–314, 336–339, 343, 370–371,
Dalkey Island, 377 374–375, 377–379
Damme, 127–128 currency, 4
dams, 127, 141 foundation, 374
Damsay, 230 identity, 375
Danelaw, 82–83, 206, 208, 304, 314–315 Kings of, 2, 6, 338, 370, 374, 376
Danes, 6, 209 military power, 6
Danevirke, 101 steatite, 205
MAM 37: 25-index – Press

index 387
Duinkerke (Dunkirk), 128 birds, 223, 227, 249, 280
Duirnish, 353, 362 cats, 227
Dún Aillinne, 285 cattle, 250–252, 283
Dun Ardtreck, 285 deer, 281, 283–284, 289, see also reindeer
Dun Bharavat, 288 dogs, 176, 227
Dun Mor Vaul, 265, 285–286 fish, 8, 115, 117, 227, 245, 251–253
Dunkirk see Duinkerke horses, 176, 227
Dunnyneill Island, 377 molluscs, 139, 227
Dun Vulan, 287 pigs, 227, 250–251, 283
Durham, 262 seal, 227
dykes, 100, 122, 126–127, 131–132 sheep, 250–252, 283, see also wool
walrus, 190, see also ivory
Eadgar, 241 whales, 114–115, 225, 245
Earl’s Bu, 265, 286 feasting, 78, 102, 209, 227, 251
Eccleston, 213 Fenlands, 123, 131
Edinburgh, 355, 361–362 Ferchar, Earl of Ross, 339
Edward I of England, 144, 363 Fergus of Galloway, 339, 342
Edward the Elder of Wessex, 207 ferries, 143
Eggja monument, 19 feudal, 83, 129–130
Egil Ragnarsen, 81 Finlaggan, 358
Egino, Archbishop of Lund, 69 Finland, 27, 41
Eigg, 360 Finno-Ugric, 27
Ekoln, Lake, 95 Fionnula, 342
Elbe River, 79 Firth, Bay of, 230
elites see also magnates fishing, 28, 42, 46, 74, 115, 122, 124–125, 128–130,
elk (moose), 279, 289 139, 167, 219, 227, 279, 288–289, 335
Emajõgi River, 47–48, 54 cod, 227, 288
emporia, 30, 108–109, 112, 138, 141, 147, 150, 200, eels, 126, 128–129, 227
see also harbours equipment, 116, 199, 225, 253
England see also Danelaw herring, 115, 129, 227, 253, 288
Anglo-Saxon, 19, 145–146, 306 sea, 148
burials, 312, 314–315 fish preservation, 74, 191, 195, 227, 253
combs, 261, 267, 269, 271, 290 Flanders, 7, 122–123
identity, 307 Counts of, 122, 125–130, 132
Kingdom of, 83 environment, 124
monuments, 315 fishing, 128–130
place names, 207 settlement, 122, 124
pottery production, 208 society, 125, 128, 130, 132
Scandinavian settlement, 207–208 trade, 123, 125, 128
society, 208 flax, 227, see also plant remains, flax
status, 311 flooding, 113–114, 123–124, 127–128, 131–132, 142,
erosion, coastal, 131, 220 see also water management
Estonia, 267 Foss, River, 146–147, 150
archaeology, 41 fragmentation, 78, 80
burials, 51 Frankish products, 114, 308
centres, 47, 53–54 Fraser, Jo, 355
harbours, 56 fraud, 159
hillforts, 49, 52 Freswick Links, 204, 211, 227, 262, 265, 267–268
society, 45–46 Frey, 180
trade, 55 Frisia
Ethelred II of England, 82 merchants of, 22, 96
Eyjólfr, 325–327 Fröjel, 29, 37
Eyrephort, 310–311, 374 fur, 26–27, 114, 279, 288
Furness Abbey, 335–336, 343
Falsgrave, 142, 149 Fyn, 75, 160
Falster, 85 Fyris, River, 95
Farewell, Cape, 329 Fysingen, Lake, 89
Faroe Islands, 199, 251
faunal remains Gaels, 130, 249, 375, 379
animal, 53, 77–78, 111, 249 gaming, 103, 105, 225, 378
MAM 37: 25-index – Press

388 maritime societies of the viking and medieval world

Gamleborg, 71, 74–75 Hamwic, 141
Garry Iochdrach, 211 hanfot system, 16–17
Gdansk, 266, 269 Hang Li-Po, 374
Gerald of Wales, 333, 374, 377 Hanseatic League, 47–48, 64
German–Danish conquest, 47 Harald Finehair of Norway, 374
Gersta, River, 126–127 Haraldr Maddaðarson, Earl of Orkney, 322–323, 354
Gevninge, 140 Haraldr Óláfsson of Man, 335–336, 339–341, 344
Ghent, 129 harbour infrastructure, 44, 139, 190
gift giving, 5, 164–165, 377 harbours, 28, 43, 114, 120, 139–141, 147, 149, 371,
glass, 83, 114, see also beads see also emporia
imports, 204 monastic, 147
Glenfalloch, 364 Harlaw, 360
globalization, 3–4 Harris see Lewis
glocalization, 4 Hartlepool, 146
Glynns, 352 Hatton, Sir Christopher, 344
Gnezdovo, 26–27, 48, 209 Haughey’s Fort, County Armagh, 285
Godfred of Denmark, 163 Heath Wood, Derbyshire, 315
Godred see Guðrøðr Hedeby (Haithabu), 2, 4–5, 22, 30, 62, 69, 81, 96,
Gokstad, 177 162–164, 196, 226, 260, 264, 266, 269
gold, 82, 105, 163 Hedon, 142–143, 148–149
Goltho, 208 heirlooms, 228
Götar, 99, 102–103, 106 Helgö, 160, 163
Götavirke, 101–102, 106 Henry I of England, 338
Gotland, 2–4, 6, 27–28, 35, 37, 44, 62, 64–65, 84 Henry II of England, 149
administration, 65–66 Henry III of England, 336, 338, 340–341, 364
culture, 66–67 Henry VIII of England, 343
hoards, 65 Henry the Livonian, 48
identity, 67 Heðinn stone, 352
settlement, 85 heraldry, 18
trade, 66, 74 Herjólfr, 329
Gotlandic picture stones, 16–18, 20, 22 Hermes, 29
Graham, Archibald, Bishop of the Isles, 355 Herøy, 172, 176, 178
grave goods, 35, 37, 44–45, 49–50, 73, 77–79, 82, 171, hierarchy, 45–46, 56, 158, 164, see also status
176, 178–179, 300–301, 304–311, 313–314 hillforts, 8, 44–45, 47, 49–56, 91, 95, 100
graves see burials Hirta, 201
Gravelines, 128 Historic Scotland, 220
Great Bernera, 201, 285 hoards, 28, 62, 64, 75, 80, 82, 84, 103–105, 163–164,
Great Army, 125, 207–208, 300, 314–315 167–168, 202–203, 222, 229
Greece, 36, 69 gold, 65
Greenland, 198, 328–329 bronze, 62
Grevelingen, 130 silver, 27, 62, 64–65, 69, 71, 74–75, 84, 105, 379
Grı́mr Hjaltason, 324, 326–329 Hofstaðir, 249
Grobiņa, 37, 41, 51 Hokkien, 375
Grødby River, 74 Holderness, 142
Grønhaug, 18 Holderness, Lords of, 142–144
Grovehill, 144 Holmboe, Jens, 173
Gudme, 73, 160 Holm Cultram, 343
Guðmundr Arason, 323, 325, 327–328, 335 Holmen, 184, 188, 190, 194
Guðmundr Svertingsson, 324 Holsegård, 82
Guðrøðr Crovan of Man, 336–344 Holwerda, J H, 112
Guðrøðr Don, son of Rögnvaldr, 338 Holy Land, 323, 342
Guðrøðr Óláfsson of Man, 339, 341, 343, 354, 357 hones, 186, 208–209, see also whetstones
Guldensgård, 82 Honorius IV, 342
Gulen, 178 Hooghen Steendam, 127
Gullberwick, 322 horn, 190, 267, 272
horse gear, 73, 103
Haithabu see Hedeby hospitality, 147, 194, 324
Håkon Håkonarsson of Norway, 195, 350, 354, 357 Hossmo, 96
Hákon Pálsson, Earl of Orkney, 339 Housebay, 232
halls, 230 housing, 223–224, see also Scandinavian, houses
MAM 37: 25-index – Press

index 389
Howar, 268 burials, 171, 174
Howar, Knap of, 281 central places, 73, 89, 93, 96
Howe, 285, 288 combs, 290
Hrafn Sveinbjarnarson, 323–325, 327, 335 deer, 285–287
Hrı́sheimar, 249 monuments, 174, 221, 244, 265
Hull, River, 140, 148 settlements, 74, 95
Humber, River, 140 society, 45–46
Humbleton Beck, 143 travel, 89, 93
hunting, 167, 286–287, 289, 291 irrigation, 130, see also water management
husa(r), 89, 93–96, 106 Iru, 47, 53–54, 56
Husum, 95–96 Islamic world, 2, 26, 31, 64, 372–376, see also coinage
hybrid culture, 4, 378–380 Islay, 211, 253, 311–313, 352, 356, 358, 363
Ranald of, 353
Iceland, 199, 251, 267, 320, 379 Isle of Wight, 140
Bishops of, 323 Isles, Bishop of the, 362–363
faunal remains, 251 Isles, Lordship of the, 7, 350
religion, 66 land division, 363–364
settlement, 249 military forces, 350–352, 355, 357–362, 365
society, 64, 66 Ivar, 374
identity, 2, 4–5, 228, 272, 307, 310, 369–370, 372, 374, Ivar’s Knowe, 267
376–377, 380 ivory, 290, 335, see also faunal remains
female, 27, 50, 78
hybrid, 369–370, 374 James IV of Scotland, 353
male, 307 James VI of Scotland, 355, 362
imports, 56, 74, 125, 138–139, 144, 165–166, 199, 204, Janssen, L J F, 111
281 Järfalla, 95
Inchicore, 311 Jarlshof, 226, 228, 264, 266–268, 286
Inchmarnock, 204, 211 Jaroslav, 27
Ingibjörg, daughter of Hákon Pálsson, 339 Jarrow, 147–148, 150
ingots, 63, 166–167, 205–206 Jersika, 44, 51
Ingvar, 105 jewellery see also beads
inheritance, 73, 76, 311–314, 316 amulets, 77–78, 82
Inishkea North, 212 bronze, 42, 64
Innocent III, 342 brooches, 27, 34–37, 66, 301, 378
interactions, 25, 29, 34–35, 73, 159, 164, 194–195, copper alloy, 165
203, 206, 209, 221, 261, 334, 373–374, 379–380, earrings, 78
see also trade female, 27, 64, 67, 78
Iona, 255, 343, 352, 355, 357, 362 necklaces, 77–78, 230
Abbey Museum, 352 oval brooches, 4, 27–28, 66, 73, 204, 211, 299–300,
Ireland 314
burials, 374, 376 pendants, 66
combs, 266, 290 pins, 64, 226, 228, 239, 241–242, 245–246, 270, 273,
deer, 281 309, 378
hoards, 379 rings, 63–64, 82
identity, 307, 377–380 slavic, 80
Kings of, 371, 377–379 Jews, 62, 370, 379
material culture, 308 Johannessen, Fr, 174
Mesolithic, 281 John I, Lord of the Isles, 360
National Museum of, 304 John II, Lord of the Isles, 353–354
pre-Viking, 377 John de Courcy, 336, 338, 341
Scandinavian influences, 205, 299, 308, 369–370, John of England, 336, 338, 341, 360
373, 377 John of Fordun, 356, 364
society, 371, 373, 377, 379–380 Jordanes, 89
status, 378
trade, 377 Kabusa, 96
weapons, 303 Kalmar, 67, 96
Irish Academy, Royal, 299 Karmøy, 7, 18
Irish Sea, 270, 272, 333 Kaupang, 3–4, 22, 162–164, 166–167, 176, 196, 264
Irish wars, 336, 358, 360–361, 371, 379–380 coins, 166
Iron Age, 14, 36, 46, 67, 171, 299 food production, 167
MAM 37: 25-index – Press

390 maritime societies of the viking and medieval world

Kaupang project, 157 Lilla Klintagård, 35
Kaupo, 45 Lilla Rytterne, 95
Keeill Vael, 212 Lilleborg, 71, 75
kelping, 222–223 Limerick, 205, 356, 371, 374
Kerrera, 357 Lincoln, 82, 206–207, 266
Khazars, 26, 63 Flaxengate, 206, 213
Kiev, 47, 209 Lindisfarne, 147–148, 150
Kilmainham-Islandbridge, 205, 212, 299, 304, 309–310 Linköping, Bishop of, 66
kilns, 179, 246–248, 312 literacy, 344
Kilwa Kisiwana, 374–375 Little Ice Age see climate change
kingship, 2, 5–6, 16, 22, 101, 190, 306, 337, 341, Liudger, 109
344–345, 379 Liutizi, 79
King’s Lynn, 149 livestock, 125, 176, 224, 227, 288–289, see also meat
Kingston-upon-Hull, 144 Livonia, 45, 47–48, 51
kinship, 25, 45–46, 76, 78, 158, 292, 310, 313–314, cemeteries, 50–51
316, 371, 375 Lochaber, 360
Kirkwall, 222 Lochawe, 352
Kjarrekstaðir, 230 Lofoten, 226
knarr, 119 Lofta, 89–90, 96
knighthood, 130, 340–341, 345 Lögmaðr, 342
knives, 73, 81, 307 Loire, 6, 255
Knowth, 266, 377–378 Lolland, 85
Koknes, 44, 51 London, 5, 82, 267, 356, 358
Kolmården, 99 longhouses, 203, 223–225, 249
Kromme Rijn River, 108–109, 113–115 lordship, 55, 125, 127, 130, 133, 199, 219, 229, 232,
Kuldiga, 51 307
Kungähalla, 266, 279 Lort Burn, 145
Kungsåra, 96 Lossit, 363
Kvalsund, 18–19, 173–180 Loudoun Hill, 360
Louis the Pious, 119
Ladoga see Staraja Ladoga Ludgo, 92
land ownership, 18, 22, 65, 73–74, 81, 83, 131–132, Lund, 79, 264, 266–267
191–192, 194, 310, 312–313, 316, 373 Bishops of, 69, 74
Langskaill, 267–268 Lundeborg, 160, 163
language, 4, 34, 46, 199, 320, 330, 373, 375 Lyminge, 140
L’Anse aux Meadows, 378 Lyndale, 353, 362
Larne, 311, 374
Læs River, 74 MacDonald of Dunyvaig, Angus 362
Latgalia, 44–45, 50, 56 MacDonald family, 350, 352, 361
Latvia MacDonnell family, 360
centres, 47, 49, 52, 55 machair, 221, 228, 232, 251, 253, 255, 286
trade, 56 Macillechoan, Hector More, 360
Laukskola, 50 MacLean, Hector, 353
laws, 16, 30–31, 36, 66, 141, 149, 159, 301–302, 354, MacLean of Duart, Lachlan, 355–358, 361–362
377 MacLean of Duart family, 352, 355, 357, 361
lead, 241 MacLeod of Dunvegan, Alexander, 353
leather, 62, 194 MacLeod of Lewis, Torquil, 353
Lecker Au log boat, 19 MacLeod of Dunvegan family, 362
ledung, 18, 96, 100, 354, 363, 365, see also naval MacRuairi of Garmoran, Lachlan, 364
organization MacRuari of Garmoran, Christina, 360
Leffinge, 124 MacSorley family, 337
leidang see ledung MacSween, John, 330
Leiden, 109, 111–112 MacSweeney family, 360
Lejre, 73, 140 magnates, 8, 27, 30, 54, 160, 167, 184–186, 188, 191,
Lejre Å, River, 140 260
Lek River, 108, 113, 115, 120 Magnús Berfœttr of Norway, 357
Leland, John, 333 Magnús Óláfsson of Man, 336, 338, 340
Lewis, 255, 265, 270, 285–286, 335, 353, 356–357 Mainland, 232, 281, 285
Lewis Chessmen, 335 Mälaren, Lake, 89, 91, 95, 99, 102, 160
Lielupe River, 51 Malaysia, 3, 373–376
MAM 37: 25-index – Press

index 391
Malcolm IV of Scotland, 340 Møllebjerg, 74
Man, Isle of, 2, 7, 333–335 Møllegård, 73
burials, 312 monasteries, 31, 146–148, 150, 192, 204
currency, 4, 336 monastic
foreign relations, 337–338 ideology, 148
Kings of, 6, 334–345, 350, 357 reform, 340
land divisions, 363 money, 163–164, see also coinage
landscape, 335 Monkwearmouth, 146
monuments, 315 Monro, Dean, 355–357
religion, 334, 341–344 Moose see elk
Scandinavian influence, 334, 341 Moray, Earl of, 360
settlement, 336 Mörkö, 96
Mansur Shah of Melaka, Sultan, 374 Morragh, 374
Manx National Heritage, 212 Mrimba, 374
Mar, Earl of, 360 Mulgrave, 149
Mardyck, 129–130 Mull, 352–353, 355–356
markets, 26, 28–29, 80, 114–115, 129, 164–165, 260, Munkegård, 70, 76–78, 82
see also emporia Munkeliv abbey, 188
beach, 139, 141 Myroscough, 343
central-place, 54 mythology, 228, 374–375
economic, 156–159, 371 Myvatnssveit, 249
marriage, 25, 102, 340, 342, 370, 376
Marwick, 220 Narva River, 47
Marx, Karl, 156 naval organization, 96, 100, 354, 362–365, see also
Mary, daughter of Rögnvaldr Óláfsson, 339 navies
mass production, 165, 167 Navesti River, 47
Matthew Paris, 340 navies, 3, 6, 16, 22, 310, 337, 352, 354–355, 357,
Mattwse, 96 see also naval organization
Maughold, 345, 352 navigation, 47, 116, 141–143, 148, 150, 324
meat, 122, 194, 227, 251, 280–281, 287, 289 Ndr. Grødbygård, 70, 76–78, 82
Meaux, 144, 149 needles, 225
Medemblik, 115 Nerlandsøya, 174
Medieval Archaeology, Society for, 210, 318 Newburn, 145
Medieval Warm Period see climate change Newcastle-upon-Tyne, 142, 145
Medina, River, 140 St Andrew’s Church, 145
Melrose, 344 St Nicholas’ Church, 149
memory, 76, 82, 311, 322–323, 376 Newgrange, 281
mercenaries, 5–8, 360 Newport, 140
merchants, 29, 35–36, 46, 114, 186, 190, 195, 207, Nidaros, 195
369–370, 375 Nieuwpoort, 128, 130
Mercia, 207 Noirmoutier, 255
Mesolithic, 280–281 Noltland, Links of, 287
metal detecting, 69, 82 Nonneseter Abbey, 192, 194
metalwork, 3, 103, 194, 199, 225 Nordians hög, 89
ecclesiastical, 300, 380 Norman conquest, 8, 340
metalworking, 64, 162, 190, 222, 226–227, 232, 307 Normandy, 225
Meuse-region, 114 Norrsjö, 89–90
Mežotne, 51–52 Norrsunda, 89
middens, 8, 221, 223, 227, 232, 245–246, 248, 250, 252 Norsa, 91, 96
Middleton Cross, 315 Northampton, 207
Midhowe, 281 Northdyke, 231
migration, 3, 27, 200, 203, 208–210, 254, 279 North Elmham, 208
Mikhailovskoje, 27 North Ronaldsay, 268
military, 27, 42, 88, 102, 126, 306–307, 310, 314, 340, Northton, 281, 285–286
363, see also weapons Northumbria, 138–139, 142, 145–146, 148, 150
milk, 288–289 Earls of, 145
mills, 74, 114, 199 monasteries, 146, 149
mines, 335 royal centres, 145
Minginish, 353, 362 settlement, 148, 150
mints, 4, 8, 149 urbanism, 142
MAM 37: 25-index – Press

392 maritime societies of the viking and medieval world

Norway, 209 Papacy, 336, 342
burials, 171, 300, 311 Pärnu River, 47–48, 54
civil wars, 192 patronage, 76, 83, 113, 128, 140, 142, 191, 377
culture, 228 ecclesiastical, 142, 144, 148, 150, 335–336, 343–345
currency, 167 Paull, 143
Kings of, 350 Paviken, 29, 37
society, 301 peat, 129
towns, 182, 186 Peel Castle, 335
trade, 308 Peipsi, Lake, 47
Norwich, 206–208, 213 Pentland Skerries, 268
Nottingham, 314 Peranakan culture, 372–373, 375–376, 379
Novgorod, 2, 26, 47, 55, 265, 267 Persia, 374–375
Novum Dam, 127 Perth, Treaty of, 272, 350, 357, 367
Petrie, George, 222, 308
oars, 15–16, 18, 172, 175, 179, 352–355, 362–364, Petrovskoje, 27
see also rowing Pierowall, 281, 313
Oban, 357 pigs, 227, 250–251, 255
Obodriti, 81 Pihkva, Lake, 47
Oder estuary, 79–80, 85 pilgrimage, 342
O’ Donnell, Sir Hugh, 361 Pippin, 31
ogham, 242, 245 piracy, 6–8, 41–42, 45, 75, 81, 84, 120, 373, see also
Olaf Cuarán of Dublin and York, 374 raiding
Óláfr Guðrøðarson of Man, 335–336, 338–343, 345 Pirita River, 53
Öland, 67, 74, 84 place names, 88–89, 91–92, 95–97, 199, 219
Olav Haraldsson of Norway, 184 Plakun, 25
Olav Kyrre of Norway, 182, 187, 245 plant remains, 246, 253
Olav Magnusson of Norway, 188 barley, 226, 253–254
Olenok Evenk, 289 cannabis, 180
Olof Eriksson of Sweden, 99, 106 flax, 253–254
Oostburg, 125 oats, 226, 253–254
Oppusa, 92–93 rye, 254
Ordnance Survey, 222 weeds, 226–227
Orkney College Geophysics Unit, 220–221 poetry, 320, 322–323, 326–327, 329–330, 339, 342
Orkney, 2, 200, 202, 219, 231, 264–266, 329 Pöide, 52–53
agriculture, 227 Polotsk, 44–45, 47
deer, 285, 287 Pomerania, 81
Earls of, 2, 5, 7, 190, 219, 286, 289, 329–330 Portable Antiquities Scheme, 210, 213
identity, 232 Portland, 3
land organization, 219, 231 Portmahomack, 204
poetry, 320 pottery, 3, 53, 77, 79, 110, 203, 208, 256
prehistory, 221, 223, 281, 287 African Red Slip, 204
Scandinavian presence, 232, 273 amphorae, 204
trade, 5, 228 Badorf, 199
Orkney Islands Council, 220 Baltic, 74, 77, 79–81, 84–85
Oronsay, 281, 313, 361 continental, 138
Orphir, 265 decorated, 288
Orry, King, 336 deposition of, 78
Oseberg, 18, 173–174, 176–180 imported, 42, 125, 376
Osilians, 45–46 mayen, 125
Oslo, 195, 264, 267, 279 Northern Irish, 225, 255–256
Ospisdale, 204, 211 Pingsdorf, 199
Ostend, 131 production of, 79, 208
Otto I, 110 redware, 225, 376
Ouse, River, 146 Rhenish, 125
Oxford University, 220 Scandinavian, 203
oysters, 139 Scottish, 255–256
Øystein Magnusson of Norway, 188, 190–191, 195 tableware, 208, 228
Tatinger, 114
Pada, 53 whiteware, 225
Padjelanta, 21 power centres, 43, 99, 106, 188, 336, 377
MAM 37: 25-index – Press

index 393
power, 5, 88, 126, 132–133, 158–159, 306, 311, 314, Rögnvaldr Guðrøðarson of Man, 335–336, 338–339,
316 341–345
consolidation of, 2, 5, 30, 147, 308, 370 Rögnvaldr Kali, Earl of Orkney, 190, 321–323,
decentralized, 73 330–331
political, 42, 44, 46, 48, 370, 377 Rögnvaldr Óláfsson of Man, 336
royal, 14, 16, 74–75, 83, 182, 184–186, 191, Rollock, Thomas, 363
194–195, 260 Roma, 66
Præste River, 74 Roman world, 15, 113, 120, 127, 146–148, 150
Premonstratensian Order, 343 Rönö, 92
prestige goods, 25, 44, 55, 114, 200, 209, 251, 272, Röntuna, 92, 96
301, 308, 371, 377–378 Rosersberg, 91
Procopius, 19 Roskilde Fjord, 354
protection, 30, 43, 109, 124, 130, 147–148, 186, 377, Rosnaree, 371
see also defences Ross
Provinciaal Utrechts Genootschap, 111 Bishop of, 364
Pskov, 47 Earl of, 360, 364
Rousay, 230, 281
Quanterness, 281, 288 rowing, 15–16, 19, 354, see also oars
quernstones, 74, 118, 138–139, 186, 226 rudders, 352–353
Quoygrew, 199, 227, 264–268, 286 Rügen, 85
Rum, 360
raiding, 5, 81–83, 102, 105, 109, 237, 249, 255, Runegård, 70, 76–77
299–300, 341–342, 354 runestones, 15, 19, 76, 83–84, 95, 105
Ramsay, Knowe of, 281 Runnviken, 92
Ranald of Islay, 353 Runsa, 91, 96
raw materials, 21, 25, 64, 125, 162, 165–166, 280, 290 Runtuna, 92
Reading, 300, 314 Rushen Abbey, 336–337, 343
Reay, 311 Rushen Castle, 336
reclamation, 123–124, 126–127, 142, 145, 149, see also Russia, 26–27, 36, 47–48, 200, 209, 267
water management
reindeer, 279, 288–289 Saaremaa, 45–46, 50–52, 55
relics, 192, 194, 380 Sabal, 343
religion, 66, 139, 141, 180, 315–316, 373, see also sacrifice, 104, 164, 176, 179–180
Christianity sailing, 7, 16, 19, 21–22, 320–324, 326, 329–330,
pre Christian, 4, 22, 66, 78, 88–89, 92, 109, 171, see also ships
180, 310, 334, 342 sails, 8, 15–16, 18, 21–22
religious centres, 55, 149 Saint Nicholas, 29
Repton, 300, 311, 314–315 Saint-Omer, 129
resources salt, 124–125, 129–130, 253
marine, 139, 144, 227, 232 salt production, 129–130
wild, 123, 126, 177, 279, 287 Sami, 21, 35, 288–289, 291
Reval, 47, see also Tallinn Sanday, 227, 265–266, 268, 285–287, 324, 326–327,
Rhine, 21, 108, 111, 113, 115, 119 335
Rhineland, 114, 138 Sandey, 324, 326–327
Ribblehead, 208 Sandhill, 149
Ribe, 5, 22, 162–165, 260, 264 Sandtun, 140
Rievaulx, 343 Sandwick North, 267–268
Riga, 8, 47, 50–51, 56 Savignac Order, 336, 343, 345
Rijksdienst voor het Cultureel Erfgoed, 112 scales, 31–33, 36–37, see also weights
Rijksdienst voor het Oudheidkundig Scalloway, 265, 285
Bodemonderzoek, 112 Scandinavian
Ringkloster, 288 burials, 310, 312, 314–315
rituals, 103, 172–173, 176–177, 179–180, 310, combs, 271
376–377, see also religion crafts, 161–162
Rjurikovo Gorodišče, 26 culture, 370, 378
Robert I of Scotland, 352, 358, 364 currency, 163
Robert’s Haven, 227 diaspora, 4, 330
rock art, 21, 179, 378 farmers, 255
Rødbjerg, 73 houses, 239, 255
Ro˛gnvaldr Brúsason, Earl of Orkney, 2, 329 identity, 209–210, 320, 330, 370, 380
MAM 37: 25-index – Press

394 maritime societies of the viking and medieval world

language, 330, 373 Skuldelev, 118
migration, 199, 209–210, 219, 237, 239, 288, 299, symbolism, 14–15, 20, 22, 172, 179
303 shipwrecks, 20, 118, 322–323, 339
power, 3, 330 Shirazi, 374
settlement, 198, 219 Sibbald, Sir Robert, 355
status, 219, 301–302, 304, 311 Sigfred of Denmark, 163
trade, 74, 141, 161–162, 165, 308, 377, 379 Sigtuna, 264, 266–267
Scarborough, 142, 149 Sigurd Magnusson of Norway, 188, 191
Scarphout, 131 Sigurd the Stout, Earl of Orkney, 232
Scarwell, 231 silk, 114, 138
Scatness, 285 silver, 30, 64, 78, 85, 167
Scheldt River, 108, 115 hack, 164, 166–167
Schleswig, 95, 264, 267 Simy Folds, 208
Schwesing, 95 Singapore, 379
Scotland see also Isles, Lordship of; pottery, Scottish Sint-Winoksbergen, 125
combs, 260–261, 266–267, 269, 271–272, 290 Sitric, 374
deer, 281, 285–287 Sjösa, 92–93, 96
burials, 202–203 Skabersjö, 96
hoards, 202–203 Skagerrak, 266
identity, 307 Skaill, Bay of, 220–222, 227, 229, 231–232, 264, 267,
Kings of, 231, 338, 350 285
monuments, 315 Hoard, 222, 228–229
National Library of, 355 etymology, 230
National Museum of, 222, 353 Skaill, Deerness, 227, 232, 264, 286
place names, 203, 255 Skamby, 103–105
pre-Viking, 3, 219, 239, 241, 255, 285 Skara Brae, 221, 229
Skäsa, 96
Royal Commission on the Ancient and Historical
Skelton, 149
Monuments of, 210–211
Skepphusa, 96
sailing, 323
Skerne, 140
Scandinavian arrival, 199, 203, 219, 237, 239, 249,
Skiðblaðnir, 18
254–255, 285–286, 291, 369
Skye, 285, 353, 355–357, 362
Scandinavian culture, 202–203, 219, 272, 289, 312
Slamrebjerg, 76–77
sculpture, 352, 360–361
Slätbaken, 100–101
settlements, 227, 261 slavery, 6, 66, 79, 81, 114, 255
Wars of Independence, 350 Slavic
Scottish Parliament, 364 migration, 78, 81
Scottish Privy Council, 352, 354–355 people, 27, 78, 165, 209
sea, 320, 373 traditions, 209
conditions, 139–140, 142, 146, 174, 310, 322 sledges, 48, 116, 176
perceptions of, 8, 148, 150, 328 Sleipnir, 18
sea-level change, 41–42, 44, 124 Smørrenge, 73
seal-hunting, 42, 227, 280 Snaefell, 333
seals, 36, 42, 227, 280, 341, 344–345, 352 Snellegem, 126
seasonality, 52, 160, 162, 293, 295 Snusgar, 222–225, 227, 229–232, 266
Seine, 6 soapstone see steatite
Selje, 172, 178 society
Semigalia, 44–45, 50–51, 56 egalitarian, 45, 56, 64–65, 73, 125, 301
sheep, 21, 122, 124, 126, 227, 251–252, 335 interactions, 27, 369–370, 375
shields, 36, 300–301, 305, 307, 309–312 stratified, 44, 64, 66, see also elites; status
shellfish, 114–115, see also faunal remains, mollusc Södersjö, 89–90
Shetelig, Håkon, 173 Sognefjord, 301
Shetland, 198, 200, 202, 231, 264, 266–267, 286 Somerled, 337–338, 340, 350, 352, 354, 357–358
deer, 281, 285–286 Sonning, 310
ships see also Oseberg; Kvalsund; burials, boat; sails; Soontagana, 52
sailing Sorte Muld, 73, 160
cargo, 8, 44 Soviet archaeology, 41, 49
construction of, 19, 21, 53, 118, 352, 354 Soviet Union, 41–42
depictions of, 15, 17–18, 139, 352–353 Sparlösa runestone, 18, 20–21
equipment, 116, 118, 172–173, see also rudders spices, 114
MAM 37: 25-index – Press

index 395
Spillings, 62, 64 Tänassilma River, 47
spindle whorls, 199, 205, 225 Tartu, 48, 53–56, 267
Stafford, 207, 268 Tattershall Thorpe, 307
Stamford Bridge, battle of, 8, 336 taxes, 6, 18, 43, 51, 55, 74, 81, 115, 130, 142, 186, 194,
standing stones, 103, see also runestones 231, 324, 336
Staraja Ladoga, 25, 47–48, 209 Tērvete, 51
status, 203, 228, 251, 365, 372, 377, 379, see also Texa, 353
hierarchy textiles, 62, 114, 176
St Bees Priory, 343 Thatcham, 288
St Boniface, 227 thegns, 76, 82–83
St Brice’s Day Massacre, 6 Theophilus, 62
steatite, 186, 198–200, 202–213, 225, 228, 245 Thetford, 206–207, 213
beads, 241–242 Thingvellir, 66
sources, 198–199, 210 Thor, 205
symbolism, 209–210 three age system, 299
Stewart dynasty, 350 Tiel, 118
Steyning, 207 Timerëvo, 27
St Kilda, 201, 211 Tingvoll, 172
St Marnock’s, Inchmarnock, 204, 211 Tiree, 253, 265, 285–286
St Mary’s, 343 Tissø, 73, 160
St Nicholas, 190 tithes, 66, 191
Stockholm airport, 89 Tiveden, 99
Stockholm Museum, 112 Tjust, 89–90
St Olaf see Olav Haraldsson Tofts Ness, 267, 285
Stolpe, Hjalmar, 103, 111 Tomar, ring of, 376
Store Frigård, 82 Tønsberg, 267
Storhaug, 18, 178–179
tools, 77, 198, 246
Stour Head, Sutherland, 324
Þorfinnr, Earl of Orkney, 329, 354
Stove, 232
Torksey, 2
St Patrick’s Island, 336
Torres Strait Islanders, 374
St Peter’s Abbey, Flanders, 131
Torsberga, 92
Stromness, 230
Tournai, 129
St Sunniva, 192, 194
towns, 5, 8, 43–44, 47, 49, 141, 145–146, 150,
St Werburgh, 343
Styrstad, 105 162–163, 166–168, 186, see also urban centres
Sundbysjön, Lake, 92 trade diasporas, 3, 369–370, 377, 380
Sunnersta, 95–96 trade, 8, 29, 33–36, 42–43, 46–47, 49–50, 52, 56,
Sutton Hoo, 18 73–74, 84, 141, 159–160, 168, 190, 370, see also
Svear, 99, 102–103, 106 interactions
Sven Forkbeard of Denmark, 76, 83 centres, 43–44, 46–49, 51–56, 139, 142, 150, 369,
Sverrir of Norway, 327 371
Svitjod, 96 equipment, 33, 37, 162, see also weights, scales
Swahili, 372–375, 377, 379 long distance, 29, 160, 186, 379
Sweden networks, 160, 165, 370, 377
boat burials, 102–103 relationships, 25, 29–31, 33, 159
central places, 88 routes, 43–44, 46–48, 51–52, 55–56, 70, 128, 199,
culture, 67 373
hoards, 102–103 Traigh Bostadh see Bostadh
Kingdom, 99, 106 transport, 4–5, 14, 48, 56, 72, 141, 148, 198, 289, 374
power, 101–102, 106 tribute, 25–27, 45, 84
settlements, 95, 101 Trondheim, 186, 190, 192, 262, 264, 266–267, 279
trade, 164 Trotternish, 353, 362
Sween, Castle, 330 Truska, 378
Szczecin Lagoon, 80, 85 Truso, 69
Tune, 177
Tåby, 105 Tuquoy, 230
Talacre, 310 Tyne, River, 140, 142, 146–147
Tallavbaun, 212 Tynemouth, 146
Tallinn, 47, 53–54, 56, 267 Tyrone, Earl of, 355, 361–362
Talsi, 51 Tyskegård, 82
MAM 37: 25-index – Press

396 maritime societies of the viking and medieval world

Udal, 202, 226, 245, 264, 285–286 Waterford, 205, 266, 272, 371, 374, 379
Uı́ Néill dynasty, 370–371 water management, 122–124, 126–128, 130–133,
Uist, 201–202, 226, 228, 232, 237, 253, 265, 285–288, see also flooding
290, 357, 360, 364 weapons, 45–46, 114, 228, 299–316
Ullr, 95 arrowheads, 300–301, 305
Ulster Museum, 205, 212 axes, 300–303, 305–306, 308, 311
Ultuna, 95–96 spearheads, 300–303, 305–307, 309–312
Unst, 267 swords, 300–309, 311–312, 360, 378
Uppåkra, 73, 160 weights, 31–36, 162
Uppsa kulle, 92 Wessex, Kingdom of, 7, 36, 141
Uppsala, 93, 95, 160 Westness, Rousay, 230, 305, 310, 312, 315
urban centres, 51, 129, 186, 207–208, see also towns Westray, 227, 230, 265–268, 286
urbanism, 2, 7–8, 22, 44, 51–52, 55–56, 124, 129, 133, wetlands, 48, 52, 104, 108, 140, 142, see also
146, 149, 168, 195, 371 reclamation
Utrecht, 113, 115, 118 Wexford, 205
Weymouth Ridgeway, Dorset, 6
Vada, 93–96 whales, 280
Vågen, 183–184, 187–188, 190, 192 Wharram Percy, 208
Valsgärde, 103 wheelhouses, 244
value, 158, 164–165, 168, 228, see also currency whetstones, 74, 138, 226, 228, 246, see also hones
van Oudheden, Museum, 109, 111–112 Whitby, 146
Varbola, 52 Whithorn, 204, 211, 266, 343
Vättern, Lake, 99, 105 Wijk bij Duurstede, 110–112
Vechten, 115 William of Aumale, 142
Veisan, 183–184, 188, 190, 192, 196 William I of Scotland, 340
Veleti, 81 Winchester, 82, 141
Velikaja River, 47 wine, 114, 118, 143–144
Vendel, 103 Wissant, 130
Verkaån River, 89, 91 Wolin, 80, 266, 269
Verron, 221 wood, 111, 118, 125
Veurne, 125–126, 128–129 Woodstown, Waterford, 310–311
Vikbolandet, 99–100, 103–106 wool, 7, 21, 116, 122, 124–126, 143–144
Vikhus, 95–96 Wrath, Cape, 324, 326, 329
Viljandi, 48 Wulfstan, 69
Visby, 44, 266 Wulpen, 131
Vistula River, 51 Wyke-upon-Hull, 142–144, 149
Vivian, 342 Wyre, 230
Vivilsjö, 89
Vlaardingen, 118, 127 Yarso, Knowe of, 281
Volga River, 25–27, 47, 63 York, 4, 139, 142, 144, 146–147, 150, 206–208,
Vorgebirge, 114 212–213, 262, 264, 266, 268
Archbishops of, 144, 345
Walacria, 115 Coppergate, 139, 206, 212, 246
Walden, F. H., 222 Fishergate, 139, 153
Wales, 210, 212, 340 Ypres, 128
Walker Beck, 144 Yzer, River, 126–127, 130
Walraversyde, 131–132
warriors, 5, 27, 315, 355, 358, 360, see also military; Zealand, 75, 85
weapons Zuyderzee, 115
Wars, World, 156 Zwin, River, 126–127