OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

OXFORD STUDIES ON THE ROMAN ECONOMY

General Editors
ALAN BOWMAN ANDREW WILSON
OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

O X F O R D S T U D I E S O N T H E R OM A N EC O N O M Y

This innovative monograph series reflects a vigorous revival of interest

in the ancient economy, focusing on the Mediterranean world under
Roman rule (c.100 BC to AD 350). Carefully quantified archaeological and
documentary data are integrated to help ancient historians, economic
historians, and archaeologists think about economic behaviour collectively
rather than from separate perspectives. The volumes include a substantial
comparative element and thus will be of interest to historians of other
periods and places.
 OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

Urban Craftsmen
and Traders in the
Roman World
Edited by
ANDREW WILSON
and
MIKO FLOHR

1
OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Oxford University Press 2016
The moral rights of the author have been asserted
First Edition published in 2016
Impression: 1
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, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence, or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2015949655
ISBN 978–0–19–874848–9
Printed Great Britain by
Clays Ltd, St Ives plc
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
 OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

Preface

The chapters in this volume have their origin in a workshop organized by the
Oxford Roman Economy Project and held at Wolfson College, Oxford, United
Kingdom, 21–23 July 2011. The workshop, entitled Beyond Marginality:
Craftsmen, Traders and the Socio-Economic History of Roman Urban Commu-
nities, was funded by the European Science Foundation as an ESF exploratory
workshop, and brought together researchers from nine different countries with
the aim of exploring the topic of Roman urban crafts and trades and to build
bridges between different national scholarly traditions and disciplines (ancient
history, archaeology, epigraphy, papyrology). Like the workshop from which it
originates, this volume focuses on four key themes: the history of research,
economic strategies of craftsmen and traders, the position of crafts and trade in
urban space, and craftsmen and traders in their social environment.
We thank the ESF for the exploratory workshop grant that made the
workshop possible, all the speakers who took part and subsequently contrib-
uted papers, and also Prof. Kristin Kuutma (Helsinki), who acted as the ESF
observer and rapporteur. We are grateful to the staff of Wolfson College for
their practical assistance during the workshop, and also to All Souls College
for the provision of accommodation for some speakers. As ever, we are
extremely grateful to Baron Lorne Thyssen for his continued support of the
Oxford Roman Economy Project.
Andrew Wilson
Miko Flohr
August 2015
OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi
 OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

Contents

List of Contributors ix
List of Figures xiii
List of Tables xvii
List of Abbreviations xviii

Introduction 1
Miko Flohr and Andrew Wilson

I APPROACHES
1. Roman Craftsmen and Traders: Towards an Intellectual History 23
Miko Flohr and Andrew Wilson
2. Twentieth-Century Italian Scholarship on Roman
Craftsmen, Traders, and their Professional Organizations 55
Carla Salvaterra and Alessandro Cristofori
3. The Archaeology of Roman Urban Workshops: A French Approach? 77
Jean-Pierre Brun

II STRATEGIES
4. Mercantile Specialization and Trading Communities: Economic
Strategies in Roman Maritime Trade 97
Candace Rice
5. Driving Forces for Specialization: Market, Location Factors,
Productivity Improvements 115
Kai Ruffing
6. Fashionable Footwear: Craftsmen and Consumers in the
North-West Provinces of the Roman Empire 132
Carol van Driel-Murray
7. Contextualizing the Operational Sequence: Pompeian Bakeries
as a Case Study 153
Nicolas Monteix

III PEOPLE
8. Disciplina, patrocinium, nomen: The Benefits of Apprenticeship
in the Roman World 183
Christel Freu
OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

viii Contents

9. Women, Trade, and Production in Urban Centres

of Roman Italy 200
Lena Larsson Lovén
10. Freedmen and Agency in Roman Business 222
Wim Broekaert
11. The Social Organization of Commerce and Crafts in Ancient
Arles: Heterogeneity, Hierarchy, and Patronage 254
Nicolas Tran
12. Hierapolis and its Professional Associations: A Comparative
Analysis 278
Ilias Arnaoutoglou

IV SPACE
13. Working Together: Clusters of Artisans in the Roman City 301
Penelope Goodman
14. Spatial Concentration and Dispersal of Roman Textile Crafts 334
Kerstin Droß-Krüpe
15. Industry and Commerce in the City of Aquincum 352
Orsolya Láng
16. The Potters of Ancient Sagalassos Revisited 377
Jeroen Poblome

Index 405
 OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

List of Contributors

Ilias N. Arnaoutoglou graduated at the Faculty of Law, Aristotelian University

of Thessaloniki. He holds a Ph.D. in ancient Greek history from the University
of Glasgow (1993). He was assistant editor for the Lexicon of Greek Personal
Names, Oxford, from 1994 to 1999. Currently, he is senior researcher in the
Research Centre for the History of Greek Law, Academy of Athens. He has
written extensively on Hellenistic and Greco-Roman associations, ancient
Greek legal history and institutions, and institutions in pre-nineteenth-century
Turkish-occupied Greece.
Wim Broekaert is a postdoctoral researcher in Ancient History at Ghent
University. He has published widely on many aspects of Roman trade. His
first monograph, Navicularii et negotiantes: A Prosopographical Study of
Roman Merchants and Shippers, was published in 2013.
Jean-Pierre Brun is a professor at the Collège de France in Paris, and, until
2012, director of the Centre Jean Bérard at Naples. He is a renowned specialist
in the archaeology of Roman agriculture and crafts, and has published widely
on olive-oil and wine production in the Greco-Roman world, and on urban
crafts in Roman Italy.
Alessandro Cristofori is Professor of Ancient History at the University of
Bologna, formerly Professor of Roman History at the University of Calabria.
His research has focused both on the epigraphy of crafts in Roman Italy and
on the historiography of craftsmen and traders in twentieth-century Italy. In
2004, he published a monograph on the epigraphy of crafts in Picenum (Non
Arma Virumque: le occupazioni nell’epigrafia del Piceno).
Carol van Driel-Murray (Leiden University) is Assistant Professor of Provin-
cial Roman Archaeology at the University of Leiden. Her research focuses on
Roman and medieval leathercraft, Roman military equipment, and gender
issues in relation to the Roman army. She has published extensively on Roman
leatherwork and footwear from archaeological excavations, skin-processing
technologies, and the role of footwear in ritual complexes. Reconstruction and
experiment figure prominently in her research into the structure and use of
military equipment such as tents and saddles.
Kerstin Droß-Krüpe studied classical archaeology, ancient history, and
business administration at Philipps-University of Marburg, Germany. Her
thesis, supervised by Kai Ruffing and Hans-Joachim Drexhage at Marburg,
OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

x List of Contributors

focused on textile production in Roman Egypt. It was published as

Wolle–Weber–Wirtschaft (2011). During her research she participated actively
in the EU research project ‘DressID’, where she presented several invited
papers. During recent years she has published several essays about the use of
Geographic Information Systems (GIS) in historical context. Currently, she is
working as scientific assistant at Kassel University.
Miko Flohr is postdoctoral researcher and lecturer at the Institute for History
of Leiden University, formerly assistant director of the Oxford Roman Econ-
omy Project. His main research focus lies with urban history in the Roman
world, with a particular emphasis on economic life in Roman Italy, and on
textile economies. His first monograph, The World of the Fullo, was published
in 2013; he has since published on the textile economy of Pompeii, and on
public investment in commercial space.
Christel Freu is Professor of Roman History in the University Laval in Quebec
and member of Année épigraphique. After publishing a monograph on repre-
sentations of the poor in late antique texts (Les Figures du pauvre chez les
auteurs italiens de l’antiquité tardive (2007)), she is currently studying labour
relationships in the Roman Empire on the basis of papyri, and literary and
epigraphic texts.
Penelope Goodman is a lecturer in Roman history at the University of Leeds
with a particular interest in the spatial characteristics of Roman urbanism. Her
first monograph, The Roman City and its Periphery: From Rome to Gaul
(2007), explored the demarcation of Roman urban centres and the uses of
space just beyond their boundaries. She has also published articles on the
locations of temples in Roman Gaul and Britain, the peripheries of Italian
cities, the boundaries of Rome, and the spatial distribution of terra sigillata
workshops. Her current research is concerned with receptions of the emperor
Augustus.
Orsolya Láng is deputy head of the department of Roman Archaeology at the
Aquincum Museum in Budapest, Hungary. She has worked extensively on the
civilian settlement of Aquincum, both in the field, and in the archives. Her
thesis and her published work focus on the history and use of urban space in
Aquincum, particularly in the north-eastern quarter of the town.
Lena Larsson Lovén is lecturer in Classical Archaeology and Ancient History
at the Department of Historical Studies, University of Gothenburg. Her main
research focus lies with gender studies, iconography, and socio-economic
history in the Roman world, with a particular emphasis on the Roman west.
She has edited several volumes and has published on aspects of Roman textile
production, dress studies, family iconography and economy, and women’s
work identities.
 OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

List of Contributors xi

Nicolas Monteix holds a position as associate professor in Roman history and

archeology at the University of Rouen. He wrote a dissertation on shops and
workshops in Herculaneum (published in 2010) and was then awarded a
postdoctoral fellowship at the École française de Rome (2007–10). He led the
Pistrina project that studied of thirty-nine bakeries in Pompeii (2008–14); his
current interest lies with the archaeology of technology in the long term.
Jeroen Poblome holds a chair in Archaeology at the University of Leuven,
Belgium. As Francqui Research Professor he coordinates research projects on
aspects of the ancient economy related to the production, distribution, and
consumption of artisanal goods, the associated study of ancient society,
focusing on the development, sustainability, and resilience of communities
and regions, in the period between 1,000 BC and AD 1,000, as well as innovative
research on digital archaeological information systems. He is involved in
fieldwork in Boeotia, Kinet Höyük, and at ancient Sagalassos. He co-founded
and co-edits HEROM: Journal on Hellenistic and Roman Material Culture.
Candace Rice is Lecturer in Classical Archaeology at the University of Edin-
burgh. She completed a D.Phil. at the University of Oxford (2013) and was a
Senior Fellow at the Research Center for Anatolian Civilizations in Istanbul. Her
research focuses on regional economic development in the Roman world and the
impact of maritime connectivity as evidenced by the archaeological record. She is
currently preparing a monograph on the economic landscape of the southern
coast of Turkey in the Roman period; additional forthcoming and published
work includes articles on Roman trade, connectivity, and maritime
infrastructure.
Kai Ruffing holds the chair of Ancient History at Kassel University. He has
published extensively on many aspects of Roman economic history, particu-
larly crafts and trade. His first monograph (1999) focused on viticulture in
Roman Egypt. His second monograph (2008) analysed specialization in crafts
and trade in the Roman east, mainly based on epigraphic and papyrological
evidence.
Carla Salvaterra is senior researcher and Vice-Rector at the University of
Bologna. She initially specialized in Egypt, and wrote a dissertation on the city
of Alexandria in the Flavian Period, but has also published on the history of
the study of labour in the Roman world.
Nicolas Tran is Professor of Roman History at the University of Poitiers and
at the Institut Universitaire de France. His research focuses on craftsmen and
traders in the western part of the Roman empire. He has published mono-
graphs on Roman professional associations (2006) and on the professional
lives of craftsmen and traders in the Roman west (2013), besides co-editing a
OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

xii List of Contributors

volume on Roman collegia (with Monique Dondin-Payre, 2012) and on the

professional knowledge of Roman artisans (with Nicolas Monteix, 2011).
Andrew Wilson is Professor of the Archaeology of the Roman Empire, at the
University of Oxford. His research interests include the economy of the
Roman Empire, ancient technology, ancient water supply and usage, Roman
North Africa, and archaeological field survey. Recent publications include:
Settlement, Urbanization and Population (ed. with Alan Bowman, 2011);
Alexandria and the North-Western Delta (ed. with Damian Robinson, 2010),
Maritime Archaeology and Ancient Trade in the Mediterranean (ed. with
Damian Robinson, 2011), The Roman Agricultural Economy (ed. with Alan
Bowman), and articles on Saharan trade (Azania, 2012) and Capitolia (JRS,
2013, with Jo Quinn).
 OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

List of Figures

3.1. The dye plant situated in house VII 14, 5 at Pompeii studied by Ph.
Borgard. The photo shows nine heaters for dyeing in several colours,
two basins, and a workshop for cleaning the textile fibres. Photo:
Jean-Pierre Brun. 83
3.2. Perfume shop in the Via degli Augustali at Pompeii excavated in 2002.
The vats belong to oil presses destroyed by the AD 62/63 earthquake.
Photo: Jean-Pierre Brun. 84
3.3. Balloon photograph of the tannery I 5 at Pompeii, showing the work
area in the peristyle, and the tanning room in the lower right corner of
the building. Photo: Jean-Pierre Brun. 85
3.4. The tank used for basket-making in house I 14, 2 at Pompeii studied
by M. Cullin-Mingaud. Photo: Jean-Pierre Brun. 86
3.5. Balloon photograph of the water-mill, used to produce tannin by
crushing bark at Saepinum, Molise. Photo: Jean-Pierre Brun. 87
4.1. Specialized versus non-specialized negotiatores (N = 311).
Image: Candace Rice. 98
4.2. Subdivision of negotiator-inscriptions indicating some form of
specialization (N = 242). Image: Candace Rice. 99
4.3. Places where ‘qui negotiantur’ inscriptions have been found. Map:
Miko Flohr. 105
4.4. Cities with a statio in the Piazzale delle Corporazioni at Ostia.
Map: Miko Flohr. 107
6.1. Selection of Roman footwear, from first century AD (top left) to
fourth century (bottom right). Image: Carol van Driel-Murray. 133
6.2. Child’s sandal from Vindolanda period IV (scale 1:3).
Image: Carol van Driel-Murray. 134
6.3. Shoe style Irthing from Didymoi, Egypt. Image: Carol van
Driel-Murray after Leguilloux (2006: pl. 26). 135
6.4. Shoe style Irthing from Hardknott UK. Image: Charlesworth and
Thornton (1973: fig. 5). Not to scale. 135
6.5. Diagrammatic selection of nailing patterns. Image: Carol van
Driel-Murray. 136
6.6. Vindolanda: off-cut stamped with C(oh). IX Bat(avorum) (scale 1:3).
Image: Carol van Driel-Murray. 137
6.7. A choice of fourth-century shoes from Cuijk (reproductions made
by Olaf Goubitz 2004). Image: Carol van Driel-Murray. 141
OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

xiv List of Figures

6.8. Dated sequence of changing sandal shapes from Voorburg (The
Netherlands). Image: Carol van Driel-Murray. 142
6.9. Sandal soles from Voorburg (mid-second century—c. AD 230).
Image: A. Dekker, Amsterdam Archaeological Centre, University
of Amsterdam. 143
6.10. De Meern: the boatman’s shoes. One of each pair: left, insole
of the nailed shoe; right, sandal (scale 1:3). Image: Carol van
Driel-Murray. 144
6.11. Woerden: the boatman’s shoes. One of each pair: left, nailed shoe;
right, sandal (scale 1:3). Image: Carol van Driel-Murray. 145
7.1. Location of the bakeries in Pompeii. Map: Nicolas Monteix. 155
7.2. Tempering structure in bakery IX 3, 19–20. Image: Nicolas Monteix. 157
7.3. Extension of the basalt pavement in bakery I 12, 1–2. The filling for
the second pavement, seen under excavation (a), cuts the one used
for the first pavement (b) that abutted the blocks. As the plan shows,
the first pavement has been partly dismantled to install the second.
Image: Nicolas Monteix. 158
7.4. Cross-section and exploded isometric view of the kneading machine
in the House of the Chaste Lovers (IX 12, 6). Image: Nicolas Monteix. 161
7.5. Vessels used for the first rising of the dough. The two on the first row
had a lead sheet as a bottom (scale: 1/10). Image: Nicolas Monteix. 162
7.6. Cross section of a running stone (catillus) reused to support a
first rising vessel in bakery IX 5, 4. The fitting vessel was preserved on
site in the same room. Image: Nicolas Monteix. 163
7.7. Oven in bakery V 4, 1. Image: Nicolas Monteix, based on survey
and drawing by François Fouriaux and Sandrine Mencarelli. 166
7.8. Lead boiler embedded in the oven of bakery VII 12, 11 and its
heating surface. Image: Nicolas Monteix. 167
7.9. From grain to bread in bakery V 3, 8. Image: Nicolas Monteix. 172
7.10. The historical evolution of bakery I 12, 1–2. Image: Nicolas Monteix. 174
9.1. Purple-dealers from the Vicus Iugarius (AE 1923: 59). Photo:
Miko Flohr. 205
9.2. Sellia Epyre, aurivestrix from the Via Sacra, Rome (CIL 6.9214).
Photo: Roger B. Ulrich. 207
9.3. The shoemaker Septimia Stratonice from Ostia (CIL 14.4698). 210
9.4. Funerary plaque of the lanarius C. Cafurnius Antonchus and his wife
Veturia Deuteria, Rome (CIL 6.9489). Photo: Lena Larsson-Lovén. 213
9.5. Pompeii, facade to the left of shop IX 7, 7: woman selling products in
shop. Reproduced from Notizie degli Scavi di Antichità (1912: 179). 215
9.6. Funerary relief from Rome depicting a butcher at work and a woman
doing accounting. Photo: Dresden, Staatliche Kunstsammlungen. 216
 OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

List of Figures xv
11.1. Funerary altar of Hermias (inscription 26). Musée de l’Arles antique.
Photo: Nicolas Tran. 256
11.2. Funerary altar of G. Paquius Pardalas (inscription 19). Musée de
l’Arles antique. Photo: Nicolas Tran. 264
11.3. Relief with harbour workers. Musée de l’Arles antique.
Photo: Nicolas Tran. 267
13.1. Progressive clustering in the Birmingham jewellery industry between
1767 and 1913. Reproduced from Mason (1998: 10–11, fig. 5a–d). 304
13.2. Retail clustering on Bond Street, London. Drawn by Alex Santos. 308
13.3. The distribution of bakeries in Amsterdam, 1742. Reproduced from
Lesger (2011: 32, fig. 2). 309
13.4. Officinae lanifricariae at Pompeii. Map: Miko Flohr. 311
13.5. An example of an officina lanifricaria, Pompeii VII 2, 17.
Photo: Miko Flohr. 311
13.6. An example of a probable fullers’ workshop at Timgad. Reproduced
from Wilson (2000: 274, fig. 12.01). 315
13.7. Probable fulleries at Timgad. Reproduced from Wilson
(2000: 279, fig. 12.08). 316
13.8. Circular furnace discovered in Insula XI, Block III, Silchester.
Reproduced from Fox (1895: fig. 2). 318
13.9. Circular furnaces at Silchester. The plan shows all furnaces
mentioned in Fox and St John Hope’s excavation reports that can
be securely associated with a precise location on the city plan.
Drawn by Alex Santos. 319
13.10. Streets characterized by retail clusters in the Forum area, Rome.
Drawn by Alex Santos. 322
14.1. Number of town quarters in Ptolemais Euergetis hosting 1 to 9
craftsmen. Image: Kerstin Droß-Krüpe. 337
14.2. Craftsmen in a neighbouring residential area according to P.Oxy.
46/3300. Image: Kerstin Droß-Krüpe. 339
14.3. Fulleries in Pompeii. Map: Miko Flohr. 341
14.4. Dyeing workshops in Pompeii. Map: Miko Flohr. 341
15.1. The topographical layout of the settlement complex of Aquincum.
Image: Orsolya Láng. 353
15.2. Plan of the civilian settlement in the third century AD.
Map: K. Kolozsvári. 355
15.3. Aerial photo of the civilian settlement. Photo: G. Rákóczi. 355
15.4. Ground-plan of the civilian settlement with the identifiable
workshops and shops of different periods. Map: K. Kolozsvári. 359
15.5. Row of workshops or/and shops along the western side of the cardo
of the settlement. Photo: P. Komjáthy. 361
OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

xvi List of Figures

15.6. Plan of the industrial–commercial quarter in the north-east zone
of the civilian settlement. Image: K. Kolozsvári and Orsolya Láng. 362
15.7. Plan of Building XXIX before the control excavations. Plan: K. Kolozsvári. 364
15.8. Finds from B. Kuzsinszky’s excavation: an altar stone and two marble
statues. Image reproduced from BudRég 3 (1891: 138, fig. 10). 365
15.9. Construction phases in Building XXIX. Image: K. Kolozsvári. 365
15.10. Proportion of workshop refuse and food residue in Building XXIX.
Image: K. Kolozsvári. 367
15.11. Press slab from Building XXIX. Photo: O. Láng. 368
15.12. One of the vats during its second recovery. Photo: O. Láng. 369
16.1. The eastern suburbs of Sagalassos featuring the local Potters’
Quarter. Map: Sagalassos Archaeological Research Project. 378
16.2. View from the Akdağ towards ancient Sagalassos. The town was
laid out on the plateaux in the centre of the image.
Photo: Sagalassos Archaeological Research Project. 379
16.3. General view of the north-west part of the Çanaklı Valley. Ancient
Sagalassos was located on the mountain slopes at the back of the
image, on the left. Photo: Sagalassos Archaeological Research Project. 382
16.4. Aerial view of the eastern suburbs. The town centre is located in the
top part of the image. Photo: Sagalassos Archaeological Research
Project. 386
16.5. The East Slope Workshops. The original workshop is datable between
Augustan times and the first half of the second century AD, and the late
Roman phase between the fourth and the sixth centuries AD.
Plan: Sagalassos Archaeological Research Project. 387
16.6. The results of geophysical prospection in the eastern suburbs. Image:
Sagalassos Archaeological Research Project. 394
16.7. The so-called Coroplast workshops, fourth to sixth centuries AD.
Plan: Sagalassos Archaeological Research Project. 396
16.8. 2013 aerial view of the so-called Coroplast workshops, fourth to sixth
centuries AD. Photo: Sagalassos Archaeological Research Project. 397
16.9. Aerial view of the presumed early Roman imperial Vereinshaus.
Photo: Sagalassos Archaeological Research Project. 398
 OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

List of Tables

6.1. Shoe types per site 140

11.1. Names of the corporati of Arles known from inscriptions 266
12.1. Hierapolis: professional associations and the management of memory 292
13.1. Toponyms from Rome that suggest artisanal clustering 320
14.1. Dyers in Ptolemais Euergetis according to papyrological evidence 336
14.2. Spatial distribution of craftsmen in Ptolemais Euergetis 336
14.3. Spatial distribution of dyers in Ptolemais Euergetis 337
14.4. Toponyms from Rome referring to crafts 346
OUP CORRECTED PROOF – FINAL, 18/1/2016, SPi

List of Abbreviations

Journal abbreviations follow the style of L’Année philologique. Other abbrevi-

ations used in this volume are:
AAST Atti della Reale Accademia delle Scienze di Torino (Turin, 1865-)
AE L’Année épigraphique (Paris, 1888–)
ANRW H. Temporini and W. Haase (eds), Aufstieg und Niedergang der
römischen Welt (Berlin and New York, 1972–)
BE J. and L. Robert et al., Bulletin Épigraphique (in REG 1938–)
CIL Corpus Inscriptionum Latinarum, consilio et auctoritate Academiae
litterarumregiae Borussicae editum (1863–)
FXanthos Fouilles de Xanthos (Paris, 1968–)
IG Inscriptiones Graecae (Berlin, 1902–)
IG2 Inscriptiones Graecae: Editio Minor (Berlin, 1913–)
IGR R. Cagnat et al., Inscriptiones Graecae ad res Romanas pertinentes
(Paris, 1906–27)
IJO II W. Ameling, Inscriptiones Judaicae Orientis, ii. Kleinasien (Tübingen,
2004)
IK Inschriften griechischer Städte aus Kleinasien (Bonn, 1972–)
ILAfr R. Cagnat and A. Merlin, Inscriptions latines d’Afrique (Tripolitaine,
Tunisie, Maroc) (Paris, 1923)
ILGN É. Espérandieu, Inscriptions latines de Gaule (Narbonnaise) (Paris, 1929)
ILLRP A. Degrassi, Inscriptiones Latinae liberae rei publicae (Florence, 1957–63)
ILS H. Dessau, Inscriptiones latinae selectee (1892–1916)
ILTG P. Wuilleumier, Inscriptions latines des Trois Gaules (Paris, 1963)
IMilet P. Herrmann, Inschriften von Milet, ii. Inschriften n. 407–1019 (Vienna, 1998)
IMM O. Kern, Die Inschriften von Magnesia am Maeander (Berlin, 1990)
IPerinthos M. H. Sayar, Perinthos-Herakleia (Marmara Ereglisi) und Umgebung:
Geschichte, Testimonien, griechische und lateinische Inschriften (Vienna, 1998)
IPriene I. Hiller von Gaertringen, Inschriften von Priene (Berlin, 1906)
LGPN P. M. Fraser and E. Matthews (eds), Lexicon of Greek Personal Names
(Oxford, 1987–)
LTUR E. M. Steinby (ed.), Lexicon topographicum Urbis Romae (Rome, 1993–2000)
MAMA Monumenta Asiae Minoris Antiqua (London, 1928–)
SEG Supplementum epigraphicum graecum (Amsterdam, 1923–71, 1979–)
TAM Tituli Asiae Minoris (Vienna, 1901–)
 Introduction
Miko Flohr and Andrew Wilson

Roman cities bustled with commercial activity. Streets were lined with long
rows of tabernae, and people involved in manufacturing, retail, and trade
occupied a defining and prominent place in urban space and urban society.
Craftsmen, shopkeepers, and traders made up a significant proportion (if not
the majority) of the population in many urban communities. In most cities,
manufacturing and retail were never far away—neither physically, nor social-
ly: even for elite families that based their wealth primarily on their property on
the countryside, there were often craftsmen and retailers operating the taber-
nae in the street frontage of their town house.1 Indeed, in many ways, it was
the people who spent their days in shops and workshops who made up the
social core of Roman urban neighbourhoods, while elites played out their role
rather more in the background—in the peristyles and courtyards, away from
the street. In everyday life in urban space, it was much easier to avoid the elite
than it was to avoid urban commerce.
If the Roman world was characterized by unusual or even unprecedented
degrees of urbanization, as seems to have been the case, and if cities were to
such a large extent defined by commerce and manufacturing, then it is hard
fully to understand Roman history without having a clear idea of the social,
economic, and cultural history of urban trade, manufacturing, and retail.2
Fortunately, since the early 1990s, the fundamental centrality of economic life
to our understanding of the Roman world has become widely acknowledged,
with a corresponding upsurge in scholarly interest in the subject. The study of
crafts and trade has been moving steadily towards the centre of the scholarly
field, as is reflected by the space devoted to it in key journals, in monographs

1
Wallace-Hadrill (1991).
2
On urbanization in the Roman World, see Bowman and Wilson (2011). For Italy, see also
De Ligt (2012: 193–246).
2 Miko Flohr and Andrew Wilson

written by generalists, and, recently, in handbooks and companions.3 Crucial-

ly, evidence for urban manufacturing and commerce is no longer the exclusive
domain of specialists working in relative isolation: an increasingly interdis-
ciplinary and international scholarly field is emerging, and it is a field that is
quickly moving away from the models that characterized the study of everyday
work in the ancient world for most of the twentieth century.
Things were, of course, different in the past. For quite some time, the
shadow of elite disgust loomed over the study of Roman crafts and trade. In
particular, the famous passage in the first book of Cicero’s De officiis has been
quoted time and again to illustrate the disdain of the Roman elite for crafts and
trade, and to illustrate the marginal or even controversial position of these
activities in Roman society: the workshop was no place for a free-born person,
working for wages was little better than slavery, and trade was problematic as
well.4 Forty years ago, Moses Finley argued at length that Cicero’s statement
represented ‘not a bad guide to prevailing values’, explicitly refuting those who
dismissed it as ‘elite snobbery’.5 Finley was not alone. Many twentieth-century
scholars made similar comments based on Cicero and other authors. Craftsmen
and traders were not generally or automatically seen as ‘normal’ Romans, and
there has been a strong emphasis on the servile or freed status of many craftsmen
and traders, which not only meant that they suffered from the stigma of slavery,
but also quietly implied that they did not have true ‘Roman’ roots.6 Essentially,
mainstream scholarly thinking on crafts and trade long adopted, implicitly or
explicitly, the perspective of the Roman elite, and marginalized both the activities
and the people involved in them—either by simply neglecting them as irrelevant,
or by emphasizing how they differed from ‘normal’ Romans, and how contro-
versial their status was. Since the mid-1990s, this way of thinking has lost ground,
and scholars have more and more abandoned the perspective of elite authors.
The present volume reflects this development towards the ‘normalization’ of
crafts and trade, and aims to add to it. The sixteen chapters that follow this
introduction discuss a wide variety of data sets—both textual and archaeological—
and focus on a wide variety of research questions, but they share some common
features. In the first place, the emphasis lies, throughout the book, on analysing
processes and decisions rather than on measuring performance. The macroeco-
nomic issues that were prominent in past decades, such as the economic impact of
certain crafts and trades for the income portfolio of specific cities, or the nature of
urban economies in general, play only a marginal role here.7 This is emphatically

3
Journal articles: Bradley (2002); Wilson (2002b); Ellis (2004); Tran (2007); Flohr (2013b).
Monograph chapters: e.g. Clarke (2003: 95–129); Petersen (2006); Cuomo (2007: 77–102). Hand-
books: Kehoe (2007); Morley (2007); Hawkins (2012); Broekaert and Zuiderhoek (2013); Liu (2013).
4 5
Cic. De off. 1.150–1. Finley (1973: 57).
6
On the role of slaves and freedmen in the Roman economy, see, e.g. Joshel (1992) and
Mouritsen (2011: 206–47).
7
On this issue, see also Flohr and Wilson, Chapter 1, this volume.
 Introduction 3

not because we think that these are unimportant questions, but because we feel
that, before we can properly address them in the light of the evidence now
available, there needs to be a prior stage of research in building up a new picture
from the ground up. Hence, the emphasis here is on the micro-scale: the direct
social and physical environment in which people lived and worked, and the
smaller and larger decisions that these people made to realize their social and
economic ambitions. Indeed, a second feature that connects most of the chapters
in this volume is that the focus is not only on crafts and trade, but also, and
perhaps more, on craftsmen and traders, and the economic strategies these people
embraced in their everyday working lives.
There are several developments in the wider field of Roman social and
economic history that have informed the approaches in this volume. It seems
relevant briefly to highlight three of these at this point. In the first place, recent
years have seen the emerging use of social network perspectives to look at
aspects of Roman social history. Groundbreaking in this respect was the 2008
monograph by Giovanni Ruffini on Social Networks in Byzantine Egypt, but
scholars working on issues more closely related to this volume have also
embraced social network theory, including, prominently, Shawn Graham in
his work on the brick industry in the Tiber Valley.8 One of the present authors
has applied a social-networks approach to the material remains of fulling work-
shops to understand the social dynamics of the world in which fullers spent their
days.9 In a 2011 article on Roman merchants, Wim Broekaert emphasized the role
of social networks in trading processes.10 A social-network perspective is essential
to the study of crafts and trade, both for understanding the dynamics of everyday
economic processes and for making sense of the social position of craftsmen and
traders. In particular, it forces us to think harder about the economic impact of the
nature and strength of the social ties between economic actors, and about the
communicative contexts in which work-related statuses were negotiated and
maintained. A social network approach also makes the traditional discourse
about the ‘low’ and ‘dependent’ statuses of craftsmen and traders increasingly
hard to defend: if the social statuses of these people were for a large part constructed
within networks in which many agents had comparable backgrounds, as seems to
have been the case, they can scarcely be characterized as ‘low’ or ‘dependent’—even
if many craftsmen and traders had social superiors to deal with.
A second point that is less often explicitly stated but seems to have become
widely accepted in recent years concerns the nature of the relation between
society and economy. In the second half of the twentieth century, several
classical scholars, most prominently Finley, championed Karl Polanyi’s idea
that pre-modern economies were ‘embedded’ in society—meaning, essentially,
that societal and cultural norms dictated the terms of everyday economic life.11

8 9
Ruffini (2008); Graham (2006). Flohr (2009; 2013a: esp. 242–87).
10 11
Broekaert (2011). Polanyi (1944: 45–58); Block (2001: pp. xxiii–xxv).
4 Miko Flohr and Andrew Wilson

While few scholars have rejected the idea that social patterns and cultural
preferences could have significant impact on economic processes, many
scholars now seem to think of the relation between society and economy in a
more reciprocal way: economic processes could also have an impact on social
and cultural patterns. This, of course, comes naturally if one approaches urban
economies from a spatial perspective: the almost total dominance of commercial
space in Roman streetscapes makes it hard to maintain that urban social and
cultural processes were not somehow affected by economic life. Yet it also
follows from the proliferation of evidence for social networks that were primar-
ily economically defined—such as the many professional associations (collegia)
that bought burial grounds for their members, or the many trading diasporas
that can be found throughout the Roman world.12 Religious feasts, such as the
Quinquatrus, could also to a large extent be organized along divisions that were
based on urban economic life.13 This prominent role of (everyday) economic life
in shaping social and cultural practice obviously further enhances the need to
understand the historical dynamics of urban crafts and trade in the Roman world.
A final issue to be highlighted here concerns the impact on recent ap-
proaches to urban crafts and trade of a loosely defined set of ideas commonly
known as the ‘New Institutional Economics’. This analytical perspective was
catapulted into the centre of debate on the Roman economy in 2007 by the
editors of the Cambridge Economic History of the Greco-Roman World.14
Approaches inspired by New Institutional Economics focus heavily on under-
standing the relation between cultural, societal, and legal institutions, and
economic processes, and, particularly, on the role played by such institutions
in overcoming information deficits and lowering transaction costs.15 This
focus has had a mixed effect on the study of urban crafts and trade. On the
one hand, scholars are encouraged to invest heavily in understanding the
everyday processes of trade and manufacturing, and the social environment
in which these took place, and this has clearly enriched our understanding of
trade and manufacturing. On the other hand, however, the common emphasis
on the meaning of individual institutions has sometimes tempted scholars
towards a level of abstraction that lacks historical specificity—essentially
leading to theoretical modelling rather than historical analysis.16 Moreover,
approaches based on New Institutional Economics have sometimes put a lot of
emphasis on the dominant role of institutions in guiding economic behaviour,
and in doing so they have tended to privilege structure and underplay the role

12
On professional associations, see Van Nijf (1997); on trading communities, see Verboven
(2011); Terpstra (2013).
13
See, e.g. Ovid, Fasti 809–48.
14
Morris, Saller, and Scheidel (2007). See also Frier and Kehoe (2007).
15
See, e.g. Frier and Kehoe (2007); Verboven (2012).
16
Cf. Verboven (2012: 913). For a spirited (and sometimes overstated) critique of New
Institutional Economics as applied to ancient history, see Boldizzoni (2011).
 Introduction 5

of agency in economic processes.17 In writing the history of urban crafts and

trade, it is not only the institutions that matter, but also their specific role in
particular places and moments in the Roman world, and the way they were
acted upon by specific economic agents: in everyday discourse, institutions
must be seen as starting points for economic strategies rather than end
points—they may be embraced in some situations, but might be circumvented
or neglected in others, and they may be long-lasting or short-lived. Neverthe-
less, the emerging trend among classical scholars to study urban crafts and
trade on the level of the everyday social and economic processes may, in some
way, be related to the emergence of New Institutional Economics as an
analytical perspective, and this broader trend is clearly reflected in the contri-
butions to this volume, even if the way in which authors in this volume relate
to the New Institutional Economics varies considerably.
The present volume is divided into four parts. It starts with three chapters
focusing on the intellectual history of the study of Roman crafts and trade—an
aspect of the field that does not often receive the attention it deserves. After
this exercise in scholarly history, the history of craftsmen and traders in the
Roman world is approached from three different angles. The second part of
the book consists of four chapters all focusing on the economic aspects of
crafts and trade, and on the core economic strategies embraced by craftsmen
and traders to secure their position in the market. Subsequently, there are five
chapters concentrating on the social contexts and social networks surrounding
the daily work of craftsmen and traders, and the degree to which these shaped
everyday economic life. The final four chapters discuss the position of crafts in
their spatial environment, addressing questions related to the genesis of urban
commercial landscapes. Obviously, the separation between the three parts is
not rigid: if social, economic, and spatial issues sometimes overlap in everyday
practice, the same is true in this volume. Moreover, an effort has been made to
provide a mix of archaeological and textual approaches for each of the three
themes discussed, and to maintain a reasonable geographical balance between
the eastern and western part of the empire, and Mediterranean and non-
Mediterranean regions.

HISTORY OF RESEARCH

While scholarly priorities, for understandable reasons, often lie more with
discussing past realities than with discussing modern discourse about them, it
is arguably essential also to understand the factors and actors that have shaped
17
See, e.g. Scheidel (2012: 9–10): ‘overriding significance of historically specific “rules of the
game”, the incentives and constraints that were instrumental in determining Roman economic
development.’
6 Miko Flohr and Andrew Wilson

our ideas about Roman craftsmen and traders over the last few centuries. The
three chapters that follow this Introduction serve to sketch the development of
this discourse. The first chapter, by the present authors, discusses the devel-
opment of the debate in general terms, focusing specifically on the German
and Anglo-Saxon scholarly traditions. It assesses the relative impact of new
evidence and new ideas on discourse about Roman urban craftsmen and
traders in the nineteenth and twentieth centuries, but it also highlights the
key role played by certain individual scholars and their networks. Subsequently,
Carla Salvaterra and Alessandro Cristofori sketch the development of Italian
scholarship on the theme in, specifically, the twentieth century. This chapter
emphasizes how political developments on a national level, both in the fascist
era and in the post-war decades, had a decisive influence on the way in which
Italian scholars discussed Roman crafts and trade. At the same time, Salvaterra
and Cristofori argue that Italian scholarship has, since the mid-1990s, increas-
ingly integrated itself into international debates. The chapter by Jean-Pierre
Brun focuses on French scholarship, and shows how recent approaches to the
archaeology of urban workshops in Roman Italy have been shaped by a long
French tradition of scholarship on urban crafts that goes back to the early
twentieth century. Brun puts particular emphasis on the gradual integration of
the material remains into the debate, arguing that the major breakthrough in
this respect came only in the late 1980s.
Together, these three chapters, though different in their approach and
scope, give a good impression of the dynamics of the debate, and of the impact
of ideological and political developments in the outside world on discourse
within classical studies: one may think of the devastating impact of the First
World War on German Altertumswissenschaft, and of the corporatism pro-
moted by the Fascist regime on Italian approaches to craftsmen and traders.
At the same time, all three highlight the crucial role of scholarly networks and
institutions in fostering debate—the networks around Mommsen and Meyer
in that respect do not differ fundamentally from the classical seminar of the
Istituto Gramsci, or, indeed, of the French collaborative effort to the archae-
ology of urban workshops in southern Italy.
While it may at first sight seem strange to treat the three traditions
separately—after all, should not any scholar working on the Roman economy
be expected to cope with the relevant scholarship in all the major western
European languages?—it is in fact clear from these chapters that the different
traditions have had rather different concerns and focuses, and that the effect of
language barriers has been significant—the impact of many works has been
limited outside the scholarly tradition in the same language. On a general level,
the three chapters suggest that the combined German/Anglo-Saxon discourse
acted as the dominant tradition from the nineteenth century right up to the
present day. Debates in the English and German languages always interacted
closely, and there is a more-or-less direct sequence of debate from Rodbertus
 Introduction 7

through Meyer and Rostovtzeff to Finley and beyond, especially in the

so-called consumer city debate. Because of linguistic and logistical barriers,
francophone and italophone discourses for a long time developed more
independently, and, more often than not, they were on the receiving end of
influence—not only from Rostovtzeff or Finley, but also, as Salvaterra and
Cristofori suggest, from scholars like Joshel.18 In contrast to anglophone
scholarship, French and Italian scholarship has been less fixated on the
‘consumer city debate’, and Finley’s views gained less traction in France and
Italy. French research has focused instead to a considerable extent on advan-
cing our knowledge of the detail and infrastructure of workshops, to enhance
the evidence base for interpretation.19 Italian scholarship has paid much
attention to the role and concept of the workshop. Nevertheless, there is, of
course, a close parallel between the emergence of social history in the Anglo-
Saxon world in the late 1960s and the appearance of a generation of Marxist
scholars in Italy (and France) around the same period. Here, it is zeitgeist
rather than anything else that is at stake. Similarly, the emergence of provincial
archaeology after the Second World War was a European development that,
especially in Germany and France, as Brun suggests, may have its roots in the
post-war (re)construction boom. The same is true, obviously, for the emer-
gence of rescue archaeology, which Brun correctly sees as a watershed in the
archaeology of urban crafts. Scholarly traditions thus in many ways developed
parallel to each other throughout the twentieth century. Arguably, however,
future scholars might mark the digital revolution of the late 1990s and
onwards as another watershed, as it has fostered a sharp increase in interaction
across national and linguistic boundaries. The time is now ripe to integrate
these various traditions and, with a new understanding of the archaeology of
urban manufacturing processes, to reassess what the mass of evidence we now
have means in terms of the wider urban economy.

ECONOMIC STRATEGIES

The second set of chapters looks at the core issue of professional crafts and
trade: the strategies embraced by craftsmen and traders to make sure that their
everyday work actually resulted in profit. While it may be true that economic
agents in the Roman world were not seeking to ‘maximize profit’ in the
modern sense of the word, they had a clear interest in running their businesses
efficiently, in order to minimize risks, or reduce costs, especially if it is true

18
Rostovtzeff (1926); Finley (1973); Joshel (1992).
19
e.g. Borgard (2002); Borgard et al. (2002, 2003); Borgard and Puybaret (2004); Leguilloux
(2002, 2004); Monteix (2010); Brun, Chapter 3, this volume; Monteix, Chapter 7, this volume.
8 Miko Flohr and Andrew Wilson

that those who earned their living in urban economies had to deal with
substantial economic insecurity on an everyday basis, as Cameron Hawkins
has recently argued.20 Moreover, those craftsmen and traders who were not
struggling at subsistence level but were slightly better off may have had socio-
economic ambitions that were best furthered with a well-functioning business.
In other words, there were reasons for all to think hard about how to approach
their everyday work.
For craftsmen and traders alike, the two key strategic decisions to be made
were, first, what to do, and, second, how to do this? Such decisions were
profoundly influenced by an array of factors, including the specific (local)
economic circumstances in which people operated. The chapter by Candace
Rice discusses the economic strategies of maritime traders. It starts from the
observation that the epigraphic evidence for negotiatores suggests that special-
ization among traders was widespread, and that, by focusing on a particular
niche, traders could become successful entrepreneurs. Rice sketches an array of
strategies for maritime trade and transport, and shows how each of these
strategies can also be traced back in our evidence, which suggests that trade
practice in the Roman Mediterranean was flexible and adapted to specific
economic circumstances. To enhance their chances for economic success,
some traders also formed permanent trading communities in key port cities.
The subsequent chapter by Kai Ruffing also emphasizes the issue of special-
ization, but on a more local level. Focusing particularly on epigraphic and
documentary evidence from Egypt and Asia Minor, Ruffing investigates three
incentives for specialization: first, the intense competition on the market, which
drove craftsmen to explore economic niches that would give them a decent
customer base; second, the possibility to increase the quality and quantity of their
output by focusing on a smaller subset of services or products, and, third,
location factors—the specific local circumstances, such as the availability of
certain resources, that caused certain places to have a high concentration of
particular crafts. Ruffing argues that the market—and particularly consumer
demand—was the main driving force for specialization, and that the evidence
illustrates Adam Smith’s observation that the extent to which specialization took
place depended on the size of the market.
While the chapters by Rice and Ruffing focus on strategies at a general level,
the other two chapters offer case studies each looking at strategies in the
manufacturing of a specific product category. One discusses the manufactur-
ing of footwear, the other the production of bread. In her chapter on the
footwear economy, Carol van Driel-Murray uses consumer theory to sketch
the economic landscape in which manufacturers operated in, particularly,
Roman Europe. Starting from the observation that there were empire-wide

20
Hawkins (2012, 2013).
 Introduction 9

trends in footwear design, van Driel-Murray argues that shop-bought shoes

were widely accessible for men, women, and children alike, and that the
footwear economy of Roman Europe should be seen as a consumer economy,
with craftsmen adapting their strategies to diversified and changing consumer
demand, sometimes customizing their designs to specific needs of individual
consumers, but more often following standardized decorative patterns that
allowed for a reasonably high productivity. Interestingly, however, the leather
economy seems to have remained an economy of outsiders: after the collapse
of Roman rule, leatherworking technology seems to have disappeared. The
bread economy worked rather differently from the footwear economy:
demand was much more predictable, and much less diversified—bread is,
basically, bread. The commercialization of bread production that is well
attested for Roman Italy led to the emergence of combined mills–bakeries
that produced standardized bread loaves on a large scale.21 The bakeries of
Pompeii are the focus of the chapter by Nicolas Monteix, who discusses the
ways in which the technology used for milling, kneading, and baking, and the
spatial organization of the operational sequence could improve the function-
ality of the workshop. Emphasizing that all workshops, essentially, were a
compromise between available space and economic ambition—all the bakeries
in Pompeii had to be integrated into pre-existing buildings—Monteix argues
that the rationalization of workshop designs increased with their size, and
shows how the analysis of the design of these workshops can help to under-
stand the economic history of Roman urban crafts. One might remark that the
correlation between workshop size and rational design that Monteix observes
at Pompeii is also found elsewhere: at Ostia, a bigger town with a larger
market, and one that unlike Pompeii, continued in existence beyond the late
first century AD, even larger, highly rationalized mill-bakeries are found, some
of which are purpose-built as new constructions.22 This chapter complements,
from an archaeological point of view, Ruffing’s text-based discussion of
specialization: Monteix shows how entrepreneurs who had found a niche
could use technology and the organization of the production process to
enhance efficiency and increase profit.
Taken together, these four chapters place a lot of emphasis on limiting the
range of products produced or offered, and on reducing the costs involved in
producing or offering them. As might be expected, all stress the central role of the
market in strategic decision-making: economic actors—whether or not one
prefers to call them ‘rational’—responded to market circumstances in very
strategic ways and adapted their business strategies according to the challenges
and opportunities they encountered. What these chapters also quietly suggest is

21
For Ostia, see, e.g. Bakker (1999); for Rome, and the use of water power for milling grain, cf.
Wilson (2000; 2002a: 12–15).
22
Wilson (2008).
10 Miko Flohr and Andrew Wilson

that, with the increased complexity and urbanization of the Roman world, and
with the increased integration of economic networks, the array of economic
strategies embraced became considerably wider than it had been before, even if
traditional ways of working did not start to disappear (or even lose prominence).
The point of Roman economic history is not that traditional, small-scale, low-
investment models of manufacturing and distribution were widespread. The
point is that, alongside these, new models developed that owed their appearance
to the unique economic circumstances in the Roman world.

SOCIAL ENVIRONMENT

Roman craftsmen and traders were, of course, not living and working in
isolation from the outside world. As households were the basis of the large
majority of urban workshops in the Roman world, economic life and family
life overlapped to a considerable degree, and took place in the same physical
environment—as is well exemplified in almost all urban excavations where
workshops have been found.23 As many urban craftsmen were also involved in
retail, they enjoyed interaction with customers, and, as they often spent their
days in shops visible to all passers-by, they were publicly known for their
professional skills, which enabled them to develop a public occupational
identity. This, in turn, fostered the emergence of informal and formal group-
ings of people working in the same (or similar) craft, which eventually would
lead to the system of professional associations that is so well attested in the
epigraphic record. At the same time, many people involved in crafts and trade
had long-lasting ties with social superiors who had legal or practical authority
over parts of their social or economic lives. Craftsmen and traders were thus
part of several, sometimes overlapping, social networks, and understanding
these networks is essential for understanding both the social and the economic
history of Roman urban crafts and trade.
The chapter on apprenticeship by Christel Freu highlights the role of formal
(and informal) education in the construction of such networks. Her chapter
shows how apprenticeship provided pupils with much more than just the skills
needed to perform a certain craft in a decent manner: it also gave them the
possibility to profit from the good name of their master, and it introduced
them to relevant social and professional networks. Freu argues that formal
apprenticeship was common practice, which suggests that it played a central
role in the reproduction of the professional networks, albeit alongside the
more informal system in which knowledge and networks were transferred

23
For Pompeii, see Flohr (2007, 2012).
 Introduction 11

within households from father to son. The subsequent chapter by Lena

Larsson Lovén discusses the position of women in urban professional net-
works. While female work tends often to remain much less visible in our
textual and iconographic sources, there is sufficient evidence to make it clear
that women played a crucial role in many sectors of urban economies. Often,
however, they did so within a family business that, publicly, was mostly
associated with their husbands. Thus, while men were often commemorated
with reference to their occupational identity, their wives were not. Indeed,
Larsson Lovén suggests that many epigraphically attested working women
were actually working in rather exceptional contexts, which means that
occupations with which these women were commonly associated give a false
impression of what women’s work was like. Wim Broekaert analyses the role
of freedmen in the economic network of their former owner. The specific
social and legal position of freedmen brought both the freedman and his
patron certain economic opportunities. Using the freedman as an agent
reduced risk for the patron, and provided the freedman with ready access to
capital and to the business network of his patron. There were various models
along which freedman agency could take shape, with varying degrees of
independence, but the bottom line is that working in close collaboration
with the former owner provided more chances of success, resulting in
‘dependent’ freedmen being more visible in our sources than ‘independent’
freedmen, who worked without their patrons. Arguably, the close ties between
patrons and social dependants were fundamental to many networks of crafts
and trade throughout the Roman world.
Whereas the first three chapters of this section stayed close to the household
and the networks of friends and dependants surrounding it, the last two
chapters present two case studies on professional associations—one in the
Latin-speaking West, the other in the graecophone East. Nicolas Tran dis-
cusses the professional associations in the harbour economy of Arles, showing
that there was a clear hierarchy of professional associations, with the navicu-
larii at the top, but with key occupations, particularly negotiatores and merca-
tores, almost completely lacking from the record. Tran argues that the specific
context of Arles meant that these people played a central role in the local
community as, for example, seviri augustales, and preferred to refer to their
political, rather than to their occupational, identity. At Arles, these people
constituted a ‘plebeian’ elite, whose existence emphasizes the integration of crafts
and trade into the urban community. This connects well with the argument of
Ilias Arnaoutoglou, who discusses the evidence of professional associations from
Hierapolis, arguing that they were particularly involved in ‘memory manage-
ment’: epigraphic evidence reveals how individuals, not necessarily themselves
members of professional associations, endowed associations with sums of money
to maintain their tomb, or perform certain rituals related to their post-mortem
well-being—to the social benefit of both the commemorated individual and the
12 Miko Flohr and Andrew Wilson

professional association. Craftsmen and traders thus were not at all marginal
people—neither in Arles, nor in Hierapolis: they were well integrated into the
urban community and, through their associations, maintained close ties with the
urban elite.
What unites these five chapters, besides their focus on social networks, is
the emphasis on the role of agency—by individuals or by groups: craftsmen
and traders were no passive victims of a subordinate social position, but
contributed, actively, to building up and maintaining their social and eco-
nomic status—men and women alike. This is not to say, of course, that
craftsmen and traders were completely autonomous. More often than not,
there were social superiors in the background who could have considerable
influence, and there were always the social and cultural expectations that
limited people’s radii of action and coloured the way in which their decisions
were perceived: any autonomy was always bounded.24 Yet, however low the
socio-economic status of some of these people may have been in the eyes of the
literary elite, almost all were, for most of the time, primarily surrounded by
others who were in similar positions, and with whom relational ties could be
established on a more-or-less equal basis. It is those everyday social ties that
coloured the world of craftsmen and traders—to a much larger extent than
any judgement by outsiders.

ECONOMIC LIFE AND THE URBAN L ANDSCAPE

The position of crafts, trade, and retail in urban space is an issue that has
enjoyed some popularity in recent years—for understandable reasons, given
the dominance of commercial activities in the streets of many Roman cities.25
There are two sides to this.26 First, there is the physical interaction of crafts
and trade with the direct environment, including the sensory impact of crafts,
and eventual measures taken to minimize nuisance caused. Second, there is
the spread of economic activities over the urban area, and the degree to which
certain activities were, in some way or another, spatially concentrated. A key
question in the debate so far has been to which extent there were moral or legal
codes dictating the location of certain activities, but it has increasingly been
questioned whether this actually was the case, and whether other factors
should not play a more prominent role in the debate.27 It is increasingly

24
On the ‘bounded autonomy’ of fullones in Roman Italy, see Flohr (2013a: 307–9).
25
See, e.g. Esposito and Sanidas (2012). See also Flohr and Wilson, Chapter 1, this volume;
Goodman, Chapter 13, this volume.
26
Cf. Flohr (2013a: 189–239), also distinguishing between ‘public’ and ‘private’ environments.
27
Goodman, Chapter 13, this volume.
 Introduction 13

recognized that there is plentiful evidence for polluting and smelly industries
inside the walls of cities—the fish-salting industry, for example, is now seen to
be far more urban than we used to think, and where it is extra-mural this is
more a result of the need for space for large-scale workshops than because of a
desire to keep smells out of the city.28 The four chapters of this section reflect
the development of the debate over the location of urban industries.
The first two chapters focus on the issue of economic clustering in Roman
cities—an issue that has not (yet) received a lot of explicit scholarly attention, but
is fundamental to our understanding of urban economic geographies in the
Roman world. Penelope Goodman approaches the issue from a predominantly
archaeological perspective. She starts from two modern (British) examples of
economic clusters, and then goes on to analyse three cases of clustering that are
identifiable in the archaeological record, discussing evidence from Pompeii,
Timgad, and Silchester, before assessing the evidence for clustering in texts.
Goodman argues that clustering was a common phenomenon in Roman cities,
and that this was essentially due to landlords—the elite—allowing the market to
do its work, as this was in their social and economic interests. The subsequent
chapter by Kerstin Droß-Krüpe focuses on a specific subset of crafts—textile
crafts—and starts from the papyrological evidence from Roman Egypt. Tax and
property registers show how, in Roman Egypt, textile crafts such as dyeing and
fulling were not clustered, but appear to have been more or less equally distrib-
uted over the urban quarters. Using modern cluster theory, Droß-Krüpe argues
that this also makes sense, as most textiles are convenience goods, rather than
shopping goods, which means that clustering would actually be disadvantageous
for entrepreneurs. This leads then to a rethinking of the craft-related toponyms
known from textual sources, which need to be approached with caution, and to a
comparison with Medieval Europe, where local governments and guilds, much
more than in Roman cities, played an active role in establishing craft clusters. At
first sight, Droß-Krüpe appears more cautious when it comes to clustering than
Goodman, but the two arguments complement rather than exclude each other:
while Goodman shows that clustering could emerge in Roman cities, and was not
an exceptional phenomenon, Droß-Krüpe shows that circumstances in the
Roman world were still much less advantageous for clustering than in the
guild-dominated cities of Medieval Europe.
The final two chapters present two case studies that discuss the develop-
ment of economic landscapes of a specific city over a longer period. These
focus on Aquincum and Sagalassos respectively. The chapter by Orsolya Lang
gives a thorough analysis of the archaeology of manufacturing and retail in the
Roman city of Aquincum. While most of this site was excavated in the late
nineteenth century, recent fieldwork has reshaped our view of the commercial
history of this town. As the civilian settlement—situated some 3 kilometres

28
Wilson (1999; 2002b; 2007: 178–80); Ellis (2011).
14 Miko Flohr and Andrew Wilson

north of the legionary camp—developed from a vicus into a municipium and,

later, into a colonia, its commercial landscape became more varied, and spread
out over a larger area. Yet the original main street of the vicus always kept a
strongly commercial character, illustrating how existing situations impacted
on subsequent developments. A similar point emerges from the contribution
by Jeroen Poblome, who discusses the urban history of the potters’ quarter at
Sagalassos. Production of red-glazed fine-ware pottery started here in the
Augustan period, probably as a result of a concerted effort of the community
to move a pre-existing Hellenistic potters’ quarter from the place where the
Roman odeon now stands. Production continued until the sixth century AD.
Poblome outlines how the area gradually became more densely built up, and
became dominated by small-scale pottery workshops: local economic circum-
stances left little room for investment on a larger scale. The analysis shows the
impact that a locally important craft could have on the urban landscape—even
if the Roman potters’ quarter was not directly in the city centre.
Readers will notice that these four chapters in several respects move in similar
directions. Two issues deserve to be highlighted here. First, in the end, all chapters
move away from the ‘moral’ approach to the position of economic life in urban
space: no argument is made in favour of the idea that cultural (negative) attitudes
of the Roman elite towards manufacturing dictated the economic topography of
cities, or restricted location choices for certain trades. Rather, the location of shops
and workshops has become a matter of economic strategy, being determined by
customer base, supply lines, and land prices. Perhaps, this may even lead us to
reconsider slightly our ideas on the spatial position of trades like prostitution:
prostitution, at Pompeii, was not concentrated in back streets in the city centre
because nobody wanted to know it existed, but because customers did not want to
be seen—economically, these locations were more viable than highly visible hot
spots along the main streets. A second issue that emerges from these chapters is
the role of path dependence in the history of urban economic topographies:
whatever is already there has a big impact on what subsequently develops. This
is true both for clusters and for commercial landscapes in general. As Goodman
argues, economic clusters, once established, reinforce themselves—and this is
precisely what can be seen at Sagalassos; similarly, in Aquincum, the cluster of
commercial activities in the original vicus had a significant impact on the com-
mercial landscape of the later city—even though the geographic emphasis shifts
away from the original main road of the settlement.

CRAFTSMEN, TRADERS, AND THE WIDER DEBATE

Methodologically, the volume highlights the significance of space: understand-

ing the spatial context in which the work took place is seminal to all aspects of
 Introduction 15

urban crafts and trade—social, cultural, and economic. This obviously points
to the important role archaeology can play in the debate (and to the problem-
atic accessibility of much of the archaeological evidence), but, as exemplified
by Droß-Krüpe’s chapter, texts, too, contain spatial information. Rather,
therefore, the argument is that the analysis of crafts and trade needs to be
spatially aware: whether discussing the organization of the production process,
social interaction among professionals, or the social position of craftsmen and
traders, space is a crucial concept that cannot be overlooked.
At the same time, the volume shows the relevance of approaches integrating
textual and archaeological evidence: textual and material evidence each pro-
vides only a partial picture of the historical realities of crafts and manufactur-
ing in the Roman world, and the different data sets are extremely badly
connected to each other—direct links between the textual and archaeological
record, such as on the tomb of Eurysaces in Rome, through the combination of
iconography and epigraphy, or in Pompeii, through election notices written
on the walls of workshops that have left identifiable remains, are rare.29
Building more bridges between text and archaeology is essential for our
understanding of urban manufacturing, trade, and retail: any approach priv-
ileging one over the other risks ending up seriously flawed.
As to the everyday world of craftsmen and traders, there is a shared emphasis
on the role of both consumers and elites. The users of products made and sold
were prominent figures in the lives of many craftsmen and traders. While their
role in shaping production and retail strategies and economic topographies
seems straightforward, it should be emphasized that they also played a role in
the social positioning of craftsmen and traders: direct interaction with customers
was a key element in the occupational identity of craftsmen and traders alike.
Indeed, it may be suggested that occupations with more direct interaction with
customers had a higher chance of developing a strong occupational identity—
and, particularly if one catered for an elite audience, this also enhanced the
chances for epigraphic commemoration.30 The impact of the elite was shaped
through their landowning and patronage—though it may be pointed out that the
chapters in this volume almost unanimously refer to local elites rather than to the
senatorial and equestrian elites whose cultural ideas pervade literary discourse.
These elites invested in trade and manufacturing because they wanted to get
something out of it—socially or economically. Crucially, however, their impact
on everyday decision-making processes seems in many cases to have been
limited: they operated in the background, and left the business to those who
had most experience of conducting it, giving craftsmen and traders a consider-
able amount of de facto autonomy.
29
The tomb of Eurysaces, of course, combines iconographic depictions of the bread-baking
process with an inscription describing the man as a pistor. See also Ruffing, Chapter 5, this
volume.
30
See the discussion of inscriptions commemorating craftsmen in Flohr (2013a: 328–32).
16 Miko Flohr and Andrew Wilson

However, ultimately, the picture that emerges from this volume is very much
one of variation. While it is patently obvious that certain strategies, institutions,
and practices were known and embraced throughout the Roman world—one can
think of specialization, clustering, professional networks, and client agency—the
contributions to this volume, collectively, also suggest that the ways in which
these took shape, and the extent to which they were popular in certain places and
at certain moments, could be profoundly influenced by specific local circum-
stances. Rather than being an impediment to understanding the history of
craftsmen and traders in the Roman world, this variation must be seen as a key
historical fact. Practically, it points to two issues. In the first place, it suggests
flexibility: when we are looking at urban crafts and trade in the Roman world, we
are looking at people who were able to adapt their ways of working according to
the problems (or opportunities) they encountered. Flexibility and adaptation
were key properties of Roman craftsmen and traders. The implication is that the
world of manufacturing, trade, and retail was also open to innovation, especially
under exceptional circumstances—as can be found, for example, in the Roman
metropolis and the other major cities of the Roman Mediterranean: extreme
urbanization caused problems and brought opportunities for which traditional
strategies and practices sometimes were less adequate. Similarly, the romaniza-
tion, and urbanization, of Europe and Britain created a new socio-economic
landscape that fostered new strategies and practices, or caused existing practices
and institutions to take on new forms or meanings. The second issue follows
directly from the first: if variation and flexibility are central to the history of
urban crafts and trade in the Roman world, a task high on the scholarly agenda
should be making sense of it. Thus, without denying the need also to explore
institutions, practices, and strategies on a conceptual or empire-wide level, and
without contesting the usefulness of exploring theoretical models for thinking
through possible historical scenarios, ‘time’ and ‘place’ also need to be central
concepts in the historiography of urban crafts and trade in the Roman world: it is
only through the thorough analysis of situations in specific localities and at
specific moments, and through comparing these, that it becomes possible to
understand how this flexible system worked, and developed—in other words, the
debate needs to take place in the tension between model-building and close
reading. The chapters in this volume offer some suggestions about ways in which
this might be done.

REFERENCES

Bakker, J.-Th. (1999). The Mills-Bakeries of Ostia: Description and Interpretation.

Amsterdam.
Block, F. (2001). ‘Introduction’, in K. Polanyi, The Great Transformation: The Political
and Economic Origins of Our Time. Boston, pp. xviii–xxxviii.
 Introduction 17
Boldizzoni, F. (2011). The Poverty of Clio: Resurrecting Economic History. Princeton.
Borgard, P. (2002). ‘A propos des teinturies de Pompei: L’Exemple de l’officina
infectoria V 1, 4’, in J. C. Béal and J. C. Goyon (eds), Les Artisans dans la ville
antique (Collection Archéologie et Histoire de l’Antiquité). Lyons, 55–67.
Borgard, P., and Puybaret, M. P. (2004). ‘Le Travail de la laine au début de l’empire; l’apport
du modèle pompeien. Quels artisans? Quels équipements? Quelles techniques?’, in
C. Alfaro, J. P. Wild, and B. Costa (eds), Purpureae Vestes: Textiles y tintes del Medi-
terráneo en época romana. Valencia, 47–59.
Borgard, P., Forest, V., Bioul-Pelletier, C., and Pelletier, L. (2002). ‘ “Passer les peaux en
blanc”: Une pratique gallo-romaine? L’Apport du site de Sainte-Anne à Dijon
(Côte-d’Or)’, in F. Audoin-Rouzeau and S. Beyries (eds), Le Travail du cuir de la
préhistoire à nos jours. Antibes, 231–49.
Borgard, P., Brun, J-P., Leguilloux, M., and Tuffreau-Libre, M. (2003). ‘Le produzioni
artigianali a Pompei: Ricerche condotte dal Centre Jean Berard’, Rivista di studi
pompeiani 14: 9–29.
Bowman, A. K., and Wilson, A. I. (2011). Settlement, Urbanization, and Population
(Oxford Studies on the Roman Economy). Oxford.
Bradley, M. (2002). ‘ “It all comes out in the wash”: Looking Harder at the Roman
Fullonica’, JRA 15/1: 21–44.
Broekaert, W. (2011). ‘Partners in Business: Roman Merchants and the Potential
Advantages of being a Collegiatus’, Ancient Society 41: 221–56.
Broekaert, W., and Zuiderhoek, A. (2013). ‘Industries and Services,’ in P. Erdkamp
(ed.), The Cambridge Companion to Ancient Rome. Cambridge, 317–35.
Clarke, J. (2003). Art in the Lives of Ordinary Romans. Berkeley.
Cuomo, S. (2007). Technology and Culture in Greek and Roman Antiquity. Cambridge.
De Ligt, L. (2012). Peasants, Citizens and Soldiers: Studies in the Demographic History
of Roman Italy 225 BC–AD 100. Cambridge.
Ellis, S. J. R. (2004). ‘The Distribution of Bars at Pompeii: Archaeological, Spatial and
Viewshed Analyses’, JRA 17/1: 371–84.
Ellis, S. J. R. (2011). ‘The Rise and Reorganization of the Pompeian Salted Fish Industry’,
in S. Ellis (ed.), The Making of Pompeii: Studies in the History and Urban Development
of an Ancient Town (JRA Supplement 85). Portsmouth, RI, 59–88.
Esposito, A., and Sanidas, G. (eds) (2012). Quartiers artisanaux en Grèce ancienne:
Une perspective méditerranéenne. Villeneuve d’Ascq.
Finley, M. I. (1973). The Ancient Economy. London.
Flohr, M. (2007). ‘Nec quicquam ingenuum habere potest officina? Spatial Contexts of
Urban Production at Pompeii, AD 79,’ BABesch 82/1: 129–48.
Flohr, M. (2009). ‘The Social World of Roman Fullonicae’, in M. Driessen,
S. Heeren, J. Hendriks, F. Kemmers, and R. Visser (eds), TRAC 2008: Proceedings
of the Eighteenth Annual Theoretical Roman Archaeology Conference. Oxford,
173–86.
Flohr, M. (2012). ‘Working and Living under One Roof: Workshops in Pompeian
Atrium Houses’, in A. Anguissola (ed.), Privata Luxuria: Towards an Archaeology of
Intimacy: Pompeii and Beyond. Munich, 51–72.
Flohr, M. (2013a). The World of the Fullo: Work, Economy, and Society in Roman Italy
(Oxford Studies on the Roman Economy). Oxford.
18 Miko Flohr and Andrew Wilson
Flohr, M. (2013b). ‘The Textile Economy of Pompeii’, JRA 26/1: 53–78.
Frier, B. W., and Kehoe, D. (2007). ‘Law and Economic Institutions’, in Scheidel,
Morris, and Saller (2007), 113–42.
Graham, S. (2006). Ex Figlinis: The Network Dynamics of the Tiber Valley Brick
Industry in the Hinterland of Rome (BAR International Series 1468). Oxford.
Hawkins, C. (2012). ‘Manufacturing’, in W. Scheidel (ed.), The Cambridge Companion
to the Roman Economy. Cambridge, 175–94.
Hawkins, C. (2013). ‘Labour and Employment’, in P. Erdkamp (ed.), The Cambridge
Companion to Ancient Rome. Cambridge, 336–51.
Joshel, S. R. (1992). Work, Identity and Legal Status at Rome. London.
Kehoe, D. (2007). ‘The Early Roman Empire: Production’, in Scheidel, Morris, and
Saller (2007), 543–69.
Leguilloux, M. (2002). ‘Techniques et équipements de la tannerie romaine: L’Exemple
de l’officina coriaria de Pompéi’, in F. Audoin-Rouzeau and S. Beyries (eds), Le
Travail du cuir de la préhistoire à nos jours. Antibes, 267–82.
Leguilloux, M. (2004). Le Cuir et la pelleterie à l'époque romaine (Collection des
Hespérides). Paris.
Liu, J. (2013). ‘Professional Associations’, in P. Erdkamp (ed.), The Cambridge Companion
to Ancient Rome. Cambridge, 352–68.
Monteix, N. (2010). Les lieux de métier: Boutiques et ateliers d’Herculaneum. Rome.
Morley, N. (2007). ‘The Early Roman Empire: Distribution,’ in Scheidel, Morris, and
Saller (2007), 570–91.
Morris, I., Saller, R., and Scheidel, W. (2007). ‘Introduction’, in Scheidel, Morris, and
Saller (2007), 1–12.
Mouritsen, H. (2011). The Freedman in the Roman World. Cambridge.
Petersen, L. (2006). The Freedman in Roman Art and Art History. New York.
Polanyi, K. (1944). The Great Transformation: The Political and Economic Origins of
our Time. Boston.
Rostovtzeff, M. (1926). The Social and Economic History of the Roman Empire. Oxford.
Ruffini, G. (2008). Social Networks in Byzantine Egypt. Cambridge.
Scheidel, W. (2012). ‘Approaching the Roman Economy,’ in W. Scheidel (ed.), The
Cambridge Companion to the Roman Economy. Cambridge, 1–21.
Scheidel, W., Morris, I., and Saller, R. (eds) (2007). The Cambridge Economic History of
the Greco-Roman World. Cambridge.
Terpstra, T. (2013). Trading Communities in the Roman World: A Micro-Economic
and Institutional Perspective. New York.
Tran, N. (2007). ‘Le “Procès des foulons”: L’Occupation litigieuse d’un espace vicinal
par des artisans romains,,’ MEFRA 119/2: 597–611.
Van Nijf, O. M. (1997). The Civic World of Professional Associations in the Roman
East. Amsterdam.
Verboven, K. (2011). ‘Resident Aliens and Translocal Merchant Collegia in the Roman
Empire’, in O. Hekster and T. Kaizer (eds), Frontiers in the Roman World: Proceed-
ings of the Ninth Workshop of the International Network Impact of Empire (Durham,
16–19 April 2009). Leiden, 335–48.
Verboven, K. (2012). ‘Cité et réciprocité: Le Rôle des croyances culturelles dans
l’économie romaine’, Annales. Histoire, Sciences Sociales 67: 913–42.
 Introduction 19
Wallace-Hadrill, A. (1991). ‘Elites and Trade in the Roman Town,’ in J. Rich and
A. Wallace-Hadrill (eds), City and Country in the Ancient World. London, 241–72.
Wilson, A. I. (1999). ‘Commerce and Industry in Roman Sabratha’, Libyan Studies 30:
29–52.
Wilson, A. (2000). ‘The Water-Mills on the Janiculum’, Memoirs of the American
Academy at Rome, 45: 219–46.
Wilson, A. (2002a). ‘Machines, Power and the Ancient Economy’, JRS 92: 1–32.
Wilson, A. I. (2002b). ‘Urban Production in the Roman World: The View from North
Africa’, PBSR 70: 231–73.
Wilson, A. I. (2007). ‘Fish-Salting Workshops in Sabratha’, in L. Lagóstena, D. Bernal,
and A. Aréval (eds), Cetariae 2005: Salsas y Salazones de Pescado en Occidente
durante la Antigüedad. Actas del Congreso Internacional (Cádiz, 7–9 de noviembre
de 2005) (BAR International Series 1686). Oxford, 173–81.
Wilson, A. I. (2008). ‘Large-Scale Manufacturing, Standardization, and Trade’, in
J. P. Oleson (ed.), Handbook of Engineering and Technology in the Classical World.
Oxford, 393–417.
 I
Approaches
 1

Roman Craftsmen and Traders:

Towards an Intellectual History
Miko Flohr and Andrew Wilson

Although the study of everyday socio-economic processes, such as crafts and

trade, and of the people involved in them has occupied a marginal position
within classical studies from its beginnings until fairly recently, there is a long
history of scholarly research that has had a decisive, though often implicit,
impact on the way in which craftsmen and traders are approached by scholars
nowadays. It is self-evident that an understanding of this scholarly history is
essential in making sense of the way in which scholars have come to think
about Roman craftsmen and traders in the way they do. Yet, thus far, the
intellectual past of scholarly discourse on this topic has received relatively little
attention. The present chapter therefore intends to sketch, in rough strokes, how
scholarly approaches to craftsmen and traders have developed since the early
nineteeth century. It does so by focusing on the contributions from Germany,
Britain, and the United States of America. These are three dominant classical
traditions that have played a key role in shaping and reshaping debates and
approaches. They have, however, done so in rather different ways, and at
different moments. In this chapter these three traditions are discussed together,
and contrasted where they diverge, and thus a core historical framework
emerges that is also of use in discussions of other research traditions, such as
those from Italy and France—both of which are treated in subsequent chapters.1
The question at stake is obviously one of intellectual history, and concerns
analysing the modern historical background against which ideas about Roman
craftsmen and traders emerged. The last decades of the twentieth century saw
a gradual increase of interest in the modern reception of the Greco-Roman
world among classical scholars, and, while most of this concerns those debates
and issues that were traditionally most central to classical studies, there is also

1
Salvaterra and Cristofori, Chapter 2, this volume; Brun, Chapter 3, this volume.
24 Miko Flohr and Andrew Wilson

a growing amount of published work on the lives and careers of the scholars
who played a role in discourse on crafts and trade in the Roman world. There is
a substantial, and growing, bibliography on key figures such as Theodor
Mommsen, Eduard Meyer, Mikhail Rostovtzeff, and Moses Finley.2 The
twentieth-century history of the debate about the ancient economy has also
been discussed by a variety of scholars, especially in recent years.3 The 2006
monograph by Jonathan Perry on the modern history of the Roman collegia
sheds light on the role of some key scholars in nineteenth- and twentieth-
century thinking on professional associations; for Germany, Dissen has recently
published a similar analysis, which focuses more on the later twentieth century.4
However, besides the work of Perry, there has been little reflection specifically
focusing on the historiography of Roman craftsmen and traders.
By definition, debates in classical studies evolve around the tension between
evidence and ideas, which means that the discovery of large quantities of new
data, the increased accessibility of data sets, or the emergence of new ways of
looking at the world, including political ideologies, tend to have a direct impact on
the way certain themes are discussed. As will become clear, both the changing
nature of the evidence available, and changing modern political and ideological
discourses, have played key roles in approaches to Roman craftsmen and traders.
Yet people matter as well: scholarly networks and scholarly competition may
foster, or curb, debate, and may privilege certain approaches over others. In what
follows, the development of discourse on Roman craftsmen and traders will be
analysed with a special emphasis both on the broader developments in evidence
and ideology, and on the role played by specific people or specific networks of
people in fostering the debate. The focus will be on five subsequent developments
that shaped discourse on the subject between the mid-nineteenth and the early
twenty-first centuries. These developments should not be seen as ‘phases’: to
some extent they overlap chronologically and developments that emerge in one
period tend to continue subsequently at varying levels of intensity.

BEFORE THE M OMMSEN ERA

There is no doubt that the second half of the nineteenth century marks a
major watershed in the study of Roman craftsmen and traders. The publica-
tion of the large epigraphic corpora in the late nineteenth century, the

2
Mommsen: Christ (1972: 84–118); Rebenich (2002); A. Overbeek (2005); Perry (2006:
23–60); Meyer: Christ (1972: 286–333); Calder and Demandt (1990); Rostovtzeff: Christ (1972:
334–49); Wes (1990); Finley: Morris (1999); Harris (2013).
3
e.g. Morley (2004: 33–50); Schneider (2009); Wagner-Hasel (2009).
4
Perry (2006); Dissen (2009). For French discourse on the topic, see also Tran (2001).
 Towards an Intellectual History 25

increasing amounts of archaeological evidence for crafts and trade in the

Roman world, and the encyclopaedia frenzy completely changed the scholarly
playing field, as well as the rules of the analytical game. Before, roughly,
1850, very few well-excavated shops and workshops from the Roman world
were known to a wide audience, and inscriptions were not easy to find,
even if they had been published. This largely restricted scholars to literary
sources and legal texts. Moreover, as scholars focusing on the history of
antiquity were, until well into the nineteenth century, mainly occupying
themselves with the works of Latin and Greek historians, there were few
incentives for them to focus on craftsmen and traders.5 In consequence,
publications focusing directly on craftsmen and traders in a more than
passing manner were rare, and stood on their own rather than being part
of a debate. Such accounts were often also not primarily interested in questions
of social or economic history. For example, one of the very few works from
before 1850 that were occasionally referred to by later scholars is Johann
August Ernesti’s De Negotiatoribus Romanis (1737), which is essentially a
lexicographical exercise focusing on what, actually, was understood under the
term negotiator by the Roman authors (especially Cicero), and only sporadic-
ally alludes to the socio-economic roles of negotiatores.6 In a rather similar way,
Johann Christian Schöttgen discussed the history of fulling in the ancient world
in his Antiquitates Trituræ et Fulloniæ (1727).7
The paradox is that, while classical scholars of the eighteenth century were
not interested in detailed study of ancient economies, scholars in other fields
were already developing ideas that would have a huge impact on later debates.
Highly relevant were the eighteenth-century developments in political econ-
omy, and particularly the thinking of people such as Hume, Steuart, and
Smith.8 Whereas the classical political economists of this period, unlike the
social theorists of the nineteenth century, did not see antiquity as fundamen-
tally different from their own world, they did observe that trade and industry
were less fully developed in antiquity than they were in their own time, and
debated why this was the case, blaming low demand and, to a lesser degree,
underdeveloped technology.9 For instance, Hume claimed that commerce was
mainly based on exchange of commodities that were tied to certain types of
soil and climate, and emphasized that no ancient author ascribed the growth
of a city to the establishment of industries.10 Steuart believed that the

5
Cf. Momigliano (1950: 291–2).
6
Ernesti (1737). Cf., e.g. Hatzfeld (1919: 1), who erroneously dates Ernesti’s work to 1802.
7
Schöttgen (1727). Cf., e.g. W. Smith (1848: s.v. fullo).
8
e.g. Hume (1752); Steuart (1767); A. Smith (1776) and Lectures on Jurisprudence. See Meek
et al. (1978).
9
Morley (1998: 109–11). Morley notes that the classical political economists stayed away
from the issue of economic mentality.
10
Hume (1752: 207). Cf. Morley (1998: 109). See also Finley (1985: 137).
26 Miko Flohr and Andrew Wilson

widespread use of slave labour, and the corresponding lack of economic

opportunities for the free population, kept demand low.11
A crucial development in this debate came when, in the nineteenth century,
scholars started to distinguish more sharply between ‘pre-industrial’ and
‘industrial’ societies.12 This had a profound effect on the perception of the
Roman world. Adam Smith had seen the development of mankind as consist-
ing of four stages, progressing from hunting, to pastoralism, to agriculture,
and, finally, to commerce.13 In this model, the classical world was interpreted
as developing from the agricultural to the commercial stage, though, with the
fall of Rome, the commercial system collapsed, only to emerge again in the
medieval period.14 Yet, by the late 1840s, Karl Marx had developed an
adapted, and more teleological, version of the four-stages theory, sketching,
for the West, a historical progress from primitive communism (tribalism)
through the slave society to feudalism and capitalism.15 In this model, the
ancient world fitted neatly into the second stage. Contrary to Smith, Marx thus
put the Roman world in a structurally different category from the early
modern world. To some extent, it might be argued that the twentieth-century
debates about the Roman economy developed in the tension between the
views of Smith and Marx.

BUILDING A CAN ON

These grand theories did not have a very deep impact on the actual study of
Roman crafts and trade before the 1890s. Nevertheless, the period between the
mid-nineteenth century and the start of the First World War in 1914 did see
some crucial developments in classical scholarship that would fundamentally
change the dynamics of the discourse: for the first time, classical scholars
engaged with the evidence for everyday economic life in a systematic way.
Three developments deserve to be highlighted.
In the first place, there was a revolution in the archaeology of crafts and
trade. An exemplary role in this was undoubtedly played by Pompeii, where,

11 12
Steuart (1767: ii. 35–40). Cf. Morley (1998: 110). Morley (1998: 110).
13
See Brewer (2008); cf. Finley (1977: 313).
14
Brewer (2008: 15–17). Smith discusses antiquity only in passing in the Wealth of Nations,
but notes taken at his Lectures on Jurisprudence give a good idea of his thoughts, esp. 24 February
1763. Cf. Meek et al. (1978: 223–8).
15
Cf. Hobsbawm (1964: 19). Marx sketches the road from primitive communism to feudal-
ism in Die Deutsche Ideologie, which was written in 1845 but not published until after his death.
Cf. Marx and Engels (1932: 11–15). On the fundamental differences between ancient and
medieval economic life, see also Pre-Capitalist Economic Formations (1857); see the edition by
Hobsbawm (1964: 72–8).
 Towards an Intellectual History 27

from 1860 onwards, the speed of excavation increased dramatically, and

numerous workshops, of several kinds, were excavated.16 Well-documented
and regularly updated guides in both the English and the German language
began to circulate, such as Gell’s Pompeiana and Dyer’s Pompeii: Its History,
Buildings and Antiquities.17 Johann Overbeck’s Pompeji in seinen Gebäuden,
Alterthümern und Kunstwerken was revised and expanded three times
between its first appearance in 1856 and 1884, including new discoveries
every time.18 Its final edition would be the basis for August Mau’s Pompeii in
Leben und Kunst, which also appeared in English and would be the leading
handbook on Pompeii for most of the twentieth century.19 These guides
discussed the discovered remains of workshops in considerable detail, so that
their existence became known to a wide scholarly audience. Moreover, the
establishment of the Notizie degli Scavi di antichità in Italy in 1876, and the
Jahrbuch des Deutschen Instituts in the same year, ensured that knowledge
of new discoveries in Pompeii and elsewhere spread quickly through the
scholarly community. What happened at Pompeii also happened at other
sites, such as Delos, Timgad, and Ostia, although slightly later and with less
direct impact on the overall debate.20 Moreover, with an increasing proportion
of archaeological activity focused on more everyday (urban) environments,
archaeologists also began to develop a more explicit interest in the archaeology
of crafts and manufacturing, as is attested by, for example, the groundbreaking
work on centres of pottery production by Dragendorff, and by the work on a
number of production sites, such as Arezzo, Rheinzabern, Lezoux, and La
Graufesenque.21
Second, there was the appearance of large corpora of evidence. An increased
frustration caused by the poor accessibility of Latin epigraphy led in the late
1840s and early 1850s to the establishment of the Corpus Inscriptionum
Latinarum (CIL), which under the leadership of Theodor Mommsen pub-
lished, in a few decades, and in a systematic manner, an enormous quantity of
inscriptions from all over the Roman world.22 More than the archaeological
discoveries at Pompeii, the CIL, quickly, completely, and definitively, changed
the scholarly field: while Ernesti, in his work on the negotiatores, had to make
do with a few literary references, the CIL published no fewer than 175 texts

16
For a discussion of the history of excavations at Pompeii, see Pesando and Guidobaldi
(2006: 25–8). Between 1860 and 1914, about 50% of the area currently excavated was unearthed.
17
Gell (1834); Dyer (1867, 1868, 1871).
18
J. Overbeck (1856, 1866, 1875); Overbeck and Mau (1884).
19
Mau (1900; English edn 1899).
20
Delos was excavated from 1873 onwards; cf. Chamonard (1922); Bruneau and Ducat (1966:
25). The excavations of Timgad (‘la Pompéi africaine’) began in 1880; cf. Ballu (1897: 3). At
Ostia, true excavations began in 1907, under Vaglieri; cf. Olivanti (2001).
21
Dragendorff (1895); Rheinzabern: Reubel (1912). See also Dechelette (1904).
22
The first thirteen volumes had appeared, at least partially, before the end of the nineteenth
century. For an overview, see <cil.bbaw.de> (accessed 3 October 2013).
28 Miko Flohr and Andrew Wilson

featuring the occupation.23 The relevance of the CIL became clear from the
upsurge of interest in the study of professional associations that emerged in
the late nineteenth century. Scholars such as Liebenam, Maué, and, particu-
larly, Waltzing noted the (potential) impact of the corpus on the debate.24
Waltzing’s classic multi-volume Étude historique des corporations professionels
can perhaps even be seen as a direct result of the publication of the first
volumes of the CIL.25 The CIL was by far the largest, but it was not the only
relevant data set made accessible through systematic publication, and some
efforts were specifically devoted to crafts and trade. Otto Jahn, who worked
closely with Mommsen in parts of his career, collected and published depic-
tions of crafts and trade on reliefs and wall paintings.26 After the turn of the
century, there was Espérandieu’s Receuil générale des bas reliefs de la Gaule
romaine, including many reliefs with scenes of everyday economic life.27
Finally, and perhaps most importantly for the scholarly understanding of crafts
and trade in the late nineteenth century, there were the handbooks, dictionaries,
and encyclopaedias focusing on the ancient world. The lemmas and entries in
these works quickly became standard points of reference in discussions of specific
crafts and trades and remained so throughout the twentieth century. Obviously,
the late nineteenth century saw Georg Wissowa initiating the extended version of
Pauly’s Realencyclopädie der classischen Altertumswissenschaft, but, as many
relevant entries in this encyclopaedia appeared only in the course of the twentieth
century, other works were more influential on topics related to crafts and
manufacturing. These include general works such as Smith’s Dictionary, Darem-
berg and Saglio’s Dictionnaire des antiquités grecs et romaines, and De Ruggiero’s
Dizionario epigrafico di antichità romane.28 More specifically focusing on aspects
of economic life was Hugo Blümner’s Technologie und Terminologie der Gewerbe,
which gave detailed descriptions of all kinds of production processes and linked
these to the material and textual evidence available.29 The second volume of Das
Privatleben der Römer by Joachim Marquardt contained comparable encyclopae-
dic information on crafts and trade.30 It may be noted that both these two works
were produced in the circles around Theodor Mommsen: Blümner, whose
dissertation was a monograph on crafts and trade, was a student of Otto Jahn,
while Marquardt edited together with Mommsen the Handbuch der Römischen
Alterthümer, of which the volumes on private life were part.31 Generally, the
dictionaries and encyclopaedias of this period used almost identical sets of

23
See <db.edcs.edu>, search on negotiator in CIL (accessed 1 March 2015).
24
Liebenam (1890: v); Maué (1886); Waltzing (1892: 5). Indeed, Mommsen’s work on the
collegia, which was published before the CIL project had begun, seems to have confirmed his
determination for it. Mommsen (1843). Cf. Perry (2006: 29).
25 26
Waltzing (1895, 1896, 1900). Cf. Perry (2006: 64). Jahn (1861, 1868).
27
Espérandieu (1907–).
28
W. Smith (1848); Daremberg and Saglio (1873–); De Ruggiero (1895–).
29 30 31
Blümner (1875–84). Marquardt (1882). Blümner (1869).
 Towards an Intellectual History 29

evidence, referring to the same passages in the literary authors, and to the same
legal texts, using (if any) the same subset of inscriptions, and the same examples
from the archaeological record, most of which came from Pompeii. This resulted
in a rather sharply defined canon of evidence for crafts and trade, from which it
would be hard to escape for much of the twentieth century: scholars often
approached the encyclopaedias and handbooks as received knowledge rather
than as the artificial constructs they often were.32

FROM RODBERTUS TO ROSTOVTZEFF

While classical scholars in Germany were busy cataloguing the evidence, debates
among economic theorists continued, and increasingly followed Marx in empha-
sizing the fundamental difference between antiquity and the modern world. The
(socialist) economist Johann Karl Rodbertus, discussing the tax system of the
Roman Empire, depicted the Roman economy as an Oikenwirtschaft—a system
organized around households that were, in principle, autarkic, and in which trade
and commercial manufacturing played no meaningful role.33 The model of
Rodbertus would be picked up and elaborated by Karl Bücher, who used it in
his version of the Wirtschaftsstufentheorie, in which he envisaged a linear progress
from the ancient Oikenwirtschaft through a Stadtwirtschaft—the medieval
period—to the modern Volkswirtschaft.34 It was Bücher’s model that would,
finally, provoke a response from the German community of classical scholars:
in a famous paper given at the 1895 meeting of Geman historians, Eduard Meyer
argued that Rodbertus and Bücher had completely misunderstood the structure of
the Greek and Roman economies, and that trade and industry played a defining
role throughout antiquity.35 As has recently been emphasized by Helmut Schnei-
der, Meyer’s response cannot be seen apart from contemporary scholarly politics:
in a field that felt threatened in its relevance to the modern world, Meyer wanted
to show the continued significance of studying the ancient world and its economy
for the present day.36 Meyer was not alone: Julius Beloch also responded with an
argument that highlighted the widespread existence of large-scale industry from
archaic Greece onwards.37 Studying the economic history of Rome, Karl Hoff-
meister attempted to reconcile Bücher’s three stages with Meyer’s optimism,
leading to the argument that Rome started as an Oikenwirtschaft, quickly devel-
oped into a Stadtwirtschaft and reached, already in the late Republic, the level of a
Volkswirtschaft.38

32
See, e.g. Bradley’s critique (2002: 21–2) of ‘dictionary fulling’.
33 34 35
Rodbertus (1865: 345–6). Bücher (1893). Meyer (1895).
36 37
Schneider (2009: 349–53). Cf. Meyer (1895: 1). Beloch (1899).
38
Hoffmeister (1899: 93).
30 Miko Flohr and Andrew Wilson

Thus, in the German debate about Rodbertus’s model of the Oikenwirtschaft

(often slightly misleadingly reduced to a ‘Bücher–Meyer controversy’), the
classical scholars were consistently and significantly more optimistic about the
structure and performance of the ancient economy than the economic theorists.39
Generally, leading German classical scholars of the late nineteenth and early
twentieth centuries thus tended to believe that trade and manufacturing, through-
out antiquity, both socially and economically played an important role. Meyer
indeed refers to the existence of centres of trade and industry ‘everywhere’ in the
archaic Aegean.40 For the Roman Empire he envisaged busy transport and
widespread commercial exchange.41 Beloch, though arguing that small-scale
manufacturing was the norm in Italy, claimed that large-scale factories were
common in the eastern provinces.42 Hoffmeister described second-century BC
Rome as a Weltverkehrsstadt that had ‘thousands of shops’ where people could
buy products from all known regions.43 However, none of these works made use
of the archaeological and epigraphic evidence that was being published in the late
nineteenth century, and Meyer and Beloch dealt mostly with the Greek rather
than the Roman world. As Schneider has observed, the responses to Bücher by
Meyer and Beloch did not contain a lot of new research—rather, they put
interpretations that had long been known into a new ideological framework.44
It would be well into the twentieth century before the possibilities of the newly
available evidence began to be seriously explored.
Nevertheless, there is no doubt that the determined responses of influential
Altertumswissenschaftler such as Meyer and Beloch to the ideas of Bücher and
Rodbertus had an enormous impact on the way in which classical scholars
came to think about Greek and Roman economies. Moreover, Meyer in
particular seems to have encouraged people to engage with the archaeological
and epigraphic evidence for crafts and trade. One of the most productive
scholars of the first decades of the twentieth century was Herman Gummerus,
who did a doctorate with Meyer on Roman agriculture and subsequently wrote
a series of articles on Roman crafts and trade, including the long discussion of
industry and trade in the Realenzyklopädie.45 Moreover, Meyer also seems to
have had a direct impact on the thought and careers of two leading figures in
Roman economic history of the first half of the twentieth century: Tenney
Frank and Mikhail Rostovtzeff. Frank, educated as a philologist more than an
ancient historian, met Meyer when he was in the USA in 1909, and subse-
quently spent parts of a sabbatical in Berlin with Meyer in 1910–11.46 Shortly

39
Cf. Schneider (2009: 354). On the Bücher–Meyer controversy, see also Reibig (2001).
40 41 42
Meyer (1910: 113). Meyer (1910: 144). Beloch (1899: 25–6).
43 44
Hoffmeister (1899: 39). Schneider (2009: 353).
45
Gummerus (1906, 1913, 1915, 1916). Wissowa, the editor of the Realenzyklopädie, be-
longed to Meyer’s inner academic circle; cf. Audring (2000). His student Georg Kuhn wrote a
dissertation (1910) on Roman craftsmen that would be frequently cited by Gummerus (1916).
46
Chambers (1990: 115).
 Towards an Intellectual History 31

afterwards, he began publishing his first articles on aspects of the Roman

economy, research that would culminate in his An Economic History of Rome
(1920).47 It was also Meyer who, together with Wilamowitz, asked Rostovtzeff
to write his social and economic histories of the Hellenistic and Roman
worlds.48
Both Frank and Rostovtzeff made extensive use of the archaeological and
epigraphic evidence that had been published in the preceding decades, and
they would integrate a whole paradigm of observations into an attractive and
coherent narrative, which would emphasize the social and economic import-
ance of urban crafts and trade, though Frank at least seems to have regarded
Meyer’s views as exaggerated.49 Frank, in his chapters on industry under the
early empire, makes a case for the existence of monopolistic production
concentrated in large factories or specialized centres of mass production in
the manufacturing of everyday consumer goods such as red-glazed pottery,
glass, bricks, iron, and bronzes.50 Luxury products, on the other hand, such as
gold and gems, were produced by specialist craftsmen on a small scale.51
A lengthy case study of Pompeii highlights the role of commerce and industry
in the town, and argues that Pompeii specialized in the production of garum
and textiles, importing iron and bronze ware from the nearby production
centres of Puteoli and Capua.52 Rostovtzeff put more emphasis on commerce,
and made the important link between economic basis and urban growth,
arguing that, in the second century AD, the many flourishing urban centres
throughout the empire derived their wealth from their role in ‘inter-provincial
commerce’ in ‘articles of prime necessity’, not luxuries.53 In manufacturing,
Rostovtzeff sees in this period an empire-wide emergence of specialized,
regional centres of production, which catered for local and regional markets,
and deprived Italy of its export industries.54 In larger centres, Rostovtzeff even
envisaged a ‘development towards capitalistic mass-production’.55
While Rostovtzeff ’s book would eventually end up as the key reference for
the so-called modernist school of thought, the role of Frank in shaping the
debate from the 1910s onwards should not be underestimated, and their two
main works do not completely overlap: Frank’s focus was primarily on the
Republic and the first century of our era, while Rostovtzeff, in his monograph,
was more interested in the second and third centuries. Moreover, Frank

47
Frank (1910, 1914, 1918, 1920).
48
Christ (1972: 340). Rostovtzeff (1926) does not refer to this request in the preface to
his book.
49
Frank (1927: 219), referring to the debate between Rodbertus, Bücher, and Meyer: ‘the
advocates of both views have gone to untenable extremes.’
50 51
Frank (1927: 220–38). Frank (1927: 241–4).
52
Frank (1927: 245–67). The Pompeian case study was reproduced almost verbatim from
Frank (1918).
53 54
Rostovtzeff (1926: 148). Rostovtzeff (1926: 161–9).
55
Rostovtzeff (1926: 165).
32 Miko Flohr and Andrew Wilson

continued his study of crafts and trade well after the 1920s, when Rostovtzeff
had long moved on to studying the Hellenistic world. Still used nowadays is
the Economic Survey of Ancient Rome (1933–40), which was initiated, edited,
and partially written by Frank. The five volumes contain detailed descriptions
that outlined the evidence of trade and production for the individual parts of
the empire. Of particular relevance was the volume by Allan Chester Johnson
on the economy of Roman Egypt, which included a long section on commerce
and industry in the papyri—it was one of the first discussions of its kind in this
emerging field.56 The final volume, written by Frank, was completed after his
death in 1939 by his students Evelyn Holst Clift and Helen Jefferson Loane;
the latter also wrote a monograph on commerce and industry in the city of
Rome, which was not included in the series but written in the same period, and
shows clear influence of Frank’s thinking.57
The circle around Frank and Rostovtzeff dominated the debate from the 1910s
to the 1930s, which seems to have been a rather American debate: most other
scholars who published on these topics between the wars were also operating
from the east coast regions of the United States, and, in general, they subscribed
to the optimistic views that dominated the field. Ethel Hampson Brewster wrote a
thesis on the depiction of craftsmen and traders in Roman satiric verse.58 Louis
C. West published extensively on trade.59 There seems to have been less interest
in studying Roman crafts and trade in Germany or Britain; exceptions included
the monograph of Martin Percival Charlesworth on Roman trade routes and Eric
Herbert Warmington’s book on the Indo-Roman trade.60 At least for Germany,
the lack of debate can be partially explained by the consequences of the First
World War, and this is not exemplified better than by Rostovtzeff himself, who
was predominantly oriented towards Germany until the war—in 1905 he was even
considered for a position in Halle—but spent his post-war life in exile in America,
apparently without maintaining active ties with many German scholars.61

THE S UBSTANTIVIST DECADES

Frank’s death in 1939, and the completion of his Economic Survey a year later,
symbolized the end of an era. While the optimistic reading of Rome’s

56
Johnson (1936: 331–88). Rostovtzeff, of course, also made use of papyrological evidence,
but mentioned crafts and trade in Egypt only in passing.
57
Frank (1940); Loane (1938).
58
Brewster (1917). See the rather negative review by Wright (1918). See also Brewster (1927,
1931); Wright (1917).
59 60
West (1917, 1924, 1932, 1939). Charlesworth (1924); Warmington (1928).
61
On Rostovtzeff and Halle, see Audring (2000: 195–9): letter from Wissowa to Meyer, dated
29 December 1905, and Meyer’s immediate response dated 30 December 1905.
 Towards an Intellectual History 33

economic history would remain dominant among classical scholars in the

decades immediately following the Second World War, and would pervade
publications of newly discovered or studied data sets, a competing school of
thought was already starting to emerge in the late 1930s, and it was consid-
erably more sceptical about the economic history of the Roman world than
Frank and Rostovtzeff had been. This school of thought would eventually
become associated primarily with Moses Finley, and his The Ancient Economy
would become the standard reference.62 However, it is important not to
underestimate the key role of Arnold Hugh Martin Jones, an Oxford-trained
scholar who took up the chair of Ancient History at London (1946) and
Cambridge (1951).63
Indeed, as early as 1940, Jones discussed the economy of Greco-Roman cities
in terms not fundamentally different from what Finley would write more than
thirty years later.64 He sketched how cities performed a role as market place for
the surrounding countryside, emphasizing that the low purchasing power of
the rural community meant that the role of industry was ‘neglegible’. Long-
distance trade was restricted to luxury goods, and large-scale export-oriented
industry did not play an important role in the ancient world. Crucially, Jones
states: ‘Both trade and industry were in fact dependent upon a rich urban class,
which cannot itself have derived any large proportion of its total wealth from
these activities.’65
Instead, urban elites derived their wealth from the ownership of land.
Throughout antiquity, cities were, Jones argued, ‘economically parasitic on
the countryside’.66 Fifteen years later, in 1955, Jones published a more detailed
study focusing on the economic life of Roman towns, which conveyed a similar
message. In this article, Jones emphasized that trade and industry contributed
very little to urban tax incomes, that traders and manufacturers were poor, and
that city governments were not interested in fostering their activities; he
emphasizes that collegia were primarily social organizations that were not
involved in urban government or engaged in economic activities.67 As a source
of wealth, Jones concluded that ‘commerce could not compete with land’.68
Unfortunately, Jones tended to avoid direct debate with others: while he was
obviously familiar with the work of Rostovtzeff and Frank, he did not explicitly
engage with it.69 Similarly, while he may have known the ideas of Bücher and
Weber about the economic basis of ancient cities, he does not refer to them.

62 63 64
Finley (1973). On Jones, see Meiggs (1970). Jones (1940: 259–69).
65 66
Jones (1940: 263). Jones (1940: 268).
67
Jones (1955: 163, 173–5). The collatio lustralis, a tax paid by traders, and its role in
understanding the socio-economic position of craftsmen had already been discussed by Jones
in his inaugural address in London: Jones (1948: 11–12).
68
Jones (1955: 192).
69
Both Frank and Rostovtzeff feature in the bibliography of Jones (1940). On Jones’s style of
referencing, see Brunt (2004).
34 Miko Flohr and Andrew Wilson

The ideas of Moses Finley about the ancient economy are so well known that
they do not need to be introduced at great length, but it may be useful to
emphasize once again that Finley never specifically focused on the Roman
economy, and had a background in Greek history.70 Moreover, while The Ancient
Economy was published only in 1973, Finley seems already to have developed
some of his core ideas in the late 1940s and the early 1950s, before leaving the
USA, under the influence of Polanyi.71 However, Jones was also a great influence:
Finley’s 1965 article on technical innovation and economic progress in the
ancient world clearly echoes Jones’s work on ancient cities, arguing: ‘Even in
the Roman Empire, the quantitative contribution of trade and manufacture was
tiny, their social position low, their future without interest.’72
This, according to Finley, made it unattractive to take risks in investment in
manufacturing, and thus is one of the reasons why there was not an awful lot
of technical development in the ancient world. Contrary to Jones, Finley did
also engage directly, and often combatively, with his opponents. Discussing
the terra sigillata industry, and Rostovtzeff ’s idea that a decentralization of
manufacturing prevented emerging industrial capitalism in Italy from con-
tinuing, Finley made no secret of his feelings, and called the theory an
‘anachronistic burlesque of the affluent society’: according to Finley, the
decline of Arezzo and the related emergence of production centres in Gaul
and Germany was a meaningless event in a minor trade, without wider
economic implications.73 As far as urban craftsmen and traders are concerned,
the core of Finleyan thought already seems to have been present in his 1965
article, and The Ancient Economy did not dramatically alter the picture.74
However, what Finley did add in the book was a link to the German debate of
the late nineteenth century through the work of Max Weber and his idea that
the ancient city was a centre of consumption, an idea that Weber had taken
from Bücher and Rodbertus.75 This link would be further elaborated in a 1977
article on the historiography of the ancient city, where Finley first explicitly
coins the term ‘consumer city’, and in the ‘Further Thoughts’ added to the
second edition of the Ancient Economy.76
While a lot has been written in recent years about the debate between
‘modernists’ and ‘primitivists’ that followed the publication of Finley’s mono-
graph, it may be useful to reiterate the point made by Saller in 2002 that the
debate has tended to overemphasize the differences between Rostovtzeff and
Frank, on the one side, and Finley and Jones, on the other.77 Both ‘camps’
seem diametrically opposed when it came to the general importance of supra-

70 71
Finley (1973). See esp. Morris (1999).
72
Finley (1965: 40). The accompanying footnote refers to Jones (1955).
73 74
Finley (1965: 42). Finley (1973: 125; cf. 138). Cf. Wagner-Hasel (2009: 178–80).
75 76
Finley (1973: 125). Finley (1977; 1985: esp. 191–6).
77
Saller (2002: esp. 252–6).
 Towards an Intellectual History 35

regional trade and (export-oriented) manufacturing in the urban economy as

a whole. However, Rostovtzeff and Frank appear to have had more nuanced
views about the average Roman city than their enthusiasm in highlighting
trading centres suggest.78 Finley, in turn, may have had a more nuanced idea
about, especially, port cities than The Ancient Economy reveals.79 A second
point of apparent difference concerns the role of commerce as a vehicle of
social mobility: Frank and Rostovtzeff were much more open to the idea of
people gaining socio-economic prominence through their involvement in
crafts and trade than Jones and Finley, who saw any scenario other than
landowning elites as highly exceptional.80 Again, however, Frank and Ros-
tovtzeff did acknowledge the central role of landed property in the socio-
economic status of elites, and Finley did acknowledge that people got wealthy
from commerce—just not very many.81 Moreover, both camps believed that
the social position of craftsmen and traders was influenced negatively by the
low status of their occupations.82 Thus, while there were fundamental points
of disagreement, it is, on the issue of urban craftsmen and traders, crucial not
to overstate the differences between the two camps: part of it comes down to a
matter of style and emphasis.

AFTER JONES AN D FINLEY

The ideas of Jones and Finley received a mixed response among classical
scholars. Despite Hopkins’s repeated claim that Finley’s model of the ancient
economy presented a ‘new orthodoxy’, this was the case only to a very limited
extent.83 As has been observed by one of us, material specialists who worked
with the evidence of ancient economic life on an everyday basis have often
been unable to connect their data to the model outlined in The Ancient
Economy.84 Partially, this was due to the fact that Finley assigned little, if
any, interpretative value to documentary and material evidence; in doing so,
he failed to provide archaeologists, epigraphists, and papyrologists with an

78
See, e.g. Rostovtzeff ’s discussion of the economic basis of cities, culminating in the
statement that ‘to move large masses of foodstuffs by the land-roads was . . . beyond the resources
of smaller and poorer cities’, which essentially means that a large majority of cities had to live off
their hinterland. Rostovtzeff (1926: 133–8). Cf. Saller (2002: 255).
79
Finley (1973: 130–1).
80
Rostovtzeff (1926: 161): ‘commerce provided the main sources of wealth in the Roman
empire.’ Finley (1973: 89): ‘the land was the chief source of wealth.’ Cf. Saller (2002: 256).
81
Rostovtzeff (1926: 142–5): many senators were still primarily landowners; wealth aquired
by commerce was invested in land; Finley (1973: 59): ‘of course there were exceptions’.
82 83
Frank (1927: 274); Finley (1973: 41–61). Hopkins (1978: 47; 1983: pp. xi–xii).
84
Wilson (2002: 231).
36 Miko Flohr and Andrew Wilson

interpretative framework they could use to make sense of their observations.85

This was in sharp contrast with Rostovtzeff and Frank, who had highlighted
the value of epigraphy, archaeology, and papyrology throughout their work.
Essentially, substantivist thinking along the lines of Finley and Jones found
most fertile ground among generalists, and even there it was not accepted
without qualification.86 Thus, rather than establishing an orthodoxy, the ideas
of Jones and Finley sparked debate, especially after the appearance of The
Ancient Economy, and, for the first time, the epicentre of that debate was in
Britain, where scholars until after the Second World War had remained
mostly indifferent to the study of economic life in the Roman world.87
Evidence for urban crafts and traders eventually came to play a key role in
debates between ‘modernists’ and ‘primitivists’ of the 1980s and 1990s, par-
ticularly after Finley’s defiant defence of his ideas in the second edition of The
Ancient Economy.88 The issue in the debate that is most relevant in the present
context is related to the model of the consumer city.
Much more than research in other parts of the world, anglophone
scholarship on urban craft production has been dominated by the consumer
city debate, which ran from the late 1970s until the early 2000s. It started
with Finley’s 1977 article on the ancient city, and the response, a year later,
by Keith Hopkins, who mostly embraced the model, though assigning a
larger role to trade and urban manufacturing than Finley had done.89
Finley’s view that urban manufacturing was essentially petty was a plank
of his model of the ancient city as consumer city (or, perhaps, his faith in
that model required him to believe that urban craft production had to be
insignificant). This involved the assumptions that urban workshops were
small and numerically insignificant, that the status of craftsmen was low,
that demand for their products was low or, even if it was not, came
principally from elite consumption, and that revenue from urban production
played no real role in elite portfolios. His views provoked reactions from
archaeologists in particular, and, in the debate over the consumer city that
raged in the 1980s and 1990s, the scale, nature, and status of urban crafts
were a recurrent feature. Initially, archaeologists tended to point to a variety
of workshops in different towns to emphasize the vitality of urban produc-
tion, while ancient historians pointed out that the numbers of identifiable
workshops for particular crafts in any one town were few, and minimized
the scale of the phenomenon. The textile industry at Pompeii was at the
centre of such debate—but, in focusing on the textile industry, the argument

85
Cf., e.g. Finley (1965: 41): ‘We are too often victims of that great curse of archaeology, the
indestructibility of pots.’
86
See esp. Hopkins (1978, 1980).
87
With some obvious exceptions: e.g. Charlesworth (1924) and Warmington (1928).
88 89
Finley (1985: esp. 191–6). Finley (1977); Hopkins (1978: esp. 75).
 Towards an Intellectual History 37

largely ignored other crafts attested at Pompeii and so failed to assess the
scale of the urban productive economy in toto.90
The most intense stage of the debate took place from the late 1980s
onwards, and was decisively shaped through a series of edited volumes that
were produced in the United Kingdom in the 1990s and focused specifically on
the relation between city and country.91 More than any other debate discussed
in this chapter, the consumer-city debate was predominantly a British affair:
while Philippe Leveau had fostered some debate on the issue in France in the
1980s, scholars from Germany and the United States only incidentally got
involved.92 Critics came up with a variety of responses to Finley. In the 1980s
there were scholars who rejected the consumer city model, but proposed an
alternative model instead, such as Leveau’s cité organisatrice and Donald
Engels’s service city.93 During the 1990s archaeologists were publishing
more urban workshops, but usually individually or in small groups, still
without really convincing the historians that what they had added up to
anything of structural significance in urban economies as a whole. Champions
of the consumer city model tried to counter its critics by rephrasing the model,
or by adding elements to it. C. R. Whittaker, for example, published in the
Journal of Roman Archaeology an article entitled ‘The Consumer City Revisit-
ed: The vicus and the City’, in which he argued that craft production was
confined to small towns, or vici, which he saw as essentially rural. This now
seems misguided, and may be due to the extensive publication of several small
towns in Britain—for example, Water Newton or Durobrivae—where the
ancient remains were not covered by modern habitation and the evidence
for craft production was well recognized in careful excavation. By contrast,
many more important ancient cities have remained inhabited, limiting the
picture available, and it is only since the early 1990s that the growth of urban
rescue archaeology outside Britain, in France, Italy, and Spain, has accumu-
lated a critical mass of information to change the picture.
By the early 2000s, studies of sites such as Sabratha and Timgad had begun
to show that there were cities with large numbers of workshops and to point
out the difficulties in the survival and recovery of evidence, and the problems
of sampling urban areas to assess the distribution and number of workshops,
and their overall contribution to the urban economy.94 The phenomenon of
urban production now began to look more significant, even if we might still be
wary of replacing the ‘consumer city’ model with the equally simplistic

90
Moeller (1976); Jongman (1988). See now also Flohr (2013).
91
Rich and Wallace-Hadrill (1991); Parkins (1997); Parkins and Smith (1998); Mattingly and
Salmon (2000). To a lesser extent also Cornell and Lomas (1995).
92
See esp. Leveau (1985). Obviously, scholars were aware of the model, and several individ-
uals ventured their opinions, but there were no debates going on.
93 94
Leveau (1984, 1985); Engels (1990). Wilson (1999, 2000, 2002).
38 Miko Flohr and Andrew Wilson

‘producer city’. In the face of this newly assembled evidence, some scholars,
like Paul Erdkamp, conceded that cities could easily make substantial money
from export-oriented manufacturing, as long as, in the end, all urban con-
sumption was based on revenues from landed property.95 More generally, a
combination of studies that did manage to synthesize evidence for zones of
urban production across broad areas of several cities—a new focus on the
long-distance Mediterranean connectivity advocated in Horden and Purcell’s
The Corrupting Sea—and general boredom with the consumer city debate led
to an adjustment of perspectives. It was increasingly recognized that the terms
of the consumer city debate had focused excessively on the idea that a city and
its territory formed a cellular self-sufficient unit; instead, a new appreciation of
the numbers, sizes, and populations of Roman cities emphasizes that cities
must be seen as part of broader economic networks; few Roman cities could be
supplied exclusively from their own hinterland, and archaeology suggests that
almost none was isolated in this way. More and more scholars came to reject
the use of all one-size-fits-all models for explaining the economic role of cities
in the Roman world, arguing that the variation between cities was simply too
big for such models to be meaningful.96
The primary relevance of the consumer city debate for the study of urban
crafts and trade is that it encouraged scholars to think harder about urban
economies, and to confront evidence with models and vice versa.97 Both
champions and sceptics of the model engaged actively with the material
remains of urban workshops and, to a lesser extent, with the epigraphic and
papyrological evidence. For example, Michael Fulford used the excavated
remains of cities in Roman Britain to argue that these cities indeed were
‘parasitical’ on the countryside.98 Andrew Wallace-Hadrill used the evidence
of Pompeii to show that Finley’s claim that elites were not involved in com-
merce and manufacturing was problematic.99 David Mattingly used survey
data from Leptiminus to sketch how the economy of this town was much more
complex than the consumer city model would allow for.100 One of the present
authors used material remains from Sabratha and Timgad, and a number of
other sites, to argue against the idea that these towns depended solely or
primarily on agriculture for their income.101 Such approaches fostered the
scholarly emancipation of material evidence for urban crafts, and encouraged
a greater focus on understanding the strengths and limitations of both
archaeological and textual evidence and how they might affect our ability to
construct a picture of ancient craft economies and to draw comparisons with
other periods. The assumed contrast between the supposedly productive

95 96
Erdkamp (2001). Mattingly (2000); Wilson (2002).
97 98
Cf. Wilson (2002: 232). Fulford (1982); Whittaker (1991: 113).
99 100
Wallace-Hadrill (1991). Mattingly (2000).
101
Wilson (1999, 2000, 2002).
 Towards an Intellectual History 39

medieval city and the unproductive classical city was at least in part a product
of not comparing like for like: classical written sources tend to treat economic
issues only anecdotally, and for the ancient world we lack the bulk of economic
documents—letters, accounts, and so on—that medieval historians are used to
working with: if one were to compare only the archaeological evidence, one
might get the impression of more craft production in Roman towns, and
sometimes on a larger scale, than in medieval European towns.102
The years after the publication of The Ancient Economy also saw increased
attention paid to craftsmen and traders beyond the issue of the consumer city.
John D’Arms responded to Finley’s ideas about the social standing of traders
with a monograph in which he made a case for the socio-economic independ-
ence of traders, and for the respectability of Roman negotiatores.103 In Germany,
a special role in the post-Finley discourse was played by the Münstersche Beiträge
zur Antiken Handelsgeschichte (MBAH), a journal founded in the early 1980s by
Hans-Joachim Drexhage and Wolfgang Habermann. The MBAH was an explicit
response to the emerging debate about the Ancient Economy, and was meant to
provide a platform for the study of evidence related to ancient economic life.104
While there is a clear emphasis on epigraphic and papyrological evidence, the
journal also published archaeological papers, particularly from provincial
Roman archaeologists working on terra sigillata.105 The journal addressed a
wide variety of themes. Drexhage himself published extensively on papyrological
and epigraphic evidence for individual trades and occupations—with a particular
emphasis on the food sector, including papers on cheese, beer, meat, pulses, and
garum.106 Other scholars published on other trades, and there are also papers
focusing on occupational terminology on a more general level.107 Another
recurring theme concerns transport and its infrastructure, including roads and
ports, but also ships, seafaring routes, and tolls.108 Trade itself is, of course, also a
central theme, and it is approached from a variety of directions—including case
studies of trade cities or trade routes as well as the study of the geographical
spread of certain objects, particularly pottery.109 The broad focus of the journal
and its evidence-oriented emphasis ensured that large amounts of material and
textual evidence for craftsmen and traders were integrated into the debate, and its
tables of contents give a good impression of the (changing) priorities within the

102 103 104

Wilson (2002). D’Arms (1981). MBAH 1 (1982), 1.
105
e.g. Raepsaet (1987); Strobel (1987); Berke (1988); Mees (1994); Stuppner (1994); Zaniers
(1994); Poblome et al. (1998); Fülle (2000).
106
H.-J. Drexhage (1993, 1996, 1997a, b, 2002a).
107
Wine: e.g. Ruffing (2001); brokers: Kudlien (1997); occupational terminology: e.g. Kneissl
(1983); Ruffing (2002); H.-J. Drexhage (2002b, 2004).
108
Roads: Schneider (1982); ports: Habermann (1982); Sidebotham (1986); Matei (1989);
Fellmeth (1991); tolls: H.-J. Drexhage (1982); Habermann (1990); H.-J. Drexhage (1994); ships:
Konen (2001); seafaring: Holtheide (1982); Bounegru (1984).
109
Trade, e.g. R. Drexhage (1982); H.-J. Drexhage (1983); Schleich (1983, 1984); Monfort
(1999); Henning (2001). On terra sigillata, see n. 99.
40 Miko Flohr and Andrew Wilson

field. Originating from the late 1970s Finley debate, the MBAH quickly came to
play a key role in German discourse on Roman urban crafts and trade—as its
successor, the Marburger Beiträge, still does.
The question remains, however, why it was precisely Finley’s The Ancient
Economy in 1973 that unleashed such a vibrant debate—after all, especially as far
as craftsmen and traders are concerned, the core arguments were already on the
table in the 1950s and 1960s. Jones and Finley did not publish their earlier ideas
in places that were widely read by classical scholars, but this can hardly have been
the only reason. Arguably, other developments in the field of classical studies
(in the widest sense of the word) played a much more fundamental role in the
impact of The Ancient Economy. In the first place, as will be discussed below,
classical studies had not remained completely indifferent to the upsurge of
interest in social history that characterized the later 1960s and the 1970s in the
humanities.110 Second, but equally importantly, the 1960s and 1970s saw a
remarkable increase in archaeological activity in Europe and Britain: many
urban sites from the Roman north-western provinces were thoroughly investi-
gated for the first time, while other sites were reinvestigated.111 While most
projects did not have a primarily economic focus, the nature of these cities, and
the absence of large urban monuments, meant that a lot of the evidence that was
dug up was better able to tell stories about manufacturing and trade than about
monumental architecture, and this evidence was increasingly being analysed.112
In a way, one should thus see The Ancient Economy, not as causing the debate,
but rather as providing the ideal occasion for a debate that was waiting to happen
anyway, though it obviously helped that Finley went in a radically opposite
direction from many of his contemporary scholars.

FROM ECON OMY TO S OCIETY AND CULTURE

Besides fuelling the vibrant debate on the ancient economy, the turn towards
social history in the last decades of the twentieth century also fostered a
significant scholarly engagement with the social and cultural aspects of
urban crafts and trade—raising many issues that had been almost completely
unexplored from the early twentieth century until the late 1960s.113

110
See esp. Treggiari (1975b); key example of this trend is Alföldy (1975).
111
Cf. Burnham and Wacher (1990: 3). In Germany, excavations at, e.g. Xanten were
intensified in the 1970s. Similar developments took place elsewhere in Europe. Examples include
the excavations at Alésia, Autun, and Kaiseraugst. See, e.g. Mangin (1981); Chardron-Picault and
Pernot (1999); Forschungen in Augst 1– (1977–).
112
See, e.g. Strong and Brown (1978).
113
Giving an overview of scholarship on Roman social history, Treggiari (1975b) lists only
the works by Loane (1938) and Maxey (1938) on occupations, and the work of Waltzing
 Towards an Intellectual History 41

An important strand of research that emerges in this period focuses on the

epigraphic and iconographic representation of craftsmen and traders in,
especially, the city of Rome. A key role in this debate was played by the
work of Susan Treggiari, who, in the late 1970s, published a series of articles
on occupations in Rome, focusing on domestic jobs in the household of Livia,
on female labour in the Roman urban economy, and on urban labour in
Rome.114 Treggiari’s work, with its strong focus on gender and (legal) status,
would set a model for several later approaches, and particularly for Sandra
Joshel’s important monograph on work, identity, and legal status in Rome,
which uses the epigraphic evidence from Rome to discuss the role of work and
legal status in constructing social identity.115 Treggiari also stood at the start of
a tradition that uses epigraphic and iconographic evidence to study the role of
women in urban crafts and trade; important here is also the work on textile
crafts done by Suzanne Dixon and Lena Larsson Lovén.116 Further, in her
important book on working women in Ostia, Natalie Kampen integrated the
work of Treggiari on female labour with iconographic representations of
women at work.117
The works of Kampen and Larsson Lovén, however, are also related to a
second development of the 1970s and 1980s: the rediscovery of the iconog-
raphy of crafts and trade. This development was part of a broader trend in
Roman art history to invest in understanding forms of art produced by or for
people from the strata below the imperial elite—a trend exemplified by, among
others, the work of Paul Zanker on freedman reliefs.118 The iconography of
crafts and trade had been largely neglected since Jahn and Gummerus, but
achieved new prominence through the work of Kampen and, particularly,
through that of Gerhard Zimmer, a student of Paul Zanker at Munich who
catalogued and (re)studied depictions of everyday work from Italy and Gaul.119
Zimmer’s Römische Berufsdarstellungen laid an indispensable foundation for
the more interpretative work on the iconographic representation of crafts that
emerged from the late 1980s onwards. Interestingly, the study of the iconog-
raphy of crafts and trade also attracted the attention of scholars specializing in
Roman art. Relying heavily on the work of both Zimmer and Kampen, Eve
d’Ambra made a detailed study of two tombs at Ostia that had been found with
reliefs depicting craftsmen at work.120 Lauren Hackworth Petersen made a

(1895–6) and De Robertis (1938) on collegia as dealing specifically with craftsmen and traders.
Though Treggiari omits several publications, this gives a good idea of the state of the field at
that point.
114
Treggiari (1975a, 1976, 1979, 1980).
115
Joshel (1992; on Treggiari, see esp. pp. 20–3).
116
Dixon (2001); Larsson Lovén (2002). See also Larsson Lovén, Chapter 9, this volume.
117
Kampen (1981, 1985). On Treggiari, see esp. Kampen (1981: 108 n. 5).
118
Zanker (1975). Fundamental was Bianchi Bandinelli (1967).
119 120
Zimmer (1982, 1985). D’Ambra (1988).
42 Miko Flohr and Andrew Wilson

detailed study of the tomb of Eurysaces as part of her work on ‘freedman art’.121
John Clarke devoted a chapter in his book on Roman popular art to the social
and cultural meaning of iconographic representations of everyday work.122
Arguably, this development contributed significantly to the integration of the
study of craftsmen and traders into mainstream classical scholarship.
A development of the 1980s and 1990s was a renewed interest in the study
of Roman professional associations, especially focusing on Roman Italy and
Asia Minor. For Italy, it all started with the prosopographical work of Royden
on magistrates of the professional collegia, and the work of John Patterson on
the social roles of collegia.123 Important, too, was the work of Beate Bollmann
on the scholae used by professional associations—it provided an essential link
between the epigraphic debate on collegia and the emerging archaeological
debate about the dynamics of urban space.124 After the turn of the millen-
nium, the debate was enriched by some studies on specific collegia, such as
DeLaine’s work on the Fabri Tignarii at Ostia, and the monograph of Jinyu Liu
on the collegia centonariorum.125 Ground-breaking for the Roman east was
the monograph by Onno van Nijf on the role of professional associations in
civic life, but it was quickly followed by two other monographs.126 Imogen
Dittmann-Schöne published a monograph on the evidence from Asia Minor,
while Carola Zimmermann made a comparative effort, studying the associ-
ations for the entire Greek-speaking east, including Egypt.127 While these
works focus on a variety of directions, they all have in common that the
primary emphasis is on the social role of professional associations—which is
also the issue about which the evidence is most explicit. The (disputed)
economic role of collegia has come to the fore only towards the end of the
first decade of the new millennium.128
A final development of the late twentieth century that deserves to be
highlighted here concerns what is sometimes referred to as the ‘spatial turn’
that, from the early 1990s onwards, began to transform approaches to urban
space in the Roman world, and fostered scholars to think about the spatial
context of crafts and trade, and their interaction with the urban environment.
A leading role in this was played by Ray Laurence and his book on the relation
between space and society in Roman Pompeii.129 Laurence argued that various
uses of space, including retail and manufacturing, in pre-modern cities were
intermingled, and showed how this worked out in practice for several sectors

121 122
Petersen (2003; 2006: 84–120). Clarke (2003).
123
Royden (1988, 1989); Patterson (1993, 1994; cf. 2006: 252–63).
124
Bollmann (1998). The link between urban space and collegiate life was further elaborated
for Ostia by Stöger (2011).
125 126
DeLaine (2003); Liu (2009); see also Liu (2008). Van Nijf (1997).
127 128
Dittmann-Schöne (2001); Zimmermann (2002). See, e.g. Venticinque (2009).
129
Laurence (1994). Raper (1977) and La Torre (1988) had also discussed the economic
geography of Pompeii but without an explicit spatial interest.
 Towards an Intellectual History 43

of the Pompeian economy: shops and workshops were to be found directly

next to or even associated with residential complexes, and not spatially
restricted to certain quarters of the city. This picture was subsequently elab-
orated and refined by several others, including Wallace Hadrill, Damian
Robinson, and one of the present authors.130 Outside Pompeii, Penelope
Goodman has analysed the position of crafts in the urban periphery in Italy
and Gaul.131 Related to this is a more text-oriented discourse on the sensory
impact of certain potentially ‘dirty’ crafts on their urban environment.132

DISCU SSION

This has, of course, been only a limited survey, which has tried to sketch, very
roughly, some broad developments in the study of Roman craftsmen and
traders; there is a broad variety of debates about specific crafts and trades that
could not be addressed here, and there are important issues related to the
history of crafts and trade that have barely been touched upon, including the
debates about (slave) labour and technology. Important European research
traditions have been mentioned only in passing, and only a couple of these will
be dealt with elsewhere in this volume. Nevertheless, some general points
deserve to be made.
First, modern discussions of the debate about the ancient economy have put
a lot of emphasis on Rostovtzeff and Finley, who are commonly seen as the
almost proverbial representatives of the two ‘camps’ in the debate about the
ancient economy. This is obviously a gross oversimplification. The merit of
either of those two scholars is obvious and need not be doubted, but, in terms
of moving the debate forward, other scholars seem to have been equally or
more important—both Rostovtzeff and Finley stand at the end, rather than at
the start, of a development. As far as the twentieth century is concerned, three
scholars must be mentioned. In the first place, one should not underestimate
the influence of Eduard Meyer on the thinking and the career of both Frank
and Rostovtzeff. Meyer was a key figure who, through his network, had a
profound impact on the debate, even if he did not himself publish a lot that has
stood the test of time. Second, it may be suggested that the key figure in the
decades between 1910 and 1940 was very much Frank, and not Rostovtzeff.
Not only did Frank, through his publications in the 1910s, lay down a solid
basis on which Rostovtzeff could build, but he also continued feeding the
discourse on the Roman economy long after Rostovtzeff had moved on to
greener pastures. Third, it is A. H. M. Jones who, much more than Finley,

130 131
Robinson (2005); Flohr (2007, 2012). Goodman (2007: 105–18).
132
Esp. fulling. Cf. Bradley (2002); Kudlien (2002). See also Papi (2002); Béal (2002).
44 Miko Flohr and Andrew Wilson

must be seen as the protagonist of the primitivist reaction—not so much

because he hired Finley in Cambridge, but because, especially regarding
urban crafts and trade, he was the first to write down many of the ideas
that, later on, would be primarily associated with Finley. Finley, at least
partially, reaped what Jones had sown, and the legacy of his work profited
immensely from the fact that he published at a time when there was a
substantial upsurge of interest in the social and economic history of the
ancient world.
A second point concerns the historical development of the debate. There
seem to be three key phases. First, there is the late nineteenth century, which
saw the emergence of a large canon of easily accessible and well-known
evidence for urban crafts and trade, and, independently of that, the develop-
ment of several models for the history of urban societies. From the 1890s
onwards, the German debate about Rodbertus’s idea of the Oikenwirtschaft
fostered classical scholars to analyse the economic background of the evidence
for crafts and trade, which in the 1920s and 1930s would culminate in the
works of Frank and Rostovtzeff. Intellectually, this period saw a shift from
cataloguing the evidence for manufacturing and trade to actually using it for
discussing aspects of economic history. It may be argued that this second
phase was quite decisively influenced by the First World War, the outcome of
which completely changed the social dynamics of the debate: essentially, it
ended a flourishing German scholarly tradition, and caused the geographic
focus of the debate to shift from Europe to North America, where it stayed
until well after the Second World War. The third phase began in the late
1960s, when classical scholars started to develop an interest in the more
mundane aspects of everyday life. Both the debate about The Ancient Economy
and the other developments in the field are essentially a product of this trend.
What sets this phase apart from the preceding phase is the increased prom-
inence of questions pertaining to the realm of social and cultural history. This
has made scholarly discourse about craftsmen and traders considerably more
complex and multifaceted, and it has put new questions on the agenda. Most
of the old questions, however, have not completely disappeared.
Finally, there is the question as to what moved the debate forward. In this
respect one cannot help but noticing the similarity between the ‘German’
debate of the 1890s and 1900s, and the ‘British’ debate of the 1980s and
1990s. In both cases, the debate was preceded by a long build-up in which
new sets of evidence were added to the field: the upsurge in Roman provincial
archaeology of the 1960s and 1970s mirrors the systematic publication of the
first epigraphic corpora in the 1870s and 1880s. Moreover, in both cases, the
developments in evidence were paralleled by theoretical debates taking place
among generalists outside or in the margins of classical studies, which, at some
point, gained prominence and provoked a (dismissive) response by specialists,
who then began to explore the possibilities of the new evidence to prove the
 Towards an Intellectual History 45

theoretical models wrong. Rodbertus and Bücher played a role that is to some
extent similar to that played by Jones and Finley eighty years later, though
the latter two of course were operating from within the field of classical
studies. In both cases, there also seems to have been a clear political element
that motivated the response by specialists: if Meyer felt that it was essential for
the future of classical studies to prove the relevance of ancient economies to
the modern world, many of the epigraphists, papyrologists, and archaeologists
who took on Finley also implicitly or, occasionally, explicitly opposed the way
in which his approach to the ancient economy marginalized epigraphic,
documentary, and material evidence in favour of the literary canon. On this
level, it is also obvious that the specialists, in both cases, got their way: in the
1890s as well as in the 1980s, the most easily observable outcome of the debate
was a markedly increased integration of new data sets into the debate about
Roman urban craftsmen and traders.

REFERENCES

Alföldy, G. (1975). Römische Sozialgeschichte. Wiesbaden.

Audring, G. (2000). Gelehrtenalltag: Der Briefwechsel zwischen Eduard Meyer und
Georg Wissowa (1890–1927). Hildesheim.
Ballu, A. (1897). Les ruines de Timgad (antique Thamugadi). Paris.
Béal, J.-C. (2002). ‘L’artisanat et la ville: Relecture de quelques textes’, in J. C. Béal and
J. C. Goyon (eds), Les artisans dans la ville antique. Paris, 5–14.
Beloch, J. (1899). ‘Die Grossindustrie im Altertum’, Zeitschrift für Sozialwissenschaft
2: 18–26.
Berke, S. (1988). ‘Zum Export mittelgallischer und früher Rheinzaberner Terra Sigillata
in das Barbaricum nördlich der mittleren Donau’, MBAH 7/1: 46–61.
Bianchi Bandinelli, R. (1967). ‘Arte Plebea’, Dialoghi di Archeologia 1: 7–19.
Blümner, H. (1869). Die gewerbliche Thätigkeit der Völker des klassischen Alterthums.
Leipzig.
Blümner, H. (1875–84). Technologie und Terminologie der Gewerbe und Künste bei
Griechen und Römern. Leipzig.
Bollmann, B. (1998). Römische Vereinshäuser. Untersuchungen zu den Scholae der
römischen Berufs-, Kult- und Augustalen-Kollegien in Italien. Mainz.
Bounegru, O. (1984). ‘Beiträge zur Handelsschiffart im westlichen Schwarzen Meer in
hellenistischer und römischer Zeit’, MBAH 3/2: 1–17.
Bradley, M. (2002). ‘ “It all comes out in the wash”: Looking harder at the Roman
fullonica’, JRA 15: 21–44.
Brewer, A. (2008). ‘Adam Smith’s Stages of History’, Bristol University Department of
Economics Discussion Papers 08/601.
Brewster, E. (1917). Roman Craftsmen and Tradesmen of the Early Empire. Menasha, WI.
Brewster, E. (1927). ‘A Weaver of Oxyrhynchus: Sketch of a Humble Life in Roman Egypt’,
Transactions and Proceedings of the American Philological Association 58: 132–54.
46 Miko Flohr and Andrew Wilson
Brewster, E. (1931). ‘A Weaver’s Life in Oxyrhynchus: His Status in the Community’,
in G. D. Hadzsits (ed.), Classical Studies in Honor of John C. Rolfe. Philadelphia,
19–45.
Bruneau, P., and Ducat, J. (1966). Guide de Délos. 2nd edn. Paris.
Brunt, P. A. (2004). ‘Jones, Arnold Hugh Martin’, in C. Matthew, B. Harrison, and
L. Goldman (eds), Oxford Dictionary of National Biography. Oxford.
Bücher, K. (1893). Die Entstehung der Volkswirtschaft. Tübingen.
Burnham, B., and Wacher, J. (1990). The ‘Small Towns’ of Roman Britain. London.
Calder, W. M., and Demandt, A. (1990) (eds). Eduard Meyer: Leben und Leistung eines
Universalhistorikers. Leiden.
Chambers, M. (1990). ‘The “Most Eminent Living Historian: The One Final Authority”:
Meyer in America’, in Calder and Demandt (1990), 97–131.
Chamonard, J. (1922). Exploration archéologique de Délos—Le Quartier du théatre.
Paris.
Chardron-Picault, P., and Pernot, M. (1999). Un quartier antique d’artisanat métal-
lurgique à Autun: Le Site du Lycée militaire. Paris.
Charlesworth, M. P. (1924). Trade Routes and Commerce of the Roman Empire.
Cambridge.
Christ, K. (1972). Von Gibbon zu Rostovtzeff: Leben und Werk führender Althistoriker
der Neuzeit. Darmstadt.
Clarke, J. (2003). Art in the Lives of Ordinary Romans. Berkeley.
Cornell, T., and Lomas, K. (1995) (eds). Urban Society in Roman Italy. London.
D’Ambra, E. (1988). ‘A Myth for a Smith: A Meleager Sarcophagus from a Tomb in
Ostia’, American Journal of Archaeology 92/1: 85–99.
Daremberg, C., and Saglio, E. (1873) (eds). Dictionnaire des antiquités grecques et
romaines. Paris.
D’Arms, J. (1981). Commerce and Social Standing in Ancient Rome. Cambridge, MA.
Dechelette, J. (1904). Les Vases céramiques ornés de la Gaule romaine (Narbonnaise,
Aquitaine, Lyonnaise). Paris.
DeLaine, J. (2003). The Builders of Roman Ostia: Organisation, Status and Society.
Madrid.
De Robertis, F. M. (1938). Il diritto associativo romano dai collegi della repubblica alle
corporazioni del basso impero. Bari.
De Ruggiero, E. (1895–). Dizionario epigrafico di antichità romane. Rome.
Dissen, M. (2009). Römische Kollegien und deutsche Geschichtswissenschaft im 19. und
20. Jahrhundert (Historia: Einzelschriften 209). Stuttgart.
Dittmann-Schöne, I. (2001). Die Berufsvereine in den Städten des kaiserzeitlichen
Kleinasiens. Regensburg.
Dixon, S. (2001). ‘How do you Count them if they’re not there? New Perspectives on
Roman Cloth Production’, Opuscula Romana 25–6: 7–17.
Dragendorff, H. (1895). ‘Terra sigillata: Ein Beitrag zur Geschichte der griechischen
und römischen Keramik’, BJ 96: 18–155.
Drexhage, H.-J. (1982). ‘Beitrag zum Binnenhandel im römischen Ägypten aufgrund
der Torzolquittungen und Zollhausabrechnungen des Faijum’, MBAH 1/1: 61–84.
Drexhage, H.-J. (1983). ‘Die Expositio totius mundi et gentium: Eine Handelsgeographie
aus dem 4. Jahrhundert n. Chr.’, MBAH 2/1: 3–41.
 Towards an Intellectual History 47
Drexhage, H.-J. (1993). ‘Garum und Garumhandel im Römischen und spätantiken
Ägypten’, MBAH 12/1: 27–54.
Drexhage, H.-J. (1994). ‘Einflüsse des Zollwesens auf den Warenverkehr im römischen
Reich - handelshemmend oder handelsförderend’, MBAH 13/2: 1–15.
Drexhage, H.-J. (1996). ‘Der Handel, die Produktion und der Verzehr von Käse nach
den griechischen Papyri und Ostraka’, MBAH 15/2: 33–41.
Drexhage, H.-J. (1997a). ‘Bierproduzenten und Bierhändler in der papyrologischen
Überlieferung’, MBAH 16/2: 32–9.
Drexhage, H.-J. (1997b). ‘Einige Bemerkungen zu Fleischverarbeitung und Fleischvertrieb
nach den griechischen Papyri und Ostraka vom 3. Jh. v. bis zum 7. Jh. n. Chr.’, MBAH
16/2: 97–111.
Drexhage, H.-J. (2002a). ‘Der erebinthas namens Kosmas und sein Produkt: Zum
Anbau und Vertrib von Kichererbsen nach dem papyrologischen Befund’, MBAH
21/2: 10–23.
Drexhage, H.-J. (2002b). ‘Zum letzten Mal zu den Komposita mit -poles?! Einige
Bemerkungen zur literarischen Überlieferung’, MBAH 21/2: 74–89.
Drexhage, H.-J. (2004). ‘Zu den Berufsbezeichnungen mit dem Suffix -âs in der
literarischen, papyrologischen und epigraphischen Überlieferung’, MBAH 23/1:
18–40.
Drexhage, R. (1982). ‘Der Handel Palmyras in römischer Zeit’, MBAH 1/1: 17–34.
Dyer, T. H. (1867). Pompeii: Its History, Buildings, and Antiquities. London: Bell and
Daldy.
Dyer, T. H. (1868). Pompeii: Its History, Buildings, and Antiquities. 2nd edn. London.
Dyer, T. H. (1871). Pompeii: Its History, Buildings, and Antiquities. 3rd edn. London.
Engels, D. (1990). Roman Corinth: An Alternative Model for the Classical City.
Chicago.
Erdkamp, P. P. M. (2001). ‘Beyond the Limits of the Consumer City’, Historia:
Zeitschrift für Alte Geschichte 50: 332–56.
Ernesti, J. A. (1737). De negotiatoribus Romanis diputatiuncula. Leipzig.
Espérandieu, E. (1907–). Recueil général des bas-reliefs de la Gaule romaine. Paris.
Fellmeth, U. (1991). ‘Die Häfen von Ostia und ihre wirtschaftliche Bedeutung für die
Stadt Rom’, MBAH 10/1: 1–32.
Finley, M. I. (1965). ‘Technical Innovation and Economic Progress in the Ancient
World’, Economic History Review 18/1: 29–45.
Finley, M. I. (1973). The Ancient Economy. Berkeley.
Finley, M. I. (1977). ‘The Ancient City: From Fustel de Coulanges to Max Weber and
Beyond’, Comparative Studies in Society and History 19: 305–27.
Finley, M. I. (1985). The Ancient Economy. 2nd edn. London.
Flohr, M. (2007). ‘Nec quicquam ingenuum habere potest officina? Spatial Contexts of
Urban Production at Pompeii, AD 79’, BABesch 82/1: 129–48.
Flohr, M. (2012). ‘Working and Living under One Roof: Workshops in Pompeian
Atrium Houses’, in A. Anguissola (ed.), Privata Luxuria: Towards an Archaeology
of Intimacy: Pompeii and Beyond. Munich, 51–72.
Flohr, M. (2013). ‘The Textile Economy of Pompeii’, JRA 26/1: 53–78.
Frank, T. (1910). ‘Commercialism and Roman Territorial Expansion’, Classical Jour-
nal 5/3: 99–110.
48 Miko Flohr and Andrew Wilson
Frank, T. (1914). Roman Imperialism. New York.
Frank, T. (1918). ‘The Economic Life of an Ancient City’, Classical Philology 13/3:
225–40.
Frank, T. (1920). An Economic History of Rome to the End of the Republic. Baltimore.
Frank, T. (1927). An Economic History of Rome. New York.
Frank, T. (1933). An Economic Survey of Ancient Rome, i. Rome and Italy of the
Republic. Baltimore.
Frank, T. (ed.) (1937). An Economic Survey of Ancient Rome, iii. Britain, Spain, Sicily,
Gaul. New Jersey.
Frank, T. (ed.) (1938). An Economic Survey of Ancient Rome, iv. Roman Africa; Roman
Syria; Roman Greece; Roman Asia. Baltimore.
Frank, T. (1940). An Economic Survey of Ancient Rome, v. Rome and Italy of the
Empire. Baltimore.
Fülle, G. (2000). ‘Die Organisation der Terra sigillata-Herstellung in La Graufesenque:
Die Töpfergraffiti’, MBAH 19/2: 62–98.
Fulford, M. (1982). ‘Town and Country in Roman Britain—a Parasitical Relationship’,
in D. Miles (ed.), The Romano-British Countryside: Studies in Rural Settlement and
Economy, part 2 (BAR British Series 103). Oxford, 403–19.
Gell, W. (1834). Pompeiana: The Topography, Edifices and Ornaments of Pompeii: The
Result of Excavations since 1819. London.
Goodman, P. (2007). The Roman City and its Periphery: From Rome to Gaul. London.
Gummerus, H. (1906). Der römische Gutsbetrieb als wirtschaftlicher Organismus nach
den Werken des Cato, Varro und Columella. Leipzig.
Gummerus, H. (1913). ‘Darstellungen aus dem Handwerk auf römischen Grab- und
Votivsteinen in Italien’, Jahrbuch des Deutschen Archäologischen Instituts 28:
63–126.
Gummerus, H. (1915). ‘Die römische Industrie: Wirtschaftliche Untersuchungen. Bd
I. Das Goldschmied- und Juweliersgewerbe’, Klio: Beiträge zur alten Geschichte 14:
129–89.
Gummerus, H. (1916). ‘Industrie und Handel bei den Römern’, in G. Wissowa and
W. Kroll (eds), Paulys Real-Encyclopädie der Classischen Altertumswissenschaft:
Neue Bearbeitung. Stuttgart, 1439–535.
Habermann, W. (1982). ‘Ostia, Getreidehandelshafen Roms’, MBAH 1/1: 35–60.
Habermann, W. (1990). ‘Statistische Datenanalyse an den Zolldokumenten des Arsi-
noites aus römischer Zeit’, MBAH 9/1: 50–85.
Harris, W. V. (2013). Moses Finley and Politics. Brill.
Hatzfeld, J. (1919). Les Trafiquants italiens dans l’Orient hellenique. Paris.
Henning, D. (2001). ‘Die antike Seehandelsroute um Kap Malea’, MBAH 20/1: 23–37.
Hobsbawm, E. (ed.) (1964). Karl Marx, Pre-Capitalist Economic Formations. London.
Hoffmeister, K. (1899). Die Wirtschaftliche Entwicklung Roms: Eine sozialpolitische
Studie. Vienna.
Holtheide, B. (1982). ‘Zum privaten Seenhandel im östlichen Mittelmeer, 1.–3. Jh. n.
Chr.’, MBAH 1/2: 3–12.
Hopkins, K. (1978). ‘Economic Growth and Towns in Classical Antiquity’, in
P. Abrams and E. Wrigley (eds), Towns in Societies: Essays in Economic History
and Historical Sociology. Cambridge, 35–78.
 Towards an Intellectual History 49
Hopkins, K. (1980). ‘Taxes and Trade in the Roman Empire’, JRS 70: 101–25.
Hopkins, K. (1983). ‘Introduction’, in P. Garnsey, K. Hopkins, and C. R. Whittaker
(eds), Trade in the Ancient Economy. London.
Hume, D. (ed.) (1752). Political Discourses. Edinburgh.
Jahn, O. (1861). ‘Darstellungen antiker Reliefs, welche sich auf Handwerk und Han-
delsverkehr beziehen’, Berichte über die Verhandlungen der Königlich Sächsischen
Gesellschaft der Wissenschaften zu Leipzig. Philologisch-Historische Classe 13:
291–374.
Jahn, O. (1868). ‘Ueber Darstellungen des Handwerks und Handelsverkehrs auf
Antiken Wandgemälde’, Abhandlungen der Philologisch-Historischen Klasse der
Königlich-Sächsischen Gesellschaft der Wissenschaften 5/4: 266–318.
Johnson, A. C. (1936). Roman Egypt to the Reign of Diocletian (= T. Frank (ed.), An
Economic Survey of Ancient Rome, ii). Baltimore.
Jones, A. H. M. (1940). The Greek City from Alexander to Justinian. Oxford.
Jones, A. H. M. (1948). Ancient Economic History. London.
Jones, A. H. M. (1955). ‘The Economic Life of the Towns of the Roman Empire’,
Recueils de la Société Jean Bodin pour l’Histoire Comparative des Institutions 7:
161–92.
Jongman, W. (1988). The Economy and Society of Pompeii. Amsterdam: Gieben.
Joshel, S. R. (1992). Work, Identity and Legal Status at Rome. Norman, OK, and
London.
Kampen, N. (1981). Image and Status: Roman Working Women in Ostia. Berlin.
Kampen, N. (1985). ‘Römische Straßenhändlerinnen: Geschlecht und Sozialstatus’,
Antike Welt 16/4: 23–42.
Kneissl, P. (1983). ‘Mercator–negotiator: Römische Geschäftsleute und die Termino-
logie ihrer Berufe’, MBAH 2/1: 73–90.
Konen, H. (2001). ‘Die Schiffsbauer und Werften in den antiken Häfen von Ostia und
Portus’, MBAH 20/2: 1–36.
Kudlien, F. (1997). ‘Der antike Makler—ein verleugneter Beruf ’, MBAH 16/1: 67–84.
Kudlien, F. (2002). ‘P. Patulcius L.f., Walker und Probulos um späthellenistischen
Magnesia’, Laverna: Beiträge zur Wirtschafts- und Sozialgeschichte der alten Welt
13: 56–68.
Kuhn, G. B. (1910). De opificum Romanorum condicione privata quaestiones. Halle.
La Torre, G. (1988). ‘Gli impianti commerciali ed artigiani nel tessuto urbano di
Pompei’, in L. Dell’Orto (ed.), Pompei, l’informatica al servizio di una città antica.
Rome, 73–102.
Larsson Lovén, L. (2002). The Imagery of Textile Making: Gender and Status in the
Funerary Iconography of Textile Manufacture in Roman Italy and Gaul. Göteborg.
Laurence, R. (1994). Roman Pompeii—Space and Society. London.
Leveau, P. (1984). Caesarea de Maurétanie: Une ville romaine et ses campagnes. Rome.
Leveau, P. (ed.) (1985). L’Origine des richesses dépensées dans la ville antique. Aix-
en-Provence.
Liebenam, W. (1890). Zur Geschichte und Organisation des römischen Vereinswesens: 3
Untersuchungen. Leipzig.
Liu, J. (2008). ‘Pompeii and Collegia: A New Appraisal of the Evidence’, Ancient
History Bulletin 22: 53–70.
50 Miko Flohr and Andrew Wilson
Liu, J. (2009). Collegia Centonariorum: The Guilds of Textile Dealers in the Roman
West. Leiden and Boston.
Loane, H. J. (1938). Industry and Commerce of the City of Rome. Baltimore.
Mangin, M. (1981). Un quartier de commerçants et d’artisans d’Alésia: Contribution à
l’histoire de l’habitat urbain en Gaule. Paris.
Marquardt, J. (1882). Das Privatleben der Römer. II: Erwerb und Unterhaltung. Leipzig.
Marx, K., and Engels, F. (1932). Die Deutsche Ideologie. Berlin.
Matei, C. (1989). ‘Notes on the Activity in the Port of Ancient Tomis’, MBAH 8/1:
39–55.
Mattingly, D. J. (2000). ‘Leptiminus (Tunisia). A “Producer” City?’, in Mattingly and
Salmon (2000), 66–89.
Mattingly, D. J., and Salmon, J. (eds) (2000). Economies beyond Agriculture in the
Classical World. London.
Mau, A. (1900). Pompeji in Leben und Kunst. Leipzig.
Mau, A., and Kelsey, F. W. (1899). Pompeii: Its Life and Art. London.
Maué, H. (1886). Die Vereine der fabri, centonarii und dendrophori im römischen Reich:
Die Natur ihres Handwerks und ihre sacralen Beziehungen. Frankfurt am Main.
Maxey, M. (1938). Occupations of the Lower Classes in Roman Society as Seen in
Justinian’s Digest. Chicago.
Meek, R. L. et al. (eds) (1978). Adam Smith: Lectures on Jurisprudence. Oxford.
Mees, A. W. (1994). ‘Datierung und Vertrieb von reliefverzierten Sigillaten aus
Banassac’, MBAH 13/2: 60–9.
Meiggs, R. (1970). ‘Obituary: Arnold Hugh Martin Jones’, JRS 60: 186–7.
Meyer, E. (1895). ‘Die wirtschaftliche Entwicklung des Altertums’, Jahrbücher für
Nationalökonomie und Statistik.
Meyer, E. (ed.) (1910). Kleine Schriften zur geschichtstheorie und zur Wirtschaftlichen
und politischen Geschichte des Altertums. Halle.
Moeller, W. O. (1976). The Wool Trade of Ancient Pompeii. Leiden.
Momigliano, A. (1950). ‘Ancient History and the Antiquarian’, Journal of the Warburg
and Courtauld Institutes 13/3–4: 285–315.
Mommsen, T. (1843). De collegiis et sodaliciis Romanorum. Kiel.
Monfort, C. C. (1999). ‘The Nature of the Roman Trade: An Archaeological Perspective’,
MBAH 18/2: 87–114.
Morley, N. (1998). ‘Political Economy and Classical Antiquity’, Journal of the History
of Ideas 26: 95–114.
Morley, N. (2004). Theories, Models and Concepts in Ancient History. London.
Morris, I. (1999). ‘Foreword to the Updated Edition’, in M. I. Finley, The Ancient
Economy, 3rd edn. Los Angeles, pp. ix–xxxvi.
Olivanti, P. (2001). ‘Les Fouilles d’Ostie de Vaglieri à nos jours’, in J. P. Descoeudres
(ed.), Ostia: Port et porte de la Rome antique. Geneva, 56–65.
Overbeek, A. (2005). ‘Realistische Geschiedschrijving. Theodor Mommsen’s “Römische
Geschichte”’. Unpublished Dissertation. Groningen.
Overbeck, J. (1856). Pompeji in seinen Gebäuden, Alterthümern und Kunstwerken: Für
Kunst- und Alterthumsfreunde. Leipzig.
Overbeck, J. (1866). Pompeji in seinen Gebäuden, Alterthümern und Kunstwerken: Für
Kunst- und Alterthumsfreunde. 2nd edn. Leipzig.
 Towards an Intellectual History 51
Overbeck, J. (1875). Pompeji in seinen Gebäuden, Alterthümern und Kunstwerken: Für
Kunst- und Alterthumsfreunde. 3rd edn. Leipzig.
Overbeck, J., and Mau, A. (1884). Pompeji in seinen Gebäuden, Alterthümern und
Kunstwerken. 4th edn. Leipzig.
Papi, E. (2002). ‘La Turba inpia: Artigiani e commecianti del Foro Romano e dintorni
(I sec. a.C.–64 d.C.)’, JRA 15/1: 45–62.
Parkins, H. M. (ed.) (1997). Roman Urbanism: Beyond the Consumer City. London.
Parkins, H. M., and Smith, C. (eds) (1998). Trade, Traders and the Ancient City.
London.
Patterson, J. (1993). ‘Patronage, Collegia and Burial in Imperial Rome’, in S. Bassett
(ed.), Death in Towns. 100–1600. Leicester, 15–27.
Patterson, J. (1994). ‘The Collegia and the Transformation of the Towns of Italy in the
Second Century AD’, in L’Italie d’Auguste à Dioclétien: Actes du colloque international
organisé par l’École française de Rome. Rome, 227–38.
Patterson, J. (2006). Landscapes and Cities: Rural Settlement and Civic Transformation
in Early Imperial Italy. Oxford.
Perry, J. (2006). The Roman Collegia: The Modern Evolution of an Ancient Concept.
Leiden.
Pesando, F., and Guidobaldi, M.-P. (2006). Pompei Oplontis Ercolano Stabiae. Rome.
Petersen, L. H. (2003). ‘The Baker, his Tomb, his Wife, and her Breadbasket: The
Monument of Eurysaces in Rome’, Art Bulletin 85/2: 230–57.
Petersen, L. H. (2006). The Freedman in Roman Art and Art History. New York.
Poblome, J., Degryse, P., Librecht, I., and Waelkens, M. (1998). ‘Sagalassos Red Slip Ware:
The Organization of a Manufactory’, MBAH 17/2: 52–64.
Raepsaet, G. (1987). ‘Aspects de l’organisation du commerce de la céramique sigillée
dans le Nord de la Gaule au IIe siècle de notre ère, 1. Les données matérielles’,
MBAH 6/2: 1–29.
Raper, R. (1977). ‘The Analysis of the Urban Structure of Pompeii: A Sociological
Examination of Land Use (Semi-Micro)’, in D. Clarke (ed.), Spatial Archaeology.
London, 189–221.
Rebenich, S. (2002). Theodor Mommsen: Eine Biographie. Munich.
Reibig, A. (2001). ‘The Bücher–Meyer Controversy: The Nature of the Ancient
Economy in Modern Ideology’. Unpublished Dissertation. Glasgow.
Reubel, G. (1912). Römische Töpfer in Rheinzabern: Ein Beitrag zur Geschichte der
verzierten Terra sigillata. Speier am Rhein.
Rich, J., and Wallace-Hadrill, A. (1991). City and Country in the Ancient World.
London.
Robinson, D. (2005). ‘Re-Thinking the Social Organisation of Trade and Industry in
First Century AD Pompeii’, in A. MacMahon and J. Price (eds), Roman Working
Lives and Urban Living. Oxford, 88–105.
Rodbertus, J. K. (1865). ‘Untersuchungen auf dem Gebiete der Nationalökonomie des
Klassischen Altertums. II. Zur Geschichte der römischen Tributsteuern seit Augustus’,
Jahrbücher für Nationalökonomie und Statistik 4: 341–427.
Rostovtzeff, M. (1926). The Social and Economic History of the Roman Empire. Oxford.
Royden, H. L. (1988). The Magistrates of the Roman Professional Collegia in Italy from
the First to the Third Century AD. Pisa.
52 Miko Flohr and Andrew Wilson
Royden, H. L. (1989). ‘The Tenure of Office of the Quinquennales in the Roman
Professional Collegia’, American Journal of Philology 110/2: 303–15.
Ruffing, K. (2001). ‘Einige Überlegungen zum Weinhandel im römischen Ägypten (1.–3.
Jh. n. Chr.)’, MBAH 20/1: 55–80.
Ruffing, K. (2002). ‘Die Berufsbezeichnungen auf -poles und -prates in der epigra-
phischen Überlieferung’, MBAH 21/1: 16–58.
Saller, R. (2002). ‘Framing the Debate over Growth in the Ancient Economy’, in
W. Scheidel and S. Von Reden (eds), The Ancient Economy. Edinburgh, 251–69.
Schleich, T. (1983). ‘Überlegungen zum Problem senatorischer Handelsaktivitäten
(1)’, MBAH 2/2: 65–90.
Schleich, T. (1984). ‘Überlegungen zum Problem senatorischer Handelsaktivitäten
(2)’, MBAH 3/1: 37–72.
Schneider, H. (1982). ‘Die Bedeutung der römischen Strassen für den Handel’, MBAH
1/1: 85–95.
Schneider, H. (2009). ‘Die Erforschung der antiken Wirtschaft vom Ende des 18.
Jahrhunderts bis zum Zweiten Weltkrieg: Von A. H. L. Heeren zu M. I. Rostovtzeff ’,
in V. Losemann (ed.), Alte Geschichte zwischen Wissenschaft und Politik: Gedenkschrift
Karl Christ. Wiesbaden, 337–85.
Schöttgen, C. (1727). Antiquitates trituræ et fulloniæ, ex antiquorum temporum
reliquiis congestæ. Utrecht.
Sidebotham, S. E. (1986). ‘Ports of the Red Sea and the Arabia–India trade’, MBAH 5/2:
16–36.
Smith, A. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations.
London.
Smith, W. (1848). A Dictionary of Greek and Roman Antiquities. London.
Stöger, H. (2011). ‘The Spatial Organization of the Movement Economy: The Analysis
of Ostia’s Scholae’, in R. Laurence and D. Newsome (eds), Rome, Ostia, Pompeii:
Movement and Space. Oxford, 215–42.
Strobel, K. (1987). ‘Einige Bemerkungen zu den historisch-archäologischen Grundlagen
einer Neuformulierung der Sigillatenchronologie für Germanien und Rätien und zu
wirtschaftsgeschichtlichen Aspekten der römischen Keramikindustrie’, MBAH 6/2:
75–115.
Strong, D., and Brown, D. (eds) (1976). Roman Crafts. London.
Steuart, J. (1767). An Inquiry into the Principles of Political Economy. London.
Stuppner, A. (1994). ‘Terra Sigillata im Nördlichen Niederösterreich’, MBAH 13/2:
70–94.
Tran, N. (2001). ‘Le Collège, la communauté et le politique sous le Haut-Empire
romain: Remarques sur l’histoire du droit à la fin du XIXe siècle’, Cahiers Centre
Glotz 12: 181–98.
Treggiari, S. (1975a). ‘Jobs in the Household of Livia’, PBSR 43: 48–77.
Treggiari, S. (1975b). ‘Roman Social History: Recent Interpretations’, Social History/
Histoire Sociale 8: 149–64.
Treggiari, S. (1976). ‘Jobs for Women’, American Journal of Ancient History 1: 76–104.
Treggiari, S. (1979). ‘Lower-Class Women in the Roman Economy’, Florilegium:
Carleton University Annual Papers on Classical Antiquity and the Middle Ages 1:
65–86.
 Towards an Intellectual History 53
Treggiari, S. (1980). ‘Urban Labour in Rome: Mercennarii and Tabernarii’, in
P. Garnsey (ed.), Non-Slave Labour in the Roman World. Cambridge, 48–64.
Van Nijf, O. M. (1997). The Civic World of Professional Associations in the Roman
East. Amsterdam.
Venticinque, P. (2009). ‘Common Causes: Guilds, Craftsmen and Merchants in the
Economy and Society of Roman and Late Roman Egypt’. Unpublished dissertation.
University of Chicago.
Wagner-Hasel, B. (2009). ‘Hundert Jahre Gelehrtenstreit über den Charakter der
antiken Wirtschaft’, Historische Anthropologie 17: 178–201.
Wallace-Hadrill, A. (1991). ‘Elites and Trade in the Roman Town’, in J. Rich and
A. Wallace-Hadrill (eds), City and Country in the Ancient World. London, 241–72.
Waltzing, J.-P. (1892). L’Épigraphie latine et les corporations professionnelles de l’empire
Romain. Gand.
Waltzing, J.-P. (1895). Étude historique sur les corporations professionelles chez les
Romains, I: Le Droit d’association à Rome. Les Collèges professionelles considérés
comme associations privées. Louvain.
Waltzing, J.-P. (1896). Étude historique sur les corporations professionnelles chez les
Romains, II: Les Collèges professionels considérés comme institutions officielles.
Louvain.
Waltzing, J.-P. (1900). Étude historique sur les corporations professionnelles chez les
Romains. Indices. Liste des collèges connus, leur organisation intérieure, leur carac-
tère religieux, funéraire et public, leurs finances. Louvain.
Warmington, E. H. (1928). The Commerce between the Roman Empire and India.
Cambridge.
Wes, M. A. (1990). Michael Rostovtzeff, Historian in Exile: Russian Roots in an
American Context (Historia: Einzelschriften 65). Stuttgart, Steiner.
West, L. C. (1917). ‘Phases of Commercial Life in Roman Egypt’, JRS 7: 45–58.
West, L. C. (1924). ‘Commercial Syria under the Roman Empire’, Transactions and
Proceedings of the American Philological Association 55: 159–89.
West, L. C. (1932). ‘The Economic Collapse of the Roman Empire’, Classical Journal
28/2: 96–106.
West, L. C. (1939). ‘Notes on Diocletian’s Edict’, Classical Philology 34/3: 239–45.
Whittaker, C. R. (1991). ‘The Consumer City Revisited: The Vicus and the City’, JRA 3:
110–18.
Wilson, A. (1999). ‘Commerce and Industry in Roman Sabratha’, Libyan Studies 30:
29–52.
Wilson, A. (2000). ‘Timgad and Textile Production’, in Mattingly and Salmon (2000),
271–96.
Wilson, A. (2002). ‘Urban Production in the Roman World: The View from North
Africa’, PBSR 70: 231–74.
Wright, F. W. (1917). ‘Roman Factories’, Classical Weekly 11/3: 17–19.
Wright, F. W. (1918). ‘Review of Brewster, Roman Craftsmen and Traders’, Classical
Weekly 12/5: 38–9.
Zaniers, W. (1994). ‘Handelsfragen der Rheinzaberner Sigillata’, MBAH 13/2: 60–9.
Zanker, P. (1975). ‘Grabreliefs römischer Freigelassener’, Jahrbuch des Deutschen
Archäologischen Instituts 90: 267–315.
54 Miko Flohr and Andrew Wilson
Zimmer, G. (1982). Römische Berufsdarstellungen. Berlin.
Zimmer, G. (1985). ‘Römische Handwerker’, Aufstieg und Niedergang der Römischen
Welt 12/2: 205–28.
Zimmermann, C. (2002). Handwerkervereine im griechischen Osten des Imperium
Romanum. Mainz.
 2

Twentieth-Century Italian Scholarship on

Roman Craftsmen, Traders, and their
Professional Organizations
Carla Salvaterra and Alessandro Cristofori

Italian scholarship of the twentieth century has contributed significantly to the

questions discussed in this volume. This is explained partly by the traditional
strength of classical studies in Italy, and by the concentration there of a
considerable proportion of the relevant ancient evidence—the archaeological
and epigraphic sources in particular. Yet an additional, and more interesting,
reason is to be found in the centrality of the concept of ‘labour’ in shaping the
social, political, and constitutional structures of Italy in various periods during
the twentieth century.
The first two decades of the twentieth century saw the growth of workers’
organizations and trade unions.1 The subsequent period under the Fascist
regime saw the top-down development of a corporate system, with the creation
of organizations, the corporazioni, uniting either the employers or the workers
of one sector. The idea was to solve the class struggle that, according to the
champions of corporativismo, had resulted in the bankruptcy of both the liberal
and the socialist model: the system was aimed at fostering the greater interests
of the nation.2 After the Second World War, labour played an essential role in
the founding ideology of the new Italian republic, as attested by the first article
of the Italian constitution.3 This political and cultural climate inevitably also

1
On the origins and development of labour unions in Italy until the first decades of the
twentieth century, see, most recently, Varni (1992), Antonioli (1997), Ciampani and Pellegrini
(2005), Napoli (2007), and Vallauri (2008). Still useful is Horowitz (1963). On the revolutionary
syndicalism as a left-wing stream within the Partito Socialista Italiano, see, above all, Laghi
(1990).
2
Among the most recent contributions to the debate on the corporative system of the Fascist
era, see Gagliardi (2010).
3
On this topic, see particularly Turone (1992) and Musso (2002).
56 Carla Salvaterra and Alessandro Cristofori

influenced the priorities of classical scholars, and caused a particular interest in

issues related to the world of labour in the Roman era.
Despite all this, from an international perspective it is clear that, in the
footnotes of publications dedicated to the role of craftsmen and traders in the
Roman economy, publications written in Italian appear as references only
very rarely in this debate. Among the reasons for this state of affairs is the
prevalence in Italy, at least until some decades ago, of ‘empirical’ approaches to
the theme, with sectorial studies focusing particularly on a single subset of
sources, which strongly reflected traditional disciplinary subdivisions. This ten-
dency is accompanied by a certain disconnection between Italian classical studies
and the methodological reflection taking place beyond the national borders, and
the joint results of investigations focusing on other historical periods. In part, this
situation may also be due to an imperfect knowledge of Italian scholarly discourse
in the international field, as a result of linguistic factors or the difficulty of finding
contributions that may have an extremely limited circulation.
Our modest aim in this chapter is to contribute to overcome some of these
difficulties, and to highlight the most influential approaches to craftsmen and
traders in the Roman world in twentieth-century Italy. The discussion is
structured thematically—although a chronologically ordered arrangement,
emphasizing the parallels between the developments in research and the
changes in the political and cultural realm, would also have been interesting.4

GENERAL APPROACHES

As already noted, there have not been many works of a methodological nature
on the theme of commercial and artisanal labour in the Roman world; in the
first part of the twentieth century, such works were produced in the context of
general reconstructions of the ancient economy by scholars influenced, in
varying ways, by Marxist and socialist thinking.5 The first contributions in this
vein include a speech given at the opening of the academic year 1905/6 at the
University of Macerata by the jurist Siro Solazzi on the topic of free work in
the Roman world, and a monograph by Giuseppe Salvioli, who also had been
educated as a jurist.6 The difficult socio-economic conditions in Sicily at the

4
The present chapter develops further some of the points already presented in Salvaterra
(2006).
5
Still important for the influence of the thinking of Marx and of socialist streams on ancient
historiography in Italy is Mazza (1976), who rightly insisted on the necessity of distinguishing
the personal lives of classical scholars such as Corrado Barbagallo, Ettore Ciccotti, Guglielmo
Ferrero, and Giuseppe Salvioli, often erroneously ranked under one Marxist front within Italian
classical scholarship.
6
Solazzi (1906). The text of the speech can also be read in Solazzi (1955: 141–60).
 Twentieth-Century Italian Scholarship 57

end of the nineteenth century—Salvioli had been appointed to the chair of

history of Italian law at the University of Palermo in 1884—led him to
investigate the economic history of antiquity and the Middle Ages. His
work, Il capitalism antico: Storia dell’economia romana, appeared posthu-
mously in Italian in 1929, but which had already been published in French
in 1906.7 Salvioli insisted particularly on the prominence, in real terms, of free
labour in urban crafts in the Roman world, and on the failure of the organ-
ization of production to develop from the small-scale workshop to the large
factory in the modern sense of the word.8 He sketched a nuanced picture of the
role of merchants in the Roman economy.9 Besides the work of Salvioli, there
is the œuvre of Ettore Ciccotti, a follower of Mazzini, who served the socialist
party in the late nineteenth and early twentieth centuries, and became attached
to Fascism in the last phase of his life—though he was highly critical of
Mussolini’s dictatorship.10 Ciccotti wrote Il tramonto della schiavitù nel
mondo antico, on the decline of slavery in the ancient world, and Commercio
e civiltà nel mondo antico, which, though mostly dedicated to the nature of
commercial relations from prehistory to the Roman era, has some relevant
sections on the profile of traders in the Greco-Roman world.11 Both works have
had an influence lasting until recent times through re-editions and reprints.

Francesco Maria de Robertis

However, the fundamental name to remember when it comes to general

approaches to craftsmen and traders is that of Francesco Maria de Robertis,
a scholar educated in the 1930s at the Scuola di Perfezionamento in Studi
Corporativi, at the University of Bari, a school heavily promoted by the Fascist
regime.12 De Robertis was nonetheless able to demonstrate in his early works

7
Salvioli (1929). The book can also be found in the edition prepared by Andrea Giardina:
Salvioli (1985). French version: Salvioli (1906). The book also was translated into German, before
it appeared in Italian: Salvioli (1912).
8
Salvioli (1985: 61–84).
9
Salvioli (1985: 115–32).
10
On Ciccotti as a historian of the ancient economy, see, briefly, Giardina in the preface of
Salvioli (1985: pp. xlviii–liii).
11
On slavery, see Ciccotti (1899), 2nd edn Ciccotti (1940), repr. as Ciccotti (1977). On
commerce, see Ciccotti (1929); for merchants in the Greek world, see esp. pp. lxxix–lxxxi; for
traders in the Roman world, see particularly pp. cxxvii, cxxxvi–cxxxix.
12
An excellent account of the education of De Robertis and the development of his thinking,
against the background of the ever-changing political realities during his long scholarly career, is
to be found in Perry (2006: esp. 93–102, 114–18, 165–6, 193–200). Now one can also refer to the
contributions focusing on De Robertis in the proceedings of a 2004 conference dedicated to the
great scholar, who had died the previous year: Musca (2007), esp. the chapter by Filippo Gallo on
De Robertis and the labour relations in Roman law. For a brief account of the conference, see
Quadrato (2004).
58 Carla Salvaterra and Alessandro Cristofori

an admirable independence from the propagandistic positions of Fascism, and

after the war he had a deep influence on Italian scholarship, regarding both the
concept of ‘labour’ in the Roman world, and the fenomeno associativo, the
professional collegia. In both these themes, he appears to have been influenced
by the social doctrines of the left wing of the Democrazia Cristiana, the party
that, until the early 1990s, was the pivot of every Italian government, and of
which the leader, Aldo Moro, was a personal friend of De Robertis, and
lecturer at the juridical faculty in Bari where De Robertis spent his entire
academic life.
Regardless of how one wishes to judge the value of his ideas, De Robertis had
the unequivocal merit of recognizing the importance of documentary sources in
discussing the social position of labour in Rome: before him, in Italy, such
discussions were almost completely dominated by literary sources. De Robertis
distinguished, in a series of works published between the 1940s and the 1960s, a
‘noble’ realm, which belonged to the ruling class of Rome, and of which the
literary sources were an expression.13 In this realm, there was substantial con-
tempt for most occupations. Opposite these ideas stood those championed in the
‘vulgar’ realm of the lower classes and the provincials, illustrated principally by the
epigraphic, papyrological, and iconographical records. In this realm, representing
the large majority of the empire’s population, work was a source of personal pride
and social prestige, as was demonstrated by the explicit mention of the occupation
of the deceased in commemorative inscriptions, and the numerous reliefs with
occupational scenes from various parts of the Roman world.
The interpretative model of De Robertis evoked and still evokes some
perplexity.14 Much can be said, for example, about his characterization of
the two realms, and about the boundary between ‘noble’ and ‘vulgar’, defin-
itions that De Robertis basically borrowed from the leading name in Italian
art history of his time, Ranuccio Bianchi Bandinelli.15 One could also note
that De Robertis’s model risks undue inflexibility, as it rigidly contrasts a
supposedly coherent ideology of the upper classes with another supposedly
coherent ideology of the lower classes.16 This led him to consider the

13
Particularly De Robertis (1945), a monograph, and De Robertis (1959, 1962), articles that
subsequently also appeared in De Robertis (1963: 21–47, 49–95).
14
Cristofori (2004: 79–80) aims to explore the inherent contradictions, showing how, on the
one hand, it is impossible to speak of a unique and coherent labour ideology among the Roman
ruling class—instead there was a plurality of ideas, that had some points in common, but were
not congruent and indeed sometimes divergent, sometimes even within the thinking of one and
the same author, like Cicero (pp. 81–90); on the other hand, an attempt is made to demonstrate,
based on the scanty evidence, that there was not always an opposition between ‘noble’ and
‘vulgar’ ideas, and that in some cases ideas expressed by the ruling class penetrated into the realm
of the occupational class (pp. 90–6).
15
De Robertis (1959: 55–6, n. 7).
16
Nörr (1965: 70–73) had already warned against the rigidity of the interpretative model of
De Robertis. In the same years in which De Robertis developed his model, the great historian of
 Twentieth-Century Italian Scholarship 59

ideas expressed by literary authors as abstract speculations of some

aristocrats, completely disconnected from the problems and values of real
life, and thus, in the end, not representative of historical reality.17 Neverthe-
less, he had the undeniable merit of launching, in Italy, a discussion about
these issues.

The Classical Seminar of the Istituto Gramsci

After De Robertis, the debate continued especially through the contributions

of a collective of scholars from various disciplines within the field of Alter-
tumswissenschaft, including historians, but also philologists, archaeologists,
and jurists. This group convened in 1974 at the Instituto Gramsci, the cultural
institute connected to the Italian Communist Party—a party whose cultural
weight at Italian universities in the 1960s, 1970s, and 1980s was without any
doubt more relevant than one would think on the basis of its marginal political
role of the period.18 The basis for this collaboration was laid at the Colloques
sur l’esclavage, the first of which was held in 1971 at Besançon. Here, Italian
scholars who later became part of the group of the Istituto Gramsci met their
French colleagues.19
Against this background appeared the collection of methodological essays,
in 1978, in Analisi marxista e società antiche, which published, in more
elaborate form, the papers presented in this study group during the first two
years of its existence, and, in 1979, L’anatomia della scimmia, in which the
archaeologist Andrea Carandini reinterpreted the evolution of the entire pre-
capitalist economy in the light of the fundamental chapter of Karl Marx’s
Grundrisse der Kritik der Politischen Ökonomie dedicated to the systems that
preceded the capitalist system.20 The strictly theoretical character of these
two publications means that they are of only limited relevance in debates

ancient philosophy Rodolfo Mondolfo proposed a reading contrasting with that of De Robertis,
showing, particularly among the thinkers of the Greek and Roman worlds, the recognition of a
positive value of labour, especially as a tool for experiencing reality. See esp. Mondolfo (1965).
The very possibility of an analysis like that of Mondolfo (though it may be criticized, cf. Pasoli
(1980: 75–9, particularly 77, n. 28)) should warn us against the risks of an interpretation
assuming the univocal and completely coherent nature of thinking among the nobility.
17
As objected by Galeno (1960: 143–4). The response by De Robertis (1962: 3, n. 1) does not
seem to have picked up the fundamental nature of the criticism that had been aimed at him.
18
Lussana (2000) recalls, briefly, the cultural activities of the Istituto Gramsci (now Fonda-
zione Istituto Gramsci) in the second half of the twentieth century. The communist party was
restricted to opposing the governments led by Christian democrats.
19
Cf. Brutti (1978: 10).
20
Capogrossi Colognesi, Giardina, and Schiavone (1978); Carandini (1979). The first work
was received with great interest by the Italian scholarly community, as attested by some of the
contributions to the Quaderni di Storia: Bretone et al. (1978); Di Benedetto (1978); Narducci
(1978); but also internationally. Cf. Clavel-Lévêque and Favory (1980); Kreissig (1982).
60 Carla Salvaterra and Alessandro Cristofori

about the themes discussed in this volume.21 Highly relevant, however, remains
the monumental collaborative effort edited by Andrea Giardina and Aldo
Schiavone, Società romana e produzione schiavistica, published in 1981, as
the result of several years of work of the classical seminar at the Istituto
Gramsci.22 Even though the volume has a strong focus on agricultural activ-
ities, ample space is also devoted to manufacturing and trade, especially in
the second volume, which focuses on goods, markets, and exchange in the
Mediterranean. The broad approach of the 1981 publication, compared to the
two earlier works mentioned, and its more careful consideration of the insights
offered by the ancient evidence, as opposed to theoretical models, explain the
outstanding resonance of the volume also among scholars adhering to non-
Marxist schools of thought.23 The activities of the classical seminar of the
Istituto Gramsci subsequently also led to the publication, in 1986, of a second
volume of economic studies, devoted to late antiquity—Società romana
e impero tardoantico.24 This book aimed to highlight the moment at which
the slave mode of production, whose origins and consolidation in the late
Republic and under the High Empire the earlier volume had explored, broke
up; it also significantly broadened the scope of the analysis by including the
provinces of the empire, whereas in Società romana e produzione schiavistica
most of the attention was focused on south-central Italy.
One of the most fertile and durable results of the activities of the classical
seminar of the Istituto Gramsci, besides the reflections on applying the
interpretative framework of Marxist thought to the ancient world, was that
it underlined the importance of collaboration between historians, philologists,
jurists, and archaeologists. In this sense, the continuity between the Istituto
Gramsci seminar and Opus, an international journal focusing on the economic
and social history of antiquity, was evident—also because some of the scholars
involved in the seminar played a key role in its foundation. The journal, edited
by the Greek philologist Carmine Ampolo and by the archaeologist Giuseppe
Pucci, was the first Italian periodical dedicated to the economy and society of

21
The theoretical merit of the volume was noted by Narducci (1978: 50), who, concluding his
thoughts on Capogrossi Colognesi, Giardina, and Schiavone (1978) and on the contemporan-
eous collection of essays by Vegetti (1977), wrote ‘one can only applaud that the realm of the
sacred texts is definitively being abandoned in favour of more practical investigation, of an
empiricism that is not always “vulgar”’.
22
Giardina and Schiavone (1981).
23
Cf., for the discussion in Italy, the debate published in Opus, a report of a conference held at
Pavia, 21 January 1982, on Società romana & produzione schiavistica. Cf. Gabba (1982); Lo
Cascio (1982); Luraschi (1982), and the subsequent and shorter contributions by Bona et al.
(1982); for the reception of the book abroad, see the synthesis of the contributions to a
conference at the Collège de France on 27 November 1981, published in Quaderni di Storia:
Nicolet et al. (1982), and reviews by Thébert (1982), Gros (1983), and Rathbone (1983),
appearing in some of the most widespread journals in classical studies.
24
Giardina (1986). See the reviews of this book by Silvestrini (1989) and Andreau and Leveau
(1992).
 Twentieth-Century Italian Scholarship 61

the ancient world, and first appeared in 1982. It was an explicit aim of the
journal to overcome the disciplinary boundaries that separated scholars based
on the different properties of the evidence they worked with—literary and
material—and to open a broad international collaboration.25 Significant in this
respect is the fact that the first volume of the journal published the proceed-
ings of a conference on Moses Finley’s Ancient Slavery and Modern Ideology,
which had been translated into Italian by Elio Lo Cascio shortly beforehand.26
This volume also included a contribution by Finley himself.27 It may also be
pointed out that Opus also readily accepted the contributions of Italian scholars
coming from cultural traditions other than Marxism, such as Guido Clemente
and Emilio Gabba. In the last years of the twentieth century, the activities
discussed here gradually vanished: shortly after the publication of Società
romana e impero tardoantico, the classical seminar of the Istituto Gramsci
ceased its activities. In 1992, the publication of Opus also was suspended.
After this period, as far as the theme of this volume is concerned, one can
point to some interesting syntheses, particularly by Guido Clemente and
Giusto Traina, or to arguments inserted in volumes of broader scope, like
the classic Economia e società nell’Italia annonaria of Lellia Cracco Ruggini,
published for the first time in 1961, but with a second, updated, edition
appearing in 1995, or the 2011 handbook on the Roman economy published
by Filippo Carlà and Arnaldo Marcone.28

EVIDEN CE-BASED APPROACHES

Besides the works mentioned so far, some publications dedicated to certain

literary texts that play a key role in discourse about labour in Rome succeeded
in reaching a rather wide audience.29 In particular, the well-known passage

25
For the programmatic aims of the journal, see the presentation of the journal by Ampolo
and Pucci (1982). Indeed, some of the most famous names of the ancient economy debate
published in Opus, including Jean Andreau, Jean-Michel Carrié, Yvon Garlan, Peter Garnsey,
Jerzy Kolendo, Jean-Paul Morel, Claude Mossé, Dominic Rathbone, André Tchernia, Pierre
Vidal-Naquet, and C. R. Whittaker.
26
Finley (1981).
27
Finley (1982). Among the other contributions to this first volume of Opus relevant to the
issues discussed here is the article of Giardina (1982), which explores, in a very broad manner,
the role of labour in antiquity from an economic, social, political, and also ideological point of
view, with a particular emphasis on late antiquity.
28
Clemente (1991); Traina (2000); Cracco Ruggini (1961, 1995); of special interest for the
present purpose are the very solid chapters dedicated to the negotiatores frumentarii and their
organizations in the era of Ambrosius (pp. 56–146); Carla and Marcone (2011: 33–58 (production)
and 167–211 (trade)).
29
Among the publications with a general reach, making a necessary but in a certain sense
arbitrary selection, one can list Tozzi (1961), Storchi Marino (1979), Gabba (1980), and several of
62 Carla Salvaterra and Alessandro Cristofori

in Cicero, De Officiis I.150–1, in which the author, taking inspiration from a

pre-existing cultural tradition, and particularly from Panaetius of Rhodes, lists
the artes that should be considered sordidae and those that, on the contrary,
can be considered honestae, concluding, in the end, with the usual praise of
agriculture; the passage, because of its abstract thinking, disconnected from
contingencies, systematic and general, has become an almost obligatory point
of departure for any discussion of the issue, in Italy as elsewhere.30 Much
interest has also been evoked by a passage in a letter to Lucilius in which
Seneca discusses Posidonius’ four categories of artes, and by a brief text from
Livy dedicated to the famous plebiscitum Claudianum of 219 or 218 BC,
championed by Caius Flaminius, which in fact prohibited senators from
engaging in maritime trade on a large scale.31 Other detailed studies have
been dedicated to the attestations of occupations in the comedies of Plautus
and Terence, and to a passage of the Bellum Catilinae, in which, surprisingly,
agriculture itself is ranked among the servilia officia.32 Elena Giannarelli,
on the other hand, has focused her attention on Christian texts from Late
Antiquity.33 There also has been a lot of interest in juridical texts, especially
because of the traditional strength of the study of Roman law in Italy. This
interest was initially directed chiefly at analysing to what extent the Roman
state intervened in the economy to regulate it, but later on it has more often
been sociological in character, analysing mostly how Roman legislation re-
flects the socio-economic developments of the time.34
A lot of attention has also been devoted to the epigraphic record, initially
mainly to inscriptions. Besides the short but interesting discussion by Danilo
Mazzoleni about the meaning of references to occupations in Christian com-
memorative epigraphy, research initially focused largely on the reconstruction
of processes of production and trade from occupational inscriptions, but, in
more recent years, more attention has been paid to the social roles of crafts-
men and traders, particularly since the publication of Sandra Joshel’s funda-
mental book on occupation and identity in Rome.35
In the fertile field of epigraphic studies, three main lines of research can
be distinguished. The first consists of scholars studying the epigraphic record

the essays collected in Gabba (1988); Clemente (1984), Lana (1984), De Salvo (1987), and
Giardina (1994).
30
Particularly on this text, which obviously is also discussed in many of the publications listed
in the preceding note, see Gabba (1979); Lotito (1981); and Narducci (1985).
31
For a comment on the passage in Seneca, see, above all, Pani (1985); for the plebiscitum
Claudianum, see, among others, Guarino (1982) and Clemente (1983).
32
On Plautus and Terence, see particularly Gabba (1985) and Lentano (1996). On the passage
by Sallustius, see Bianco (1985).
33
Giannarelli (1986, 1991).
34
See, e.g. Di Porto (1984), Giacchero (1987); Merola Reduzzi (1990); Polichetti (2001); and
Mantello (2008).
35
Mazzoleni (1986); Joshel (1992).
 Twentieth-Century Italian Scholarship 63

of individual occupations, or closely related professions over a wide area

of the entire Roman world—exemplary is the monograph by Ida Calabi
Limentani on artists and artisans in the Roman world, based principally on
the evidence from inscriptions on stone.36 Closely related, because of the
emphasis on the work itself, are general investigations dedicated to a single
occupation, using all kinds of evidence available, as, for example, reflected by
the work of Fabio Vicari on the textile sector, and that of Francesca Diosono
on professions related to wood-working and trade.37 A second line of
research consists of projects that aim to investigate, in a general way,
production and trade in a certain part of the Roman world, especially
in order to reconstruct its economic structure. This line of research has a
long history that goes back at least to an early twentieth-century article by
Aristide Calderini on crafts and occupations in the epigraphy of Gallia
Transpadana.38 Recently, there has been an upsurge in such approaches.39
The last and perhaps most productive line of research concerns investiga-
tions on individual epigraphic documents, which often provide the oppor-
tunity to make more general points about our knowledge of a particular
occupation, or to reconstruct a microhistory of the person involved, which
then can be used as an example for the social roles of people involved in that
specific craft.40
An evolution similar to that detectable in the study of inscriptions on
stone—from an emphasis on production and trade to an increasing emphasis
on the people involved in the processes of manufacturing and trade—can be
found in the studies of the epigraphy on instrumentum domesticum, the study
of which has strongly developed in Italy from the 1960s and 1970s onwards,
with pioneering studies of the epigraphy of amphorae from Ostia by Fausto
Zevi, and, for northern Italy, by Paolo Baldacci.41 In recent decades the study
of texts, stamped or painted, on amphorae, tableware, lamps, glass bottles,
bricks, and other moveable objects, has seen a remarkable development,
mostly through studies dedicated to specific sets of evidence from a limited

36
Calabi Limentani (1958). See also, e.g. Panciera (1980) on the olearii, Chioffi (1999) on
professions related to the production and sale of meat, Mennella (2001) on marble merchants
active in the region of present-day Piemonte, and De Salvo (2004) on the navicularii.
37
Vicari (2001); Diosono (2008).
38
Calderini (1907). Among the older studies, to mention some of the more representative
examples, are Panciera (1957) on Aquileia, and Bivona (1984) on Sicily.
39
See, e.g. Rizzo (1993) on Sicily; Buonopane (2003a) on textile professions in Altinum;
Cristofori (2004) on Picenum.
40
Among other publications of this type, one may note Granino Cecere (1994) on the diffusor
olearius D. Caecilius Abascantus; Mennella (2001) on a negotiator vestiarius; Tisé (2001) on the
inscription of a mercator from Brundisium; Buonopane (2003b) on a vestiarius centonarius from
Aquileia; Boscolo (2005) on the piscatrix Aurelia Nais; and Gasperini (2005) on a materiarius
from Sena Gallica.
41
Zevi (1966); Baldacci (1967–8).
64 Carla Salvaterra and Alessandro Cristofori

part of Roman Italy, though there have also been some contributions of a more
general character.42
Still in the realm of documentary evidence, one should remember also the work
on labour contracts found on Egyptian papyri, which were discussed in the works
of Orsolina Montevecchi and Aristide Calderini in the 1950s, even though perhaps
the most interesting contribution in this field of research is the monograph by
Montevecchi and Mariadele Manca Masciandri on wet-nursing contracts.43
As far as the field of archaeology is concerned, the separation from the
traditional approach, with its emphasis on objects and their aesthetic value,
and in particular on objects that had artistic qualities, was particularly slow in
Italy. A decisive turn was brought about when Dialoghi di Archeologia began
to appear in 1967. This journal was promoted by Ranuccio Bianchi Bandinelli,
assisted by his many pupils, and in particular by Andrea Carandini, Filippo
Coarelli, Mario Torelli, and Fausto Zevi. The separation of Bianchi Bandinelli
from the more traditional settings of Italian archaeology was also motivated by
his adherence to Marxist ideology and was manifest in his important research
on ‘popular’ art in Italy and the provinces, in which the artistic object was
approached as a product of the world in which it was created.44
This change of direction was also upheld after the Dialoghi di Archeologia
ceased to be published in 1992: it was immediately succeeded by the new journal
Ostraka, a direct and explicit continuation, starting under the leadership of
Mario Torelli, one of the principal pupils of Bianchi Bandinelli, and of the jurist
Vincenzo Scarano Ussani, to guarantee the multidisciplinary character of the
journal, which was explicitly emphasized by Torelli in an editorial in the first
volume.45 As a result, sociological approaches to artistic and artisanal produc-
tion became widespread in Italy as well.

42
For detailed studies, see, e.g. Buchi (1979) on ceramic production in the territory of Aquileia;
Marengo (1981) on the production in Picenum from Q. Clodius Ambrosius; Zaccaria (1985) on the
commercial relations between Aquileia and Illyricum; Palazzo and Silvestrini (2001) on the
Brindisian amphorae from Apani, Cipriano and Mazzochin (2002) on some series of Dressel 6B;
Geremia Nucci (2006) on the stamps of the plumbarii of Ostia; and Malfitana (2006) on the possible
falsification of the ‘brand’ Arretinum. More general studies include Taborelli (1982) on glass;
Carandini (1989) on wine amphorae; Manacorda (1989) on the social and economic aspects of
amphorae production in the Republican period; Pucci (1993) on stamped terra sigillata; Nonnis
(2001) on the production and commercialization of Adriatic amphorae and (2003) on the social
implications of the production and commercialization of ceramics in the republican era. Funda-
mental, too, are the proceedings of the conference Epigrafia 1994.
43
Montevecchi (1950); Calderini (1954); Montevecchi and Manca Masciandri (1984); see
also Pintaudi (2001).
44
Fundamental in this respect was the new introduction prepared for the second edition of
Storicità dell’arte classica: Bianchi Bandinelli (1950), first edition Bianchi Bandinelli (1943). The
ideas of Bianchi Bandinelli are most fully developed in Roma: L’arte romana nel centro del potere
(1969) and Roma: La fine dell’arte antica (1970). Fundamental studies on Bianchi Bandinelli are
Barbanera (2000, 2003); see also Barbanera (1998) for a useful synthesis of archaeological studies
in Italy.
45
Torelli (1992).
 Twentieth-Century Italian Scholarship 65

Moreover, attention paid to the remains of products that were mass con-
sumed or had significant economic impact has also increased, even though such
evidence helps us mostly to reconstruct production processes and flows of
commerce rather than the social roles of craftsmen and traders; in this field of
studies, there is one more general discussion, by Tiziano Mannoni and Enrico
Giannicheda, and a quantity of publications dedicated to individual goods,
among which should be remembered the work of Patrizio Pensabene on the
production and circulation of marbles.46 In very recent years, one observes a
progressive shift of interest from products towards producers.47 This is also true
for approaches to workshops, which are now being studied as social and
economic spaces in which the craftsmen and traders operated.48 Limited atten-
tion, on the other hand, compared to what happens on the other side of the
Alps, has been devoted to the iconographic record, though one could mention
the work of Fabrizio Bisconti on labour scenes in early Christian art and, more
recently, the investigation of Sebastiana Mele on Roman Spain.49

PROFESSIONAL ASSOCIATIONS

The theme of professional associations deserves separate treatment. Research in

this area flourished in Italy in the 1930s, in connection with the system of
professional associations established by the Fascist regime.50 Research in this
period focused mainly on the juridical nature of the relation between the state
and the associations, aiming to reconstruct how the Roman state promoted and
controlled the fenomeno associativo, in a way that cannot always be discarded as
apologetic for the Fascist regime: among the enormous amount of studies
published in these years are the first monograph on this theme by De Robertis,
whose nuanced position, free of propagandistic tendencies, was also recognized
by foreign reviewers.51 There was also the publication of a short but rather
explicit lecture by Aristide Calderini, Le associazioni professionali in Roma
antica, in which, with a certain polemical vigour, he insisted on the fundamental

46
Mannoni and Giannichedda (1996); Pensabene (1983, 1990, 1995) are among his most
significant contributions to this theme.
47
See, e.g. Sampaolo (1995) and the essays collected in Santoro (2004).
48
For some publications of the last two decades on this line of research, see n. 61. Specifically
on workshops, see, e.g. Manacorda (1990) on brick/tile kilns, and Taborelli (1998) on glass kilns.
49
Bisconti (2000) and Mele (2008). See also Giuliani (2003), and Marcone (2000), on the
iconography of the Igel column. Cf. Mele (2010).
50
See the bibliography of Frangioni (1998), starting from the index under ‘Roma (antica)’
(p. 515), which helps to find relevant items in this work, in which books are simply ordered by
the surname of their author.
51
De Robertis (1938); cf. De Visscher (1939) and Duff (1942), with the remarks by Perry
(2006: 94–95).
66 Carla Salvaterra and Alessandro Cristofori

differences between the collegia of the Roman world and the labour organiza-
tions of subsequent periods, including the Fascist corporations.52
This ‘juridical’ line of research continued, at a much lower level of intensity,
after the Second World War, when perspectives naturally changed because of
the return of the free labour union and because of the influence of progressive
developments in Catholic thinking. This continuity is represented above all in
the works of De Robertis, which were essentially further developments of his
impressive monograph of 1938.53
From the 1970s onwards, after a rather long period of stagnation, probably
owing to the dangerous association of the Fascist past evoked by the theme—
the silence was only interrupted by the works of De Robertis and by a
contribution from Mario Attilio Levi—a new line of research emerged focusing
more on the nature of the professional associations themselves and on their
socio-economic environment.54 The protagonist of the renaissance of collegia
studies was undoubtedly Lellia Cracco Ruggini, through a series of long articles
that put significant emphasis on Late Antiquity and on the eastern part of the
Roman world, where the documentation allowed for in-depth analysis of
sociological issues that had earlier remained mostly unexplored.55 It was the
merit of Guido Clemente, to have undertaken, for the first time since Waltz-
ing’s synthesis, and following the line of research of Cracco Ruggini, the study
of the patrons of professional collegia.56 Though starting from another his-
toriographical setting, Mario Mazza investigated, in the same years, the polit-
ical value of the professional associations in Roman Asia Minor.57
After the turn brought about by Mazza and by Cracco Ruggini and her
school, the socio-historical approach has remained vital, as expressed by con-
tributions from Giovanni Mennella, Marcella Chelotti, and Filippo Boscolo
(among others) on the evidence from Roman Italy, by the impressive mono-
graph by Lietta de Salvo on the associations of navicularii, by the reinterpret-
ation, in social terms, of the famous inscription of the bakers of Sardis by Marco
di Branco, and, finally, by the useful synthesis of Francesca Diosono.58 This
direction of the debate has, to some extent, also ‘contaminated’ the most recent
work on collegia by scholars specializing in Roman law, as is, for example, clear
from a recent discussion by Giovanna Merola of a possible reference to a lex
collegii related to maritime transporters preserved in the Digesta.59

52
Calderini (n.d.), with the remarks of Perry (2006: 113–14), which do not completely do
justice to the value of what Calderini wrote. Among the other studies of significance from this
period, one can include Monti (1934); Bandini (1937); and Accame (1942).
53 54
De Robertis (1955, 1971). Levi (1963).
55 56
Cracco Ruggini (1971, 1973, 1976a, b). Clemente (1972).
57
Mazza (1974).
58
Mennella and Apicella (2000) provided an epigraphic update of the anthology of Waltzing
for Roman Italy. See further Boscolo (2002, 2004–5); Chelotti (2007); Mennella (2010); De Salvo
(1992); Di Branco (2000); Diosono (2007).
59
Merola (2007).
 Twentieth-Century Italian Scholarship 67

RECENT YEARS

In the most recent research, the general tendency in Italy now seems to be to
put research on craftsmen and traders in the Roman world into closer relation
with the research being conducted in international contexts, and to debates on
periods other than antiquity. Significant in this sense are the contributions
dedicated in the 2000s to the position of artisanal and commercial activities in
their topographical context—a theme already anticipated by Emilio Gabba
and Filippo Coarelli in the 1970s and 1980s.60 Exemplary of this tendency, also
because of where it was published—in the Journal of Roman Archaeology—is
the work of Emanuele Papi on craftsmen and traders around the Forum of
Rome.61
Another fruitful line of research, the development of which has to be seen in
conjunction with the positive reception in Italy as elsewhere of the seminal
monograph by Joshel, focuses on the role that the occupation of craftsman or
trader plays in relation to the juridical and social status of the person who
exercises it. In this respect one can mention the interesting work of Ulrico
Agnati and David Nonnis on the entrepreneurial activities of the ruling classes
of late republican Italy.62
A third area in which Italian scholarship is quickly recovering ground is that
of the meaning of professional activities in the negotiation of gender roles,
which has seen an interesting upsurge of interest since the late 1990s, although
one should not forget the ideas offered by the monographs of Vito Antonio
Sirago and Franca Comucci Biscardi from the 1980s, which focused more
generally on the role of women in the Roman world.63 This culminated in the
conference on women and work in the epigraphic record organized by Fran-
cesca Cenerini and Alfredo Buonopane, and the useful lexicographical work
by Elena Malaspina.64
In the span of over a century of studies on craftsmen and traders in the
Roman world, one can thus confirm that Italian scholarship has, progressively,
integrated itself into international discourse on the theme, and provided a not
insignificant contribution: the hope is that it will be possible to continue in this
direction in the coming years, at the same time preserving the particular
approaches that have characterized Italian classical studies.

60
Gabba (1975); Coarelli (1982).
61
Papi (2002). Along the same line of research, see, e.g. Ortalli (1993); Lo Cascio (2000);
Baratto (2003–4); Proto (2006); and Zevi (2008).
62
Agnati (1996, 1997) and Nonnis (1999). See also Manacorda (1985) and Torelli (1996).
63
Recent approaches include that of a prominent scholar of women’s studies in antiquity,
Cantarella (1999), specifically on the professional activities of women at Pompeii; see also
Morretta (1999) on the role of women in the production and trade of Baetican olive oil. For
earlier approaches on women in general, see Sirago (1983) and Comucci Biscardi (1987).
64
Buonopane and Cenerini (2003); Malaspina (2003).
68 Carla Salvaterra and Alessandro Cristofori

REFERENCES

Accame, S. (1942). ‘La legislazione romana intorno ai collegi nel I secolo a.C.’,
Bulletino del museo dell’impero romano 13: 13–48.
Agnati, U. (1996). ‘Alcune correlazioni tra mestiere e status libertatis nella Roma
tardo-repubblicana e imperiale’, RAL 7: 601–24.
Agnati, U. (1997). Epigrafia, diritto e società: Studio quantitativo dell’epigrafia latina di
zona insubre. Como.
Ampolo, C., and Pucci, G., Opus, 1/1.
Andreau, J., and Leveau, Ph. (1992). ‘Recensione’, Annales (ESC) 47: 402–8.
Antonioli, M. (1997). Il sindacalismo italiano: Dalle origini al fascismo. Studi e ricerche. Pisa.
Baldacci, P. (1967–8). ‘Alcuni aspetti dei commerci nei territori cisalpini’, Atti CeSDIR
1: 5–50.
Bandini, V. (1937). Appunti sulle corporazioni romane. Milan.
Baratto, C. (2003–4). ‘Le tabernae nei fora delle città romane tra l’età repubblicana e il
periodo imperiale’, RdA 27: 67–87; 28: 45–65.
Barbanera, M. (1998). L’archeologia degli italiani: Storia, metodi e orientamenti
dell’archeologia classica in Italia. Rome.
Barbanera, M. (2000). Ranuccio Bianchi Bandinelli e il suo mondo. Bari.
Barbanera, M. (2003). Ranuccio Bianchi Bandinelli: Biografia ed epistolario di un
grande archeologo. Milan.
Bianchi Bandinelli, R. (1943). Storicità dell’arte classica, Florence.
Bianchi Bandinelli, R. (1950). Storicità dell’arte classica. 2nd edn. Florence.
Bianchi Bandinelli, R. (1969). Roma: L’arte romana nel centro del potere. Milan.
Bianchi Bandinelli, R. (1970). Roma: La fine dell’arte antica. Milan.
Bianco, O. (1985). ‘Servilia officia (Sall. Cat. 4, 1)’, Index 13: 127–33.
Bisconti, F. (2000). Mestieri nelle catacombe romane: Appunti sul declino dell’icono-
grafia del reale nei cimiteri cristiani di Roma. Città del Vaticano.
Bivona, L. (1984). ‘Il contributo dell’epigrafia latina allo studio dei mestieri nella Sicilia antica’,
in Atti del II Congresso internazionale di studi antropologici siciliani. Palermo, 25–33.
Bona, F., Canfora, L., Capogrossi Colognesi, L., Carandini, A., Foraboschi, D., Lepore,
E., Panella, C., and Torelli, M. (1982). ‘Discussione’, Opus 1/2: 413–36.
Boscolo, F. (2002). ‘Due iscrizioni di tradizione manoscritta e il collegium centonar-
iorum Comensium’, Patavium 20: 91–105.
Boscolo, F. (2004–5). ‘I battellieri del lago di Como in età romana’, AAPat 117: 221–40.
Boscolo, F. (2005). ‘Aurelia Nais, piscatrix de horreis Galbae e i piscatores di Roma’,
RSA 35: 181–8.
Bretone, M., Canfora, L., Clemente, G., Pani, M., and Silvestrini, M. (1978). ‘Dibattito
su “Analisi marxista e società antiche” ’, QS 4/8: 5–38.
Brutti, M. (1978). ‘Introduzione’, in L. Capogrossi Colognesi, A. Giardina, and
A. Schiavone (eds), Analisi marxista e società antiche. Rome, 9–41.
Buchi, E. (1979). ‘Impianti produttivi del territorio aquileiese in età romana’, AAAd 15:
439–59.
Buonopane, A. (2003a). ‘La produzione tessile ad Altino: Le fonti epigrafiche’, in
G. Cresci Marrone and M. Torelli (eds), Produzioni, merci e commerci in Altino
preromana e romana. Rome, 285–97.
 Twentieth-Century Italian Scholarship 69
Buonopane, A. (2003b). ‘Un vestiarius centonarius ad Aquileia: Sulla genuinità di CIL,
V, 50’, AN 74: 301–14.
Buonopane, A., and Cenerini F. (eds) (2003). Donna e lavoro nella documentazione
epigrafica: Atti del I Seminario sulla condizione femminile nella documentazione
epigrafica. Faenza.
Calabi Limentani, I. (1958). Studi sulla società romana: Il lavoro artistico. Milan and Varese.
Calderini, A. (n.d.). Le associazioni professionali di Roma antica. Milan.
Calderini, A. (1907). ‘Arti e mestieri nelle epigrafi della Gallia Transpadana (special-
mente nelle raccolte milanesi)’, RIL 40: 523–44.
Calderini, A. (1954). ‘Contratti di lavoro di XX secoli fa’, Studi Romani 2: 649–62.
Cantarella, E. (1999). ‘Qualche considerazione sul lavoro femminile a Pompei’, Saitabi,
49: 259–72.
Capogrossi Colognesi, L., Giardina, A., and Schiavone, A. (eds) (1978). Analisi marxista
e società antiche. Rome.
Carandini, A. (1979). L’anatomia della scimmia: La formazione economica della società
prima del capitale. Con un commento alle “Forme che precedono la produzione
capitalistica” dai “Grundrisse” di Marx. Turin.
Carandini, A. (1989). ‘L’economia italica tra tarda repubblica e medio impero con-
siderata dal punto di vista di una merce: Il vino’, in Amphores romaines et histoire
économique: Dix ans de recherche. Rome, 505–21.
Carlà, F., and Marcone, A. (2011). Economia e finanza a Roma. Bologna.
Chelotti, M. (2007). ‘Le associazioni di artigiani, commercianti e artisti nella regio
seconda augustea’, in E. Lo Cascio and G. D. Merola (eds), Forme di aggregazione
nel mondo romano. Bari, 243–58.
Chioffi, L. (1999). Caro: il mercato della carne nell’Occidente romano. Riflessi epigrafici
e iconografici. Rome.
Ciampani, A., and Pellegrini, G. (eds) (2005), La storia del movimento sindacale nella
società italiana: Vent’anni di dibattiti e di storiografia. Soveria Mannelli.
Ciccotti, E. (1899). Il tramonto della schiavitù nel mondo antico. Turin.
Ciccotti, E. (1929). Commercio e civiltà nel mondo antico. Milan; repr. Bologna, 1977.
Ciccotti, E. (1940). Il tramonto della schiavitù nel mondo antico. 2nd edn. Udine.
Ciccotti, E. (1977). Il tramonto della schiavitù nel mondo antico, ed. M. Mazza. Bari.
Cipriano, S., and Mazzocchin, S. (2002). ‘Analisi di alcune serie bollate di anfore Dressel 6B
(Ap. Pulchri, Flav. Fontan e Fontani, L. Iuni. Paetini, L. Tre. Optati)’, AN 73: 305–40.
Clavel-Lévêque, M., and Favory, F. (1980). ‘Recensione’, Labeo 26: 390–401.
Clemente, G. (1972). ‘Il patronato nei collegia dell’Impero Romano’, SCO 21: 142–229.
Clemente, G. (1983). ‘Il plebiscito Claudio e le classi dirigenti romane nell’età dell’im-
perialismo’, Ktèma 8: 253–59.
Clemente, G. (1984). ‘Lo sviluppo degli atteggiamenti economici della classe dirigente
fra il III e il II sec. a.C.’, in W. V. Harris (ed.), The Imperialism of Mid-Republican
Rome. Rome, 165–83.
Clemente, G. (1991). ‘Arti, mestieri, vita associativa, collegia’, in S. Settis (ed.), Civiltà
dei Romani: Il potere e l’'esercito. Milan, 83–91.
Coarelli, F. (1982). ‘L’ “Agora des Italiens” a Delo: Il mercato degli schiavi?’, in
F. Coarelli, D. Musti, and H. Solin (eds), Delo e l’Italia, = Opuscula Instituti Romani
Finlandiae, 2: 119–45.
70 Carla Salvaterra and Alessandro Cristofori
Comucci Biscardi, F. (1987). Donne di rango e donne di popolo nell’età dei Severi.
Florence.
Cracco Ruggini, L. (1961). Economia e società nell’“Italia annonaria”: rapporti fra
agricoltura e commercio dal IV al VI secolo d.C. Milan.
Cracco Ruggini, L. (1971). ‘Le associazioni professionali nel mondo romano-bizantino’,
in Settimane di studio del Centro italiano di studi sull’alto medioevo. XVIII. Artigianato
e tecnica nella società dell'alto medioevo occidentale. 2–8 aprile 1970, vol. i. Spoleto,
59–193.
Cracco Ruggini, L. (1973). ‘Stato e associazioni professionali nell’età imperiale romana’,
in Akten des VI. Internationalen Kongresses für Griechische und Lateinische Epigraphik
München 1972. Munich, 271–311.
Cracco Ruggini, L. (1976a). ‘Collegium e corpus: La politica economica nella legislazione
e nella prassi’, in G. G. Archi (ed.), Istituzioni giuridiche e realtà politiche nel tardo
impero (III–V sec. d.C.): Atti di un incontro tra storici e giuristi. Firenze, 2–4 maggio
1974. Milan, 63–94.
Cracco Ruggini, L. (1976b). ‘La vita associativa nelle città dell’Oriente greco: Tradizioni
locali e influenze romane’, in D. M. Pippidi (ed.), Assimilation et résistance à la culture
gréco-romaine dans le monde ancien: Travaux du VIe Congrès International d’Études
Classiques (Madrid, Septembre 1974). Bucharest and Paris, 463–91.
Cracco Ruggini, L. (1995). Economia e società nell’ “Italia annonaria”: Rapporti fra
agricoltura e commercio dal IV al VI secolo d.C. 2nd edn. Bari.
Cristofori, A. (2004). Non arma virumque. Le occupazioni nell’epigrafia del Piceno.
Bologna.
De Robertis, F. M. (1938). Il diritto associativo romano dai collegi della Repubblica alle
corporazioni del Basso Impero. Bari.
De Robertis, F. M. (1945). Storia sociale di Roma: Le classi inferiori. Bari; repr. Rome,
1981.
De Robertis, F. M. (1955). Il fenomeno associativo nel mondo romano: Dai collegi della
Repubblica alle corporazioni del Basso Impero. Naples.
De Robertis, F. M. (1959). ‘Sulla considerazione sociale del lavoro nel mondo romano’,
in Problemi economici dall’antichità ad oggi: Studi in onore del prof. Vittorio
Franchini nel 75 compleanno. Milan, 54–70.
De Robertis, F. M. (1962). Ancora sulla considerazione sociale del lavoro nel
mondo romano (II: l’ambiente aulico), in Studi in onore di Amintore Fanfani, i.
Milan, 3–37.
De Robertis, F. M. (1963). Lavoro e lavoratori nel mondo romano. Bari.
De Robertis, F. M. (1971). Storia delle corporazioni e del regime associativo nel mondo
romano. Bari.
De Salvo, L. (1987). ‘Il giudizio sulla mercatura nel mondo romano’, AFLM 20: 9–32.
De Salvo, L. (1992). Economia privata e pubblici servizi nell’Impero romano: I corpora
naviculariorum. Messina.
De Salvo, L. (2004). ‘Nuove testimonianze (e riletture) di testi epigrafici relativi a
navicularii e battellieri’, in L. Ruscu (ed.), Orbis antiquus: Studia in honorem Ioannis
Pisonis. Cluj-Napoca, 120–4.
De Visscher, F. (1939). ‘Recensione’, RHE 35: 78–82.
Di Benedetto, V. (1978). ‘Appunti su marxismo e mondo antico’, QS 4/8: 53–97.
 Twentieth-Century Italian Scholarship 71
Di Branco, M. (2000). ‘Lavoro e conflittualità sociale in una città tardoantica: Una
rilettura dell’epigrafe di Sardi CIG 3467 (= Le Bas – Waddington 628 = Sardis VII, 1,
n. 18)’, Antiquité tardive 8: 181–208.
Di Porto, A. (1984). Impresa collettiva e schiavo manager in Roma antica (II sec. a. C –II
sec. d. C.). Milan.
Diosono, F. (2007). Collegia: Le associazioni professionali nel mondo romano. Rome.
Diosono, F. (2008). Il legno: Produzione e commercio. Rome.
Duff, P. W. (1942). ‘Review of F. M. De Robertis, Il diritto associativo romano dai
collegi della Repubblica alle corporazioni del basso impero’, JRS 32: 129–31.
Epigrafia (1994). Epigrafia della produzione e della distribuzione: Actes de la VIIe
rencontre franco-italienne sur l’épigraphie du monde romain organisée par l’Université
de Rome—La Sapienza et l’École française de Rome, Rome 5–6 juin 1992. Rome.
Finley, M. I. (1981). Schiavitù antica e ideologie moderne. Rome and Bari.
Finley, M. I. (1982). ‘Problems of Slave Society: Some Reflections on the Debate’, Opus
1/1: 201–7.
Frangioni, L. (1998). Corporazioni & dintorni: Saggio bibliografico sulle corporazioni e i
gruppi professionali dall’età romana alla fascista (e oltre). Florence.
Gabba, E. (1975). ‘Mercati e fiere nell'Italia romana’, SCO 24: 141–63; repr. in Gabba
(1988), 143–61.
Gabba, E. (1979). ‘Per un’interpretazione politica del “de officiis” di Cicerone’, RAL 34:
117–41.
Gabba, E. (1980). ‘Riflessioni antiche e moderne sulle attività commerciali a Roma nei
secoli II e I a.C.’, in J. H. D’Arms and E. C. Kopff (eds), The Seaborne Commerce of
Ancient Rome: Studies in Archaeology and History (MAAR 36). Rome, 91–102; repr.
in Gabba (1988), 89–105.
Gabba, E. (1982). ‘Per la storia della società romana tardo-repubblicana’, Opus 1/2:
373.
Gabba, E. (1985). ‘Arricchimento e ascesa sociale in Plauto e Terenzio’, Index 13: 5–15;
repr. in Gabba (1988), 69–82.
Gabba, E. (1988). Del buon uso della ricchezza: Saggi di storia economica e sociale del
mondo antico. Milan.
Gagliardi, A. (2010). Il corporativismo fascista. Rome and Bari.
Galeno, G. (1960). ‘Letture’, Labeo 6: 143–4.
Gasperini, L. (2005). ‘Sul materiarius di Sena Gallica (CIL XI, 6212)’, Picus 25: 129–38.
Geremia Nucci, R. (2006). ‘I plumbarii ostiensi: Una sintesi delle nuove evidenze’,
ArchClass NS 7: 448–67.
Giacchero, M. (1987). ‘Il mondo della produzione e del lavoro nell’Edictum de pretiis’,
in Studi in onore di Arnaldo Biscardi, vi. Milan, 121–32.
Giannarelli, E. (1986). ‘Il tema del lavoro nella letteratura cristiana antica: Fra costru-
zione ideologica e prassi letteraria’, in S. Felici (ed.), Spiritualità del lavoro nella
catechesi dei Padri del III–IV secolo. Rome, 213–24.
Giannarelli, E. (1991). ‘Il concetto di lavoro nel monachesimo antico: Temi e problemi’,
Codex Aquilarensis 5: 31–53.
Giardina, A. (1982). ‘Lavoro e storia sociale: antagonismi e alleanze dall’ellenismo al
tardoantico’, Opus 1: 115–46.
Giardina, A. (ed.) (1986). Società romana e impero tardoantico. Bari.
72 Carla Salvaterra and Alessandro Cristofori
Giardina, A. (1994). Il mercante nel mondo romano, in A. Giardina and A.J. Gurevic, Il
mercante dall’antichità al Medioevo, Rome and Bari, 5–59.
Giardina, A., and Schiavone, A. (eds) (1981). Società romana e produzione schiavistica.
Rome and Bari.
Giuliani, R. (2003). ‘Scene di mestiere nelle catacombe: Il restauro del cubicolo dei
Bottai nel cimitero di Priscilla’, MiChA 9: 9–18.
Granino Cecere, M. G. (1994). ‘D. Caecilius Abascantus, diffusor olearius ex provincia
Baetica (CIL VI 1885)’, Epigrafia 705–19.
Gros, P. (1983). ‘Recensione’, Gnomon 55: 609–20.
Guarino, A. (1982). ‘Quaestus omnis patribus indecorus’, Labeo, 28: 7–16.
Horowitz, D. L. (1963). The Italian Labour Movement. Cambridge, MA.
Horowitz, D. L. (1972). Storia del movimento sindacale in Italia. Bologna.
Joshel, S. (1992). Work, Identity and Legal Status at Rome: A Study of the Occupational
Inscriptions. Norman, OK, and London.
Kreissig, H. (1982). ‘Zu einer Diskussion am Gramsci-Institut über sozialökonomische
Probleme des Altertums’, JWG 23/2: 169–76.
Laghi, I. (1990). La rivoluzione sociale del primo sindacalismo: Antologia del sindacalismo
rivoluzionario. Rome.
Lana, I. (1984). L’idea del lavoro a Roma. Turin.
Lentano, M. (1996). ‘Acquisire, conservare, consumare: Spunti per una lettura in
chiave economica degli Adelphoe di Terenzio’, Aufidus 29: 77–100.
Levi, M. A. (1963). ‘Iscrizioni relative a collegia dell’età imperiale’, Athenaeum 41:
384–405.
Lo Cascio, E. (1982). ‘ “Modo di produzione schiavistico” ed esportazioni italiche’,
Opus 1/2: 389–97.
Lo Cascio, E. (ed.) (2000). Mercati permanenti e mercati periodici nel mondo romano:
Atti degli incontri capresi di storia dell'economia antica (Capri 13–15 ottobre 1997).
Bari.
Lotito, G. (1981). ‘Modelli etici e base economica nelle opere filosofiche di Cicerone’, in
Giardina and Schiavone (1981), iii, 80–126.
Luraschi, G. (1982). ‘Per una valutazione della società romana e dell’economia schiavis-
tica’, Opus 1/2: 399–412.
Lussana F. (ed.) (2000). La Fondazione Istituto Gramsci: Cinquant’anni di cultura,
politica e storia. Un catalogo e una guida. Florence.
Malaspina, E. (2003). ‘La terminologia latina delle professioni femminili nel mondo
antico’, Mediterraneo antico 6/1: 347–91.
Malfitana, D. (2006). ‘Appropriazione di copyright, falsificazione o ingannevole messaggio
pubblicitario nel marchio ARRETINVM?: Nota sui rapporti fra archeologia e storia del
diritto romano’, MBAH 25/2: 149–68.
Manacorda, D. (1985). ‘Schiavo manager e anfore romane: A proposito dei rapporti tra
archeologia e storia del diritto’, Opus 4: 141–51.
Manacorda, D. (1989). ‘Le anfore dell’Italia repubblicana: Aspetti economici e sociali’,
in Amphores romaines et histoire économique: Dix ans de recherche. Rome, 443–67.
Manacorda, D. (1990). ‘Le fornaci di Visellio a Brindisi: Primi risultati dello scavo’,
Vetera christianorum 27/2: 375–415.
Mannoni, T., and Giannichedda, E. (1996). Archeologia della produzione. Turin.
 Twentieth-Century Italian Scholarship 73
Mantello, A. (2008). ‘Etica e mercato tra filosofia e giurisprudenza’, SDHI 78: 3–78.
Marcone, A. (2000). ‘Tra archeologia e storia economica: Il mausoleo dei Secundini a
Igel’, Athenaeum 88/2: 485–97.
Marcone, A. (2005). ‘Riflessioni sugli as petti giuridici dell’artigianato romana’, in
M. Polfer (ed.), Artisanat et économie romaine. Montagnác, 7–16.
Marengo, S. M. (1981). ‘I bolli laterizi di Quinto Clodio Ambrosio nel Piceno’, Picus
1: 105–13.
Mazza, M. (1974). ‘Sul proletariato urbano in epoca imperiale: Problemi del lavoro in
Asia minore’, Syculorum Gymnasium 27: 237–78; repr. in Mazza (1986), 75–117.
Mazza, M. (1976). ‘Marxismo e storia antica: Note sulla storiografia marxista in Italia’,
Studi storici 17/2: 95–124.
Mazza, M. (1986). La fatica dell'uomo: Schiavi e liberi nel mondo romano. Catania.
Mazzoleni, D. (1986). ‘Il lavoro nell’epigrafia cristiana’, in S. Felici (ed.), Spiritualità del
lavoro nella catechesi dei Padri del III–IV secolo. Rome, 263–71; repr. in Mazzoleni
(2002), 39–48.
Mazzoleni, D. (2002). Epigrafi del Mondo Cristiano antico. Rome.
Mennella, G. (2000). ‘Un negotiator vestiarius cisalpinus et transalpinus a Fara Novarese’,
Epigraphica 62: 125–35.
Mennella, G. (2001). ‘Evidenze epigrafiche sul commercio del marmo nel Piemonte
romano’, in M. G. Angeli Bertinelli and A. Donati (eds), Varia epigraphica: Atti del
colloquio internazionale di epigrafia. Bertinoro, 8–10 giugno 2000. Faenza, 329–37.
Mennella, G. (2010). ‘Il collegium nautarum e l’integrazione delle risorse forestali
nell’economia di Luna’, in F. Javier Navarro (ed.), Pluralidad e integración en el
mundo romano: Actas del II coloquio internacional Italia Iberia–Iberia Italia,
Pamplona-Olite del 15 al 17 de octubre de 2008. Pamplona, 279–85.
Mennella, G., and Apicella, G. (2000). Le corporazioni professionali nell’Italia romana:
Un aggiornamento al Waltzing. Naples.
Merola, G. D. (2007). ‘Una lex collegii marittima? A proposito di D. 14, 2, 9’, in E. Lo
Cascio and G. D. Merola (eds), Forme di aggregazione nel mondo romano. Bari,
259–72.
Merola Reduzzi, F. (1990). ‘Servo parere’: Studi sulla condizione giuridica degli schiavi
vicari e dei sottoposti a schiavi nelle esperienze greca e romana. Naples.
Mondolfo, R. (1965). ‘Tecnica e scienza nel pensiero antico’, Athenaeum 43: 279–94.
Montevecchi, O. (1950). I contratti di lavoro e di servizio nell’Egitto greco-romano e
bizantino. Milan.
Montevecchi, O., and Manca Masciandri, M. (1984). I contratti di baliatico. Milan.
Monti, G. M. (1934). Le corporazioni nell’evo antico e nell’alto medioevo. Bari.
Morelli, F. (2011). ‘Dal Mar Rosso ad Alessandria: il verso (ma anche il recto), del
“papiro di Muzirus” (SB XVIII 13167)’, Tyche 26: 149–233.
Morretta, S. (1999). ‘Donne imprenditrici nella produzione e nel commercio dell’olio
betico (I–III sec. d.C.)’, Saitabi 49: 229–45.
Musca, D. A. (ed.) (2007). Francesco Maria De Robertis: L’uomo, il docente, lo studioso:
Atti di un Incontro di studio. Bari, 20 novembre 2004. Bari.
Musso, S. (2002). Storia del lavoro in Italia: Dall’Unità a oggi. Venice.
Napoli, M. (ed.) (2007). Alle radici del sindacalismo italiano. Milan.
Narducci, E. (1978). ‘Note in margine a due libri recenti’, QS 4/8: 39–50.
74 Carla Salvaterra and Alessandro Cristofori
Narducci, E. (1985). ‘Valori aristocratici e mentalità acquisitiva nel pensiero di Cicerone’,
Index 13: 93–125.
Nicolet, C., Andreau, J., Vallet, G., Mossé, C., and Morel, J.-P. (1982). ‘Interventi su
Società romana e produzione schioavistica, a cura di A. Giardina e A. Schiavone’, QS
8/16: 287–323.
Nonnis, D. (1999). ‘Attività imprenditoriali e classi dirigenti nell’età repubblicana: Tre
città campione’, CCG 10: 71–109.
Nonnis, D. (2001). ‘Appunti sulle anfore adriatiche di età repubblicana: Aree di
produzione e di commercializzazione’, AAAd 46: 467–500.
Nonnis, D. (2003). ‘Le implicazioni socio-politiche della produzione e della distribuzione
nell’Italia repubblicana: Per un repertorio prosopografico’, in C. Zaccagnini (ed.),
Mercanti e politica e nel mondo antico. Rome, 245–74.
Nörr, D. (1965). ‘Zur sozialen und rechtlichen Bewertung der freien Arbeit in Rom’,
ZRG 82: 67–105.
Ortalli., J. (1993). Mestieri, merci, mercati: Luoghi e memorie per l’economia di Bologna
romana, in R. Scannavini (ed.), Piazze e Mercati nel centro antico di Bologna: Storia
e Urbanistica dall’età romana al medioevo, dal Rinascimento ai giorni nostri.
Bologna, 251–81.
Palazzo, P., and Silvestrini, M. (2001). ‘Apani: Anfore brindisine di produzione
“aniniana” ’, Daidalos 3: 57–107.
Panciera, S. (1957). Vita economica di Aquileia in età romana. Aquileia.
Panciera, S. (1980). ‘Olearii’, in J. H. D’Arms and E. C. Kopff (eds), The Seaborne
Commerce of Ancient Rome: Studies in Archaeology and History (MAAR 36). Rome,
235–50.
Pani, M. (1985). ‘La polemica di Seneca contro le artes (Ep. 90): Un caso di sconcerto’,
in Xenia: Scritti in onore di Piero Treves. Rome, 141–50; repr. in Pani (1992),
99–112.
Pani, M. (1992). Potere e valori a Roma fra Augusto e Traiano. Bari.
Papi, E. (2002). ‘La turba inpia: Artigiani e commercianti del foro Romano e dintorni
(I sec. a.C.–64 d.C.)’, JRA 15/1: 45–62.
Pasoli E. (1980). ‘Scienza e tecnica nella considerazione prevalente del mondo antico:
Vitruvio e l’architettura’, in Scienza e tecnica nelle letterature classiche. Sassari, 63–80.
Pensabene, P. (1983). ‘Osservazioni sulla diffusione dei marmi e sul loro prezzo nella
Roma imperiale’, DArch 1/1: 55–63.
Pensabene, P. (1990). ‘Trasporto, diffusione e commercio dei marmi: Aggiornamenti e
nuove interpretazioni’, Journal of the European Study Group on Physical, Chemical,
Biological & Mathematical Techniques Applied to Archaeology 27: 231–64.
Pensabene, P. (1995). Le vie del marmo: I blocchi di cava di Roma e di Ostia. Il
fenomeno del marmo nella Roma antica. Rome.
Perry, J. (2006). The Roman collegia: The Modern Evolution of an Ancient Concept.
Leiden.
Pintaudi, R. (2001). ‘Quattro contratti di trasporto verso Panopolis’, in T. Gagos and
R. S. Bagnall (eds), Essays and Texts in Honor of J. David Thomas. Oakville, CT,
211–15.
Polichetti, F. (2001). Figure sociali, merci e scambi nell’Edictum Diocletiani et collegarum
de pretiis rerum venalium. Naples.
 Twentieth-Century Italian Scholarship 75
Proto, F. (2006). ‘Tabernae, officinae ed altri impianti a carattere commerciale della
Regio I di Pompei: Un campione d’indagine socio-economica’, Rivista di studi
pompeiani 17: 15–28.
Pucci, G. (1993). ‘I bolli sulla terra sigillata: Fra epigrafia e storia economica’, in
W. V. Harris (ed.), The Inscribed Economy: Production and Distribution in the
Roman Empire in the Light of instrumentum domesticum: The Proceedings of a
Conference Held at the American Academy in Rome on 10–11 January, 1992 (JRA
Supplementary Series, 6). Ann Arbor, 73–9.
Quadrato, E. (2004). ‘Francesco Maria De Robertis: Appunti di un incontro di studio’,
Rivista di Diritto Romano 4: 1–8 <http://www.ledonline.it/rivistadirittoromano/
allegati/dirittoromano04quadrato.pdf> (accessed 21 July 2015).
Rathbone, D. W. (1983). ‘The Slave Mode of Production in Italy’, JRS 73: 160–8.
Rizzo, F. P. (1993). La menzione del lavoro nelle epigrafi della Sicilia antica (Per una
storia della mentalità). Palermo.
Salvaterra, C. (2006). ‘Labour and Identity in the Roman World: Italian Historiography
during the Last Two Decades’, in B. Waldijk (ed.), Professions and Social Identity: New
European Historical Research on Work, Gender and Society. Pisa, 15–38.
Salvioli, G. (1906). Le Capitalisme dans le monde antique: Études sur l’histoire de
l’économie romaine. Paris.
Salvioli, G. (1912). Der Kapitalismus im Altertum: Studien über die römische
Wirtschaftsgeschichte. Stuttgart.
Salvioli, G. (1929). Il capitalismo antico: Storia dell’economia romana. Bari.
Salvioli, G. (1985). Il capitalismo antico: Storia dell’economia romana, ed. A. Giardina.
Rome and Bari.
Sampaolo, V. (1995). ‘I decoratori del Tempio di Iside a Pompei’, MNIR 54: 200–13.
Santoro, S. (ed.) (2004). Artigianato e produzione nella Cisalpina, i. Proposte di metodo
e prime applicazioni. Florence.
Silvestrini, M. (1989). ‘Recensione’, QS 15/30: 199–208.
Sirago, V. A. (1983). Femminismo a Roma nel Primo Impero. Soveria Mannelli.
Solazzi, S. (1906). Il lavoro libero nel mondo romano. Macerata.
Solazzi, S. (1955). Scritti di diritto romano, i. Naples.
Storchi Marino, A. (1979). ‘Artigiani e rituali religiosi nella Roma arcaica’, RAAN 54:
333–57.
Taborelli, L. (1982). ‘Vasi di vetro con bollo monetale (Note sulla produzione, la
tassazione e il commercio degli unguenti aromatici nella prima età imperiale)’,
Opus 1/2: 315–40.
Taborelli, L. (1998). ‘Un antico forno vetrario ad Ancona’, Picus 18: 219–24.
Thébert, Y. (1982). ‘Recensione’, Annales ESC 37: 788–91.
Tisé, B. (2001). ‘Osservazioni su CIL IX, 60’, in S. Alessandrì and F. Grelle (eds), Dai
Gracchi alla fine della Repubblica: Atti del V convegno di studi sulla Puglia romana.
Mesagne, 9–10 aprile 1999. Galatina, 155–70.
Torelli, M. (1992). ‘Editoriale’, Ostraka 1/1: 3–5.
Torelli, M. (1996). ‘Industria laterizia e aristocrazie locali in Italia: appunti prosopografici’,
CCG 7: 291–6.
Tozzi, G. (1961). Economisti greci e romani: Le singolari intuizioni di una scienza
moderna nel mondo classico. Milan.
76 Carla Salvaterra and Alessandro Cristofori
Traina, G. (2000). ‘I mestieri’, in A. Giardina (ed.), Storia di Roma dall’antichità a oggi.
Roma antica. Rome, 113–31.
Turone, S. (1992). Storia del sindacato in Italia: Dal 1943 al crollo del comunismo.
Rome and Bari.
Varni, A. (1992). Il sindacato nella società italiana fra ’800 e ’900. Florence.
Vallauri, C. (2008). Storia dei sindacati nella società italiana. Rome.
Vegetti, M. (1977). Marxismo e società antica. Milan.
Vicari, F. (2001). Produzione e commercio dei tessuti nell’Occidente romano (BAR
International Series, 910). Oxford.
Zaccaria, C. (1985). ‘Testimonianze epigrafiche dei rapporti tra Aquileia e l’Illirico in
età imperiale romana’, AAAd 26/1: 85–127.
Zevi, F. (1966). ‘Appunti sulle anfore romane’, ArchClass 18.2: 208–47.
Zevi, F. (2008). ‘I collegi di Ostia e le loro sedi associative tra Antonini e Severi’, in
C. Berrendonner, M. Cébeillac-Gervasoni, and L. Lamoine (eds), Le Quotidien
municipal dans l’Occident romain. Clermont-Ferrand, 477–505.
 3

The Archaeology of Roman

Urban Workshops: A French Approach?
Jean-Pierre Brun

Since 2000, the Centre Jean Bérard, the Naples-based research centre of the
Centre National de la Recherche Scientifique and the École Française de
Rome, has been developing a research programme on ancient urban work-
shops in southern Italy, focusing principally on Pompeii and Herculaneum,
but also including other excavations such as those at Saepinum, Paestum, and
Levanzo in Sicily.1 The aim of the programme is to advance our understanding
of certain artisanal activities that can be particularly difficult to identify in
places where the evidence is less well preserved, in order to propose criteria for
interpreting those poorly preserved remains.
The programme comes from a tradition of study that has been evolving in
France since the beginning of the twentieth century and that has been renewed
by the development of rescue archaeology since 1990.

ROMAN URBAN CRAFTS I N F RENCH S TUDIES

Even though the history of technology in antiquity has long been limited by
the scarce availability of written sources that deal mainly with agriculture and,
in France, by the non-participation of scholars in the theoretical debate on the
nature of the ancient economy that occurred in Germany at the beginning of
the twentieth century, the French did develop a specific interest in the artisanal

1
This programme involves colleagues from the Centre Jean Bérard, the Centre Camille
Jullian of CNRS (Philippe Borgard), the CNRS (with Magali Cullin-Mingaud, Marie T. Libre),
the École française de Rome (Nicolas Monteix, Emmanuel Botte), the Centre Archéologique du
Var (Martine Leguilloux). It is supported by a grant from the Ministère des Affaires Étrangères
and by a financial support from the Agence Nationale de la Recherche (programme Artifex).
78 Jean-Pierre Brun

world using literary, epigraphic, and archaeological sources. These studies

started at the turn of the twentieth century with some excellent articles
published in Daremberg and Saglio’s Dictionnaire des antiquités and some
syntheses about the history of Greek economic activities.2 After the First
World War, they were followed by the grand syntheses on Africa by Stéphane
Gsell and by Camille Jullian’s work on Gaul that explored economic issues and
the state of the question at that time.3
In the 1930s, the study of the history of technology became particularly
fashionable. Historians of this period, fascinated by the take-off of modern
innovations, wanted to study technological development and its relationship
with historical and social development. This was one of the main projects of
the Annales School, clearly defined by Lucien Febvre in a manifesto published
in 1935.4 The historians from this school made some important contributions,
but these mainly concerned the medieval period, and they greatly underesti-
mated the role of ancient technology. As far as antiquity was concerned,
they followed the opinions of one specialist, Commandant Lefebvre des
Noëttes, a cavalry officer who argued, in a series of books published between
1924 and 1932, that ancient technology was deeply deficient, and explained
the development of slavery as a result of these shortcomings. The title of one
of his books—L’Attelage et le cheval de selle à travers les âges: Contribution à
l’histoire de l’esclavage—is highly significant.5
From the beginning of the twentieth century, archaeological excavations
started to return increasing amounts of evidence, which Albert Grenier sum-
marized for Gaul in his 1934 Manuel d’archéologie gallo-romaine. In this book,
he compared the small-scale artisanal activities located in towns and villages to
large-scale industries, quarries, ceramic productions, mines, and metallurgical
workshops located in the countryside close to the natural resources they
exploited.
Three years later, the same author published a synthesis in the series of the
Economic Survey of Ancient Rome edited by Tenney Frank.6 In his introduc-
tion on the working methods he intends to follow there is no allusion to the
ongoing debate about the nature of the ancient economy: he exposed a careful
and empirical process based mainly on texts, considering that ‘in economic
matters, the archaeological evidence needs, as well as inscriptions, a certain
degree of generalization, but this is often arbitrary’; thus he planned to use
archaeology only when the interpretation of the remains was well established.
Significantly, he insisted on the use of evidence from pottery production, and

2
Daremberg, Saglio, and Pottier (1877–91); Guiraud (1900); Francotte (1900–1).
3 4
Gsell (1913–28); Jullian (1920–6). Febvre (1935: 531–5).
5
Lefebvre des Noëttes (1924, 1931, 1935). It is well known that these theories have been
abandoned—see, e.g. Amouretti (1991) and Rapsaet (2002).
6
Grenier (1937: esp. 382 (for the method) and 586–7).
 The Archaeology of Roman Urban Workshops 79

he mentions textile production, but only through written sources.7 Thus, the
scarcity of texts and archaeological remains, with the notable exceptions of
pottery and metalworking, have meant that artisanal research was almost
absent in French studies of the ancient economy until the mid-1980s.
From the 1950s to the 1970s, a period that witnessed the destruction of
many archaeological sites during post-war reconstruction, there was however
a slow accumulation of data, and several large syntheses were published,
including L’Histoire générale du travail, L’Histoire générale des techniques,
Le Dictionnaire archéologique des techniques, and Histoire des techniques.8
They followed the pre-war theories that postulated a stagnation of techno-
logical development during antiquity and they integrated archaeological docu-
mentation into the discourse but only through iconographic representations
and artefacts, mainly in stone, metal, or terracotta. The artisan himself was
presented through the filter of the aristocratic mentality reflected by the
surviving literature, and the workshops were not discussed. Scholars stood
on the threshold of studying the productive installations, but they did not
enter the subject.
Changes came with the 1980s. In the Histoire de la France urbaine, whose
first volume was published in 1980, the economic role of towns was scrutin-
ized. Venceslas Kruta’s presentation of Iron Age oppida insisted on their role
as artisanal centres, while Christian Goudineau argued that towns in Roman
Gaul were parasitic on the countryside.9 The latter picture, which used all the
archaeological data available at the time, is still impressive. According to
Goudineau, the role played by towns as production centres was limited to
local needs and the overall urban output was low, even if he recognizes that it
is impossible to determine the proportion of urban to rural production. As the
commercial role of cities was limited mainly to local trade between town and
country—with the exception of towns located along the main trade routes:
Mediterranean and Rhône–Rhine axes—his conclusion is clear: the town in
antiquity was definitely a ‘consumer city’.
Soon after, Philippe Leveau expounded another interpretation published in
several articles.10 He argued that towns and their elites organized the coun-
tryside and improved agricultural production through their estates through a
new type of management, introducing new crops and farming methods. In this
sense, ancient cities were not parasitic: they paid for their provisions through
services. But the role played by urban production was not considered as
important. In Leveau’s doctoral thesis on Caesarea of Mauretania, which
provides a wide range of new data, especially on the countryside, the city is

7
Grenier (1937) devoted twenty-two pages to pottery, four to metallurgy, and two to textile
production, and does not even mention leather production.
8
Rémondon (1959); Duval (1962); various authors (1963–4); Gilles (1978).
9 10
Février et al. (1980: 206–10, 365–81). Leveau (1983); Février and Leveau (1982).
80 Jean-Pierre Brun

reckoned to be a political town whose purpose was to organize and administer

the territory.11
Another approach emerged at the end of the 1970s, with scholars such as
Marie-Claire Amouretti for the Greek period, Jean-Paul Morel for the Roman
Republic, and Jean-Pierre Sodini for late antiquity. They paid attention to
technical processes, to social and cultural environments, and to the artisans
themselves as men, trying to re-create their technical movements within the
chaîne opératoire and to understand their daily lives.12 They all insisted on the
interdependence between agriculture and craftsmanship. The craftsmen pro-
vided the necessary tools and devices to peasants and landowners, from iron
tools to jars and amphorae for storing and trading the crops. The agricultural
development of antiquity is inconceivable without the parallel development of
artefacts produced in workshops located in cities or villages. Jean-Paul Morel
recently argued that artisanal products are indispensable for farming, especially
for advanced and highly productive farming techniques, and, further, that in
some regions these products were made principally in towns, where in some
cases these activities were undertaken at manufactories.13
Research into artisanal establishments really took off only from the end of
the 1980s, following the development of rescue archaeology. These excava-
tions shed new light on artisanal activities that encouraged the development of
research programmes on themes such as urban and rural metallurgy or on
well-preserved sites such as those destroyed by Vesuvius.14 New data were
presented at several conferences: at Lyons in 1996 and 2002, at Erpeldange in
1999, 2001, and 2004, at Autun in 2007, at Aix-en-Provence in 2007, at Rome
in 2009, at Naples in 2009.15 Several workshop monographs were also pub-
lished.16 Partial syntheses are now available—for example, on the professional
associations by Nicolas Tran, published in 2006, on the fish-salting industry in
Italy by Emmanuel Botte in 2009, and on the workshops of Herculaneum and
Pompeii by Nicolas Monteix in 2011.17

DEFINING THE ‘ ARTISAN’

A key question to ask is what type of activities we intend to study when we are
looking at artisanal activities. Several French scholars have attempted to define

11 12
Leveau (1984). Amouretti et al. (1977); Morel (1976, 1992); Sodini (1979).
13 14
Morel (1985, 1996, 2009). Mangin (1981, 1985).
15
Various authors (1996); Béal and Goyon (2002); Polfer (1999, 2001, 2005); Brun (2009);
Chardron-Picault (2010); Fontaine, Satre, and Tekki (2011); Monteix and Tran (2011).
16
Many publications of rescue excavations present workshops of various types: e.g. Rivet
(1992) and Chardron-Picault and Pernot (1999).
17
Tran (2006); Botte (2009); Monteix (2011).
 The Archaeology of Roman Urban Workshops 81

this area of research, but within very different boundaries. For Jean-Pierre
Sodini, writing in 1979, the word ‘artisan’ included all professional activities
except agriculture, food preparation, and intellectual professions. In his essay on
‘the artisan’ published in 1992, Jean-Paul Morel did not give a precise definition,
mentioning only the disdain of aristocratic circles for craftsmen recorded by
Cicero and Seneca. In 2005, Alain Ferdière, independently of Sara Santoro,18
proposed to limit the field of artisanal activities to craftsmanship, excluding
artisans who prepared food and those involved in construction (especially
builders).19 In two pieces published in 2011, Nicolas Monteix and Jean-Paul
Morel revisited the argument. Monteix sees Ferdière’s and Santoro’s definitions
as too restrictive and decides not to use the term artisanat because of its
ambiguity and anachronism, preferring instead the broader term métier.20
Jean-Paul Morel recognized the problem: ‘définir strictement [l’artisanat] est
devenu un exercice sinon nécessaire, du moins obligé: exercice redoutable pour
l’Antiquité, et pas seulement pour elle.’ He proposed the following definition:
‘the artisan is a man who, apart from agricultural work, makes something, using
mainly his hands and working alone, or with the help of a small team.’21 Such a
definition includes workers making products, such as potters, producing food,
such as bakers, or providing services, such as barbers, but also those undertaking
domestic activities such as weaving for one’s own family.
Personally, I think that we should consider as ‘artisanal’ any type of
productive activity that is carried out in a non-domestic context and paid
for by a customer or a patron in money or kind. Using such a broad definition
enables a wider investigation of the entire range of urban economic activities,
including the preparation of food, building activities—from the extraction of
stone in the quarries to construction work—and obviously all craftsmanship.
It may also include workshops in villae that were actually enterprises.
The archaeology of these activities should be studied with the same atten-
tion as that previously paid to aristocratic houses, sanctuaries, or necropoleis,
because these remains are often the only key to the infrastructure of ancient
civilization. Ancient literature written by aristocrats presents the upper-class
vision of the working world, and there are very few inscriptions mentioning
craftsmen compared to the thousands of inscriptions dealing with the careers
of the aristocrats or with the worship of gods. Archaeological remains are the
only evidence left by the lower classes, who had neither the power nor the
intellectual means to transmit the memory of their activities and their way of
life. These remains are also poorly preserved, and thus difficult to identify.
Furthermore, their location, often outside city centres, means that they are
sometimes ignored: there is thus a need to study them more carefully than has
previously been done.

18 19 20
Santoro (2004: 24–6). Ferdière (2005). Monteix (2011: 8–9).
21
Morel (2011).
82 Jean-Pierre Brun

THE ‘ ARTIFEX ’ PROGRAMME

Consequently, the aim of our research at Pompeii, and in southern Italy,

was first and foremost archaeological and practical, rather than historical.
Generally speaking, ancient written sources paid little attention to artisanal
activities and to their contribution to economic life, if we exclude the produ-
cers of luxurious items. Archaeology is barely able to compensate for this lack
because many activities used perishable materials, and thus their importance
remains underestimated. In any book about the archaeology of artisanal
productions pottery, metalworking, and glassmaking take the main stage
because they leave durable remains such as kilns, pottery sherds, slag, and
fused glass.22 On the other hand, weavers and dyers, tanners and cobblers,
wood-workers and carpenters, basket-makers and perfumers, are almost
absent. Their installations are difficult to characterize, and they used materials
that are no longer present. If we want to present a more balanced view of
artisanal production within the ancient economy in the future, archaeologists
must learn how to recognize the remains of these activities.

Workshops at Pompeii and Herculaneum

The excavated remains of the towns buried in the AD 79 eruption of Vesuvius

give us a unique opportunity to study these remains. Obviously the necessary
observations and analyses were not done during the initial excavations, but the
state of preservation is still much better than anywhere else. The catastrophe
fixed as many as 270 workshops in time, which contained tools, rubbish, and
sometimes artefacts or graffiti. In many cases they can be clearly identified and
interpreted.

Dyeing Plants

Four wool-washing and dyeing plants that were in use in AD 79 were cleaned
and excavated.23 The first ones, probably belonging to the same complex, were
installed after the earthquake of 62 or 63, in two shops V 1, 4 and V 1, 5
opening onto the Via di Nola. At the entrance, both of them present a large
boiler lined with lead that was used for degreasing wool with alum, together
with several other boilers used for dyeing either in heating solutions or in
fermenting ones. Reconstruction of the boilers and dyeing experiments in

22
e.g. Fontaine, Sartre, and Tekki (2011).
23
Borgard (2002); Borgard and Puybaret (2004).
 The Archaeology of Roman Urban Workshops 83

2008 and 2009 helped further understanding of the various features.24 The
Casa della Regina d’Inghilterra (VII 14, 5.17–19) facing onto the Via dell’-
Abbondanza was studied in 2001–2: in the garden of the house, after the
earthquake of 62 or 63, two working complexes were built (see Fig. 3.1). The
first one is composed of a boiler and two benches in limestone used to clean
the wool. The second part is equipped with nine boilers for dyeing the fibres.
A similar association is visible in workshop I 8, 19, which was studied between
2003 and 2006: a semi-underground room was used to prepare the wool,
which was dyed in an adjacent ground-floor space following the processes
already mentioned. Further investigations were undertaken in workshop VII
12, 23, which is unique for its probable wool-cleaning installation, with its two
types of boilers (high and low). In parallel, the textile remains that survived
through carbonization were studied.25

Perfumeries

Perfumeries are usually difficult to identify, because their remains are not at all
characteristic except when a workshop located near the centre of a town is

Fig. 3.1. The dye plant situated in house VII 14, 5 at Pompeii studied by Ph. Borgard.
The photo shows nine heaters for dyeing in several colours, two basins, and a
workshop for cleaning the textile fibres. Photo: Jean-Pierre Brun.

24 25
Alfaro et al. (2011). Study by Fabienne Médard, not yet published.
84 Jean-Pierre Brun

Fig. 3.2. Perfume shop in the Via degli Augustali at Pompeii excavated in 2002. The
vats belong to oil presses destroyed by the AD 62/63 earthquake. Photo: Jean-Pierre Brun.

equipped with a press-bed. Wilhelmina Jashemski, in her book about the

gardens of Pompeii, and later David Mattingly, in an article published in
1990, suggested that an oil press found in a shop opening off the Via degli
Augustali (VII 4, 24) belonged to a perfume workshop.26 Excavating this shop
in 2001, I found two large tanks underneath the floor laid after AD 65, which
residue analysis proved to have contained vegetal oils, and also a boiler
nearby, probably used for the preparation of the perfumes (see Fig. 3.2).27
In May 2011, in collaboration with the Spanish team of Valencia, we excavated
three other shops along the street, and were able to identify more perfume
shops.28

Tanneries

A large tannery, occupying the main part of the fifth block of Regio I, was
excavated in the years 1873 and 1874; an architectural and stratigraphic study
was carried out between 2001 and 2010. During the second quarter of the first
century AD, two tanneries were installed in existing houses. Then, the great
earthquake of AD 62 or 63, having severely damaged the houses, offered the
opportunity for someone to buy the entire block and to create a single large

26 27
Jashemski (1979: 276); Mattingly (1990). Brun and Monteix (2009).
28
Brun et al. (2012).
 The Archaeology of Roman Urban Workshops 85

Fig. 3.3. Balloon photograph of the tannery I 5 at Pompeii, showing the work area in
the peristyle, and the tanning room in the lower right corner of the building. Photo:
Jean-Pierre Brun.

tannery equipped for washing the skins and tanning them in a set of fifteen
cylindrical pits (see Fig. 3.3).29

Basket-making

A study of ancient basket-making recently published presents the excavations

conducted in the house I 14, 2 where bunches of carbonized flexible shoots
and a long tank used to soak them were discovered (see Fig. 3.4). The results of
the excavations have been integrated in a full synthesis on what we know
about this rarely detectable craft.30

We have thus progressed in our understanding of some branches of

ancient technology not limited to these four examples. The programme also
studied plumbers’ workshops (Nicolas Monteix), blacksmiths’ establishments
(Marie-Pierre Amarger), painters’ workshops (Marie Tuffreau-Libre), colour
workshops and pottery workshops (Laëtitia Cavassa), and bakeries (Monteix).31

29 30
Leguilloux (2004). Cullin-Mingaud (2010).
31
Borgard et al. (2005); Amarger and Brun (2007). Pompeian bakeries are discussed in
Monteix, Chapter 7, this volume.
86 Jean-Pierre Brun

Fig. 3.4. The tank used for basket-making in house I 14, 2 at Pompeii studied by
M. Cullin-Mingaud. Photo: Jean-Pierre Brun.

Results

As a result of the programme, it is now possible to describe artisanal activities,

such as, the entire wool- or skin-making cycle, and illustrate them with photos
and plans. Taking into account observations made on the well-identified work-
shops from Pompeii and Herculaneum, we extended our enquiries to other
areas of southern Italy. For example, the excavations of the tannery of Pompeii
helped us to interpret other less well-preserved tanning installations known in
Italy, such as at Pompeii itself,32 at Saepinum (see Fig. 3.5), and under the
church of Santa Cecilia at Rome.33 It is obvious that tanneries were common,
with several workshops of this type existing in each town, but our knowledge
does not yet allow the presentation of a chronological synthesis of this industry.
The aim is to understand what traces would remain if the walls, instead of
being 1 or 2 metres high, were preserved at only ground level or, worse, at the
level of the foundations. To give some examples, archaeologists can identify a
perfume factory thanks to press-beds, tanks, boilers, and broken perfume

32
Excavations directed by Steven Ellis and Gary Devore: Devore and Ellis (2008: 9 and fig. 21).
33
Parmigiani and Pronti (2004) reject the interpretation of the cylindrical tanks as tanning
pits and propose identifying them as silos, which is not possible as the silos are bottle-shaped for
chemical reasons. But the traditional interpretation has strong support: see Coarelli (1974: 309,
316); Breccia Fratadocchi (1976: 217–22). A full description of the tannery of Saepinum is
provided by Brun and Leguilloux (2014).
 The Archaeology of Roman Urban Workshops 87

Fig. 3.5. Balloon photograph of the water-mill, used to produce tannin by crushing
bark at Saepinum, Molise. Photo: Jean-Pierre Brun.

flagons. A tannery might leave tools, cylindrical tanks, alum amphorae from
Lipari, and specific bone assemblages. A dye plant could be equipped with
tanks and various types of boilers lined with lead; because of the work done by
the programme, it was possible easily to identify such an installation in a
recent excavation at Frejus.34 A wool-cleaning installation is equipped at
Pompeii with a concrete or stone bench and with at least one heater where
the wool fibres were washed in hot water containing roots of soapwart
(saponaria officinalis). These remains are characteristic enough to be identi-
fied everywhere.35 For example, in a rescue excavation I directed in 1997 in the
Roman village of Pignans (Var), I found benches, hearths, and channels that
were the remains of a wool-washing workshop.

THE P LACE OF ARTISANAL ACTIVITIES I N

EC ONOMIC LIFE

Excavating various workshops even from a strictly archaeological point of

view cannot be disconnected from research into the evolution of economic life
at Pompeii and its place within the economy of Italy at the beginning of the
empire, assuming that Pompeii was representative of a wealthy medium-sized

34
Pasqualini et al. (2005–6).
35
I will not enter here into the debate on the identification of these installations. Moeller
(1976: 30–5) thought these were used for washing wool; Jongman (1991: 168–9) contested this
interpretation, thinking that they were perhaps linked with some food preparation installations,
and Flohr (2013a) has recently defended this idea; Borgard and Puybaret (2004) are convinced
that they are installations for cleaning and washing wool, two of them being associated with dye
plants. Monteix (2011: 170–2; 2013) also follows Moeller’s interpretation.
88 Jean-Pierre Brun

town. As a result of our work, we have now a less dramatized vision than
Amedeo Maiuri’s theories would lead us to believe.36 Most of the known
workshops were actually built after the earthquake, but these economic activ-
ities often succeeded previous activities of the same nature. Recent research
shows a complex and versatile situation: some workshops were repaired and
adapted after the earthquake, others, such as the tannery, were totally trans-
formed, and yet further workshops were built inside aristocratic houses, which
were sold or rented by their previous owners.
Having changed the perception of the chronology, we have to assess the
place of artisanal activities within economic life. Did the artisans of Pompeii
serve only the needs of the inhabitants of the town and its territory or was part
of the production exported towards larger cities such as Rome? The question is
pointless for builders, plumbers, painters, or bakers, but worth asking for
products such as perfumes, fish sauce, leather, or textiles.
The latter activity was the focus of a polemic between the supporters of
relatively large-scale production for export and the supporters of production
limited to the needs of the inhabitants of the towns and the countryside. The
first option is supported by Moeller, who thought that this industry was
controlled by the fullers and involved the local elites.37 Willem Jongman and
recently Miko Flohr have rightly refuted this hypothesis.38 Jongman con-
cluded that the wool washed and dyed at Pompeii was not exported. Philippe
Borgard, on the contrary, has argued that the wool could have brought a
significant income to some artisans and traders. Does the dispersion and the
size of washing and dyeing plants and the absence of large factories indicate
locally focused activity? We must recall that medieval towns with an export
textile industry did not, in fact, have larger or more concentrated workshops
and, using archaeology without the help of the archives, it would have been
impossible to estimate their nature and productive output.
We face the same questions with leather production. The large tannery of
Pompeii is clearly a manufactory employing at least twenty more or less
specialized workers to crush the bark, carry out the beamhouse work,
manipulate the hides in the pits, lift the water, dry, cut, and grease the leather.
With its fifteen large tanning pits, its size can be compared with eighteenth-
century tanneries from Europe: for example, the model of the tannery pro-
posed by the Encyclopédie of D’Alembert and Diderot is equipped with twelve
pits.39 Most of the twenty-four tanneries recorded at Aniane (Languedoc) in
1859 were about the same size as the one at Pompeii.40 For example, the
tannery owned by Antoine Jouillé was equipped with eight cylindrical tanning
pits and twelve rectangular tanks. Does the similarity of size imply that

36 37
Maiuri (1942: 217–20). Moeller (1976).
38
Jongman (1991); Flohr (2013b).
39 40
Diderot and Le Rond d’Alembert (1765: xiv. 889–92). David (2002).
 The Archaeology of Roman Urban Workshops 89

production was, at least, proto-industrial? This depends mainly on the num-

ber of tanneries operating at the same time in the town. At Aniane, in 1745,
the town included at least twenty-nine tanneries with a total of fifty-one
tanning pits. At Pompeii, the presence of several plants would be a strong
suggestion for an export industry, but we know only one plant of this type,
which is, however, close to a non-excavated part of the town, where similar
plants could be hidden.
Knowing the number and size of the tanning pits, can we quantify output?
We would like to attempt to estimate the maximum level of production, but
the modern example of Aniane shows that the production of leather could be
intense for some years, drop dramatically, and then restart. Our figures, based
on the material remains of workshops, will never be able to take into account
these unpredictable variations.
In any case, tanning was certainly profitable, and the very size of the
installation seems to respond to a booming demand for this essential product.
On the other hand, we must recognize that funerary monuments erected by
tanners are rare. But it is not surprising: this job was considered by society as
an ars sordida, which would not be mentioned by wealthy tanners who became
integrated into local aristocracy and became landowners. Epigraphy is thus a
misleading clue, giving a slightly false insight into the social facade of wealth
but not into the origin of that wealth.
Even in the well-known case of Pompeii, because of the lack of archives, it is
difficult either to support or to reject hypotheses about export industries.
Systematically studying a single town has its limits: to go further we would
need to extend excavations over the surrounding territory. Although Pompeii
has been largely excavated and gives a picture of a city in the late first century
AD, that picture is still partial and distorted. Excavations have tended to focus
on the centre of the town, and we know very little about the suburbs, which
were certainly crucial in term of economic activities. North of the Porta
Vesuvio, a metallurgical complex was excavated in 1901. The size of the
workshop is much larger than the metallurgical workshops inside the city
walls.41 More recently, in June 2011, the Soprintendenza partially excavated a
suburb located between Pompeii and Torre Annunziata, where there was at
least one big kiln perhaps used to fire bricks and tiles.42 There are indications
that larger plants of any type could lie outside Pompeii and indeed outside
every other city, like Bordeaux, where on the site Jean Fleuret, a tannery has
recently been excavated.43

41
The excavations in Fondo Barbatelli are mentioned in the Notizie degli Scavi (1897: 534;
1900: 501; 1901: 599; 1902: 399).
42
Unpublished excavations mentioned by the Soprintendenza.
43
Unpublished excavations directed by Christophe Sireix; the leather, fully preserved under
the water table, is currently being studied by Martine Leguilloux, who identified several inscrip-
tions and recognized that the tanning process mixed vegetal tanning and alum treatment.
90 Jean-Pierre Brun

CONCLUSIO N

Were the ancient towns only consumer cities? Was there a division of labour, a
standardization, a rationalization of the work processes and to what extent?
We have been looking at these questions for more than a century without
reaching any clear and uncontested answers, essentially because of the scarcity
of written sources and their nature. The surviving texts are almost exclusively
interpretations that allow us to see how the ruling class perceived their
economic and social world, but they do not reveal the realities of the work
and their evolution, nor do they provide quantified data. Using these sources,
we are unable to break out of this vicious circle.
Our programme proposed to leave theoretical considerations to one side for
now and look instead at the archaeological documentation, asking precise
questions such as: what are the technical and environmental conditions in
which various artisanal processes were developed? When did the fundamental
mechanical innovations appear, such as the screw, camshaft, and connecting
rod? How, when, why, and to what extent did these innovations spread around
the Mediterranean area and across the Empire?
Archaeological clues are often ambiguous, but they constitute the only form
of documentation that is in fact increasing and that can be quantified and
serialized. Thus, scholars should focus on gathering new data from archaeo-
logical excavations until we reach a statistically representative picture of the
ancient situation, before elaborating new models in order to explain these
observations: theorization may be useful for opening new research perspec-
tives, and often it pushes us to ask the right questions of the field research, but
ultimately it never answers them.

REFERENCES

Alfaro, C., Brun, J.-P., Borgard, Ph., and Pierobon, R. (eds) (2011). Purpurae vestes III:
Textiles y tintes en la ciutad Antigua (Archéologie de l’artisanat antique, 4). Valencia.
Amarger, M.-P., and Brun, J.-P. (2007). ‘La forge de l’insula I 6,1 de Pompéi’, Quaderni
di Studi Pompeiani 1: 147–68.
Amouretti, M.-C. (1991). ‘L’attelage dans l’Antiquité: Le Prestige d’une erreur scien-
tifique’, Annales: Économies, Sociétés, Civilisations 46/1: 219–32.
Amouretti, M.-C., Comet, G., Pomponi, F., Durbiano, C., et al. (1977). Campagnes
méditerranéennes, permanences et mutations. Aix–Marseilles.
Béal, J.-C., and Goyon, J.-C. (eds) (2002). Les artisans dans la ville antique
(Archéologie et Histoire de l’Antiquité, 6). Lyons and Paris.
Borgard, Ph. (2002). ‘À propos des teintureries de Pompéi: L’Exemple de l’officina
infectoria V, 1, 4’, in J.-C. Béal and J.-C. Goyon (eds), Les Artisans dans la ville
antique. Lyons, 55–67.
 The Archaeology of Roman Urban Workshops 91
Borgard, Ph., and Puybaret, M.-P. (2004). ‘Le Travail de la laine au début de l’Empire:
L’Apport du modèle pompéien. Quels artisans? Quels équipements? Quelles tech-
niques?’ in C. Alfaro, J. P. Wild, and B. Costa (eds), Purpureae vestes: Textiles y tintes
del Mediterráneo en epoca romana. International Symposium on Textiles and Dyes
in the Mediterranean Roman World, Eivissa 8–10. novembre 2002. Valencia, 47–59.
Borgard, Ph., Brun, J.-P., Leguilloux, M., Monteix, N., Cullin-Mingaud, M., and
Libre, M. T. (2005). ‘Recherches sur les productions artisanales à Pompéi et à Hercu-
lanum’, in P. Guzzo and M.-P. Guidobaldi (eds), Nuove ricerche archeologiche a Pompei
ed Ercolano (Studi della Soprintendenza Archeologica di Pompei, 10). Naples, 295–317.
Botte, E. (2009). Salaisons et sauces de poissons en Italie du Sud et en Sicile durant
l’Antiquité. Naples.
Breccia Fratadocchi, M. M., Ricci, S., and Sarlo, B. M. (1976). ‘Considerazioni su un
nuovo ambiente sottostante la basilica di Santa Cecilia in Trastevere’, BdA 61/3–4:
217–28.
Brun, J.-P. (ed.) (2009). L’Artisanat antique en Italie méridionale et en Gaule, Mélanges
offerts à Maria-Francesca Buonaiuto (Collection du Centre Jean Bérard, 32; Arché-
ologie de l’artisanat antique, 2). Naples.
Brun, J.-P., and Monteix, N. (2009). ‘Les parfumeries en Campanie antique’, in Brun
(2009), 115–33.
Brun, J.-P., Bustamente, M., Chapelin, G., and Ribera, A. (2012). ‘Parfumeries de la via
degli Augustali’, Chronique des activités archéologiques de l’École française de Rome
<http://cefr.revues.org/540>. (accessed 21 July 2015).
Brun, J.-P., and Leguilloux, M. (2014). Les installations artisanales romaines de Saepi-
num. Tannerie et moulin hydraulique. Naples.
Chardron-Picault, P. (ed.) (2010). Aspects de l’artisanat en milieu urbain: Gaule
et Occident romain, Actes du colloque international d’Autun, 20–22 sept. 2007
(Suppléments de la RAE, 28). Dijon.
Chardron-Picault, P., and Pernot, M. (1999). Un quartier antique d’artisanat
métallurgique à Autun, Saône-et-Loire—Le Site du Lycée militaire (Documents
d’Archéologie française, 76). Paris.
Coarelli, F. (1974). Roma (Guide archeologiche Laterza). Rome and Bari.
Cullin-Mingaud, M. (2010). La Vannerie dans l’Antiquité romaine: Les Ateliers de
vanniers et les vanneries de Pompéi, Herculanum et Oplontis. Naples.
Daremberg, Ch., Saglio, E., and Pottier, S. (1877–91). Dictionnaire des antiquités
grecques et romaines. Paris.
David, P. (2002). Le Travail du cuir à Aniane et en Languedoc (Cahiers d’arts et
traditions rurales, 15). Montpellier.
Devore, G., and Ellis, S. J. R. (2008). ‘The Third Season of Excavations at VIII.7.1–15 and
the Porta Stabia at Pompeii: Preliminary Report’, Fasti Online Documents & Research
<www.fastionline.org/docs/FOLDER-it-2008-112.pdf> (accessed 7 July 2015).
Diderot, D., and Le Rond d’Alembert, J. (1765). Encyclopédie ou dictionnaire raisonné
des sciences, des arts et des métiers, tome neuvième. Neuchâtel.
Duval, P.-M. (1962). ‘L’Apport technique des Romains’, in M. Dumas (ed.), Histoire
générale des techniques. Paris, 218–54.
Febvre, L. (1935). ‘Réflexions sur l’histoire des techniques’, Annales d’histoire écono-
mique et sociale 7: 531–5.
92 Jean-Pierre Brun
Ferdière, A. (2005). ‘L’Artisanat en Gaule romaine’, in L. Rivet (ed.), Société française
d’étude de la céramique antique en Gaule: Actes du congrès de Blois, 5–8 mai 2005.
Marseilles, 7–14.
Février, P.-A., and Leveau, Ph. (eds) (1982). Villes et campagnes dans l'Empire romain.
Actes du colloque d’Aix, mai 1980. Aix-en-Provence.
Février, P. A., Fixot, M., Goudineau, Chr., and Kruta, V. (1980). ‘La Ville antique’,
in G. Duby (ed). Histoire de la France urbaine i. Paris.
Flohr, M. (2013a). ‘The Textile Economy of Pompeii’, JRA 26/1: 53–78.
Flohr, M. (2013b). The World of the fullo: Work, Economy and Society in Roman Italy
(Oxford Studies on the Roman Economy). Oxford.
Fontaine, S., Satre, S., and Tekki, A. (eds) (2011). La Ville au quotidien: Regards croisés
sur l’habitat et l’artisanat, Afrique du Nord, Gaule et Italie. Aix-en-Provence.
Francotte, H. (1900–1). L’Industrie dans la Grèce ancienne. Bruxelles.
Gilles, B. (1978). Histoire des techniques. Paris.
Grenier, A. (1937). ‘La Gaule romaine’, in T. Frank (ed.), An Economic Survey of
Ancient Rome. Baltimore, 379–644.
Gsell, St (1913–28). Histoire ancienne de l’Afrique du Nord. Paris.
Guiraud, P. (1900). La main-d’œuvre industrielle dans l'ancienne Grèce. Paris.
Jashemski, W. F. (1979). The Gardens of Pompeii, Herculaneum and Villas Destroyed
by Vesuvius. New Rochelle.
Jongman, W. (1991). The Economy and Society of Pompeii. Amsterdam.
Jullian, C. (1920–6). Histoire de la Gaule. Paris.
Lefebvre des Noëttes, R. (1924). La Force animale à travers les âges. 2nd edn. Paris.
Lefebvre des Noëttes, R. (1931). L’Attelage et le cheval de selle à travers les âges.
Contribution à l’histoire de l’esclavage. Paris.
Lefebvre des Noëttes, R. (1935). De la marine antique à la marine moderne: La
Révolution du gouvernail—Contribution à l’étude de l'esclavage. Paris.
Leguilloux, M. (2004). ‘Identification des tanneries romaines par le mobilier arché-
ologique et l’archéozoologie’, in E. De Sena and H. Dessales (eds), Metodi e approci
archeologici: L’industria e il commercio nell’Italia antica. Colloque de l’American
Academy in Rome et de l’École Française de Rome. Avril 2002 (BAR International
Series, 1262). Oxford, 38–48.
Leveau, Ph. (1983). ‘La ville antique et l’organisation de l’espace rural, villa, ville,
village’, Annales ESC (July–August), 920–42.
Leveau, Ph. (1984). Caesarea de Maurétanie, une ville romaine et ses campagnes. Rome.
Maiuri, A. (1942). L’ultima fase edilizia di Pompei. Rome.
Mangin, M. (1981). Un quartier de commerçants et d’artisans d’Alésia: Contribution à
l’histoire de l’artisanat urbain en Gaule. Paris.
Mangin, M. (1985). ‘Artisanat et commerce dans les agglomérations secondaires du
Centre-Est des Gaules sous l’Empire’, in Ph. Leveau (ed.), L’Origine des richesses
dépensées dans la ville antique. Marseilles, 113–31.
Mattingly, D. J. (1990). ‘Painting, Presses, and Perfume Production at Pompei’, Oxford
Journal of Archaeology 9/1: 71–90.
Moeller, W. O. (1976). The Wool Trade of Ancient Pompeii (Studies of the Dutch
Archaeological and Historical Society, 3). Leiden.
 The Archaeology of Roman Urban Workshops 93
Monteix, N. (2011). Les lieux de métier: Boutiques et ateliers d’Herculanum (Collection
du Centre Jean Bérard, 34). Rome.
Monteix, N. (2011). ‘De l’artisanat aux métiers: Quelque réflexions sur le savoir-faire du
monde romain à partir de l’exemple pompéien’, in Monteix and Tran (2011), 7–26.
Monteix, N. (2013). ‘The Apple of Discord: Fleece-Washing in Pompeii’s Textile
Economy. A Response to Miko Flohr’, Journal of Roman Archaeology 26/1: 79–88.
Monteix, N., and Tran, N. (eds) (2011). Les Savoirs professionnels des gens de métier:
Études sur le monde du travail dans les sociétés urbaines de l’empire romain
(Archéologie de l’artisanat antique, 5). Naples.
Morel, J.-P. (1976). ‘Aspects de l’artisanat dans la Grande-Grèce romaine’, in La
magna Grecia nell’età romana: Atti del XV convegno di studi sulla Magna Grecia,
Taranto, 5–10 Ottobre 1975. Naples, 263–323.
Morel, J.-P. (1985). ‘La manufacture, moyen d’enrichissement dans l’Italie romaine?’,
in Ph. Leveau (ed.), L’Origine des richesses depensees dans la ville antique: Actes du
colloque 1984. Aix-en-Provence, 87–111.
Morel, J.-P. (1992). ‘L’artisan’, in A. Giardina (ed.), L’Homme romain. Paris, 267–302.
Morel, J.-P. (1996). ‘Élites municipales et manufacture en Italie’, in M. Cebeillac-
Gervasoni (ed.), Les Élites municipales de I'Italie péninsulaire des Gracques a
Néron: Actes de la table ronde de Clermont-Ferrand, 28–30 novembre 1991. Naples
and Rome, 181–98.
Morel, J.-P. (2009). ‘Entre agriculture et artisanat: Regards croisés sur l’économie de
l’Italie tardo-républicaine’, in J. Carlsen and E. Lo Cascio (eds), Agricoltura e scambi
nell’Italia tardo-repubblicana. Bari, 63–90.
Morel, J.-P. (2011). ‘Habitat et artisanat: Quelques impressions finales’, in Fontaine,
Satre, and Tekki (2011), 233–41.
Parmigiani, N., and Pronti, A. (2004). Santa Cecilia in Tevere: Nuovi scavi e ricerche
(Monumenti di antichità cristiana XVI). Vatican City.
Pasqualini, M., Excoffon, P., Michel, J.-M., and Botte, E. (2005–6). ‘Fréjus, Forum Iulii.
Fouilles de l’Espace Mangin’, Revue archéologique de Narbonnaise, 38–9: 283–341.
Polfer, M. (ed.) (1999). Artisanat et productions artisanales en milieu rural dans les
provinces du nord-ouest de l’empire romain: Actes du colloque, Erpeldange, Luxem-
bourg les 4 et 5 mars 1999 (Monographies Instrumentum, 9). Montagnac.
Polfer, M. (ed.) (2001). L’artisanat romain: Évolutions, continuités et ruptures, Italie et
provinces occidentales: Actes du 2e colloque d’Erpeldange, 26–28 octobre 2001
(Monographies Instrumentum, 20). Montagnac.
Polfer, M. (ed.) (2005). Artisanat et économie romaine, Italie et provinces occidentales:
Actes du 3e colloque d’Erpeldange, 14–16 octobre 2004 (Monographies Instrumen-
tum, 32). Montagnac.
Rapsaet, G. (2002). Attelages et techniques de transport dans le monde gréco-romain.
Brussels.
Rémondon, R. (1959). ‘Le monde du travail’, in L.-H. Parias (ed.), Histoire générale du
travail. Paris, 257–369.
Rivet, L. (1992). ‘Un quartier artisanal d’époque romaine à Aix-en-Provence, Bilan de
la fouille de sauvetage du “parking Signoret” en 1991’, Revue archéologique de
Narbonnaise 25: 325–96.
94 Jean-Pierre Brun
Santoro, S. (2004). ‘Artigianato e produzione nella Cisalpina romana: Proposte di
metodo e prime applicazioni’, in S. Santoro (ed.), Artigianato e produzione nella
Cisalpina, 1. Proposte di metodo e prime applicazioni. 2nd edn (Flos Italiae. Doc-
umenti di archeologia della Cisalpina Romana, 3). Florence, 19–69.
Sodini, J.-P. (1979). ‘L’artisanat urbain à l’époque paléochrétienne’, Ktèma 4: 71–119.
Tran, N. (2006). Les membres des associations romaines: Le Rang social des collegiati en
Italie et en Gaules sous le Haut-Empire (Collection de l’École française de Rome,
367). Rome.
Various authors (eds) (1963–4). Le dictionnaire archéologique des techniques. 2 vols.
Paris.
Various authors (1996). Aspects de l’artisanat du textile dans le monde méditerranéen,
Égypte, Grèce, monde romain. Lyons.
 II
Strategies
 4

Mercantile Specialization and Trading

Communities: Economic Strategies in
Roman Maritime Trade
Candace Rice

Archaeological material leaves little doubt as to the significance of maritime

transport in the Roman Mediterranean; thousands of shipwrecks, tens of
thousands of ceramic sherds, glass products, and imported building products
from bricks to coloured marbles clearly illustrate that transport over longer
distances was a constant feature of the Roman world. Yet despite the archaeo-
logical visibility of many of the transported commodities (or their containers),
identifying the mechanisms through which they moved is difficult. What do the
main modes of long-distance commerce suggest about market conditions and
the organization of trade? What types of people were involved and what types
of decisions were made? There is little direct archaeological or textual evidence
for the ways in which Roman producers, merchants, shippers, and buyers
coordinated their transactions. We may forget that every producer needed a
buyer and a shipper, or that every shipowner needed a cargo, and that every
consumer needed all of these. In the face of prolific material evidence, it is often
too easy to exclude the human element.
This chapter seeks to explore the mechanisms and transactions through
which maritime transport was conducted and the social networks and physical
infrastructure in port cities that made it possible. In doing so, the argument
moves through the processes of maritime trade to establish the pre-eminent
characteristics of Roman maritime trade and highlight the major mechanisms
involved, bringing the role of the individuals involved to the fore. This chapter
briefly introduces the merchants who played a major role in maritime trade,
not just in terms of buying and selling, but also in the decision-making
process, and then examines the trading process from product acquisition, to
shipment, to purchase, finally tying these elements together in a discussion of
the phenomenon of ‘trading communities’.
98 Candace Rice

T H E ME R C H A N T S

The primary Latin terms for those involved in maritime trade are negotiator,
mercator, navicularius, and occasionally nauta. In Greek, the terms are ἔμπορος,
ναύκληρος, and ναῦτη& (hereafter transliterated).1 The distinction between the
terms negotiator and mercator is not entirely apparent; interpretations are
complex, as one could be both a negotiator and a mercator, and furthermore
the definition of a negotiator changes between the republic and the empire.2 In
general, it would seem that the term negotiator signifies trade on a larger scale,
while mercator is often used to describe someone working on a fairly small scale.3
One indication of this is that mercator frequently appears in inscriptions without
any indication of further specialization, while the application of the term nego-
tiator is frequently further clarified by some sort of specialization.
A detailed look at the Latin inscriptions related to negotiatores is revealing.
A total of 311 inscriptions from the Epigraphik-Datenbank Clauss - Slaby (EDCS)

300

250

200
Number of attestations

150

100

50

0
Specialized Non-specialized

Fig. 4.1. Specialized versus non-specialized negotiatores (N = 311). Image: Candace Rice.

1
Negotiatores, mercatores, and emporoi were, of course, not limited to the maritime trading
sphere.
2
Andreau (2000).
3
Numerous studies have examined the role of negotiatores in the Roman world. See most
recently Tran (2014); as well as Feuvrier-Prevotat (1981); Kneissl (1983); Andreau (2000);
Verboven (2007a); and Arnaud (2011: 73–5); and Verboven (2007a) analyses the inscriptions
related to negotiatores, sorting them according to geographical location.
 Mercantile Specialization and Trading Communities 99
Trade
30
Location
Trade and location
12 qui negotiantur
Local venue
Unknown

39 124

28

Fig. 4.2. Subdivision of negotiator-inscriptions indicating some form of specialization

(N = 242). Image: Candace Rice.

refer to negotiatores; these have been grouped according to specialization.4 As

can be seen in Fig. 4.1, 242 of the inscriptions (78 per cent) provide some degree
of amplification of the title negotiator. The most common indicator of special-
ization is the trade in which the negotiator was engaged (Fig. 4.2). There are 124
such inscriptions—which is 40 per cent of all inscriptions referring to negotia-
tores. Negotiatores vinarii, or wine traders, are the most common with twenty-
nine attestations. Geographically, most of these inscriptions come from Rome
(eleven), and eleven more are from Italy (three at Ostia), but there are also several
in Gaul (three in Narbonensis and five in Lugdunensis), as well as Germany and
Dalmatia. Numerous other trades are also recorded, including negotiatores artis
cretariae (pottery dealers), negotiatores frumentarii (grain merchants), negotia-
tores ferrarii (iron merchants), negotiatores eborarii (ivory merchants), and
several cloth and clothes merchants (negotiatores lanarii, sagarii, vestiarii).
Inscriptions also record those negotiatores who specialized in trade with a
certain region rather than a certain commodity—for example, the negotiatores
Cisalpini et Transalpini. Such inscriptions account for 9 per cent of all negoti-
ator inscriptions. A further group of inscriptions (12 per cent) that deal with
location are the very specific qui negotiantur inscriptions, discussed in more
detail below. An additional 3 per cent of inscriptions record negotiatores who
specialized in a particular product from a particular region. At Rome, for
example, an inscription records a certain Tiberius Claudius Docimus who
traded in wine and salsamenta from Mauretania.5 Twelve of the inscriptions
(4 per cent) refer to negotiatores who sold products in a specific part of their
own city—for example, the negotiatores ex area Saturni from Ostia.6

4 5 6
EDCS <db.edcs.eu> (accessed 3 December 2013). CIL 6.9676. CIL 14.397.
100 Candace Rice

The terminology was not consistent. Olive-oil dealers, for example, only
rarely referred to themselves as negotiatores olearii, but more commonly
identify themselves as simply olearii; there are twenty-nine Latin inscriptions
mentioning olearii, only four of which refer to negotiatores olearii. Rome has
the highest concentration (nine), but, as expected, there are several from Spain
(five from Baetica and two from Hispania Citerior), and four of those from
Rome are traders in Baetican oil. There are also three inscriptions from North
Africa and others from Gaul and Dalmatia.

THE TRADING PROCESS

The specialization of merchants is an important indicator of the developed

nature of Roman maritime trade: specialization comes with higher risk and is
typically unsustainable in underdeveloped economies,7 yet specialized traders
were not uncommon in the Roman period.8 The fact that merchants also
chose to highlight their specialization indicates that they were successful in
their chosen ventures. The rest of this chapter briefly focuses on the practi-
calities of the trading process in order to discuss the ways in which merchants
managed their business affairs so as to secure profits.
The initial stage of product movement typically involved short transport,
overland, by river or by sea, or by a combination thereof, in order to move
products from their production site to a harbour for export. The evidence
suggests that the products of multiple estates and productive centres were
commonly combined for maritime shipment. It was often the case that from
the time the products were loaded onto a ship, and potentially from the time
they were transported from their production site, the goods were no longer
linked directly to their producers. While there is ample evidence to cite in
support of this mode of organization, this chapter is intended as a brief sketch
of the evidence and I therefore focus on three examples.
The first example comes from Baetulo, where, in 2004, excavations revealed
two large dumps of Roman amphorae, as well as three dolia.9 The deposits

7
Cf. Doerflinger (1983).
8
Erdkamp (2005: 177–8) asserts that specialized grain (p. 108) and wine merchants
(pp. 128–9) must be related to the vast demand of the city of Rome, and that only ‘populous
and wealthy cities’ had specialized merchants and large-scale trade. Rome was surely the largest
consumer base of the Roman world, and significant maritime infrastructure was developed to
facilitate its supply, both in Italy and further abroad. The integration of markets and subsequent
specialization of production and trade are, however, visible in much greater quantities across the
Roman world than Erdkamp allows. If we consider only the epigraphic evidence related to
negotiatores, Verboven (2007b) has clearly shown that areas such as Gaul, Germany, and Britain
contained a significant proportion (nearly 32%) of the epigraphic references to negotiatores.
9
Comas and Padrós (2008).
 Mercantile Specialization and Trading Communities 101

were located near previously identified docking facilities and have been inter-
preted as the results of harbour loading activities.10 Both deposits span the late
first century BC through the first century AD and contain primarily Tarraco-
nensian amphorae. The first deposit (over 3,000 sherds) contained 72 per cent
Tarraconensian amphorae (54 per cent Pascual 1 and 18 per cent Dressel
2–4).11 Eight per cent of the amphorae were Gauloise 4, which may or may not
have been imported; they could be local imitations. There were also Dressel
7–11 amphorae from both Baetica and Tarraconensis and Dressel 20s from
Baetica. The second deposit (over 10,000 sherds) contained 74 per cent
Tarraconensian Dressel 2–4 and only 15 per cent Pascual 1 amphorae. The
remaining 11 per cent were primarily Baetican (Dressel 20 and Dressel 7–11).
What is particularly interesting about this second deposit is that six amphorae
stamps and six graffiti were recovered. While not all were readable, three of the
amphorae stamps indicated the production sites of the amphorae. Two stamps
(AC and M) are associated with the production site of Can Tintorer and one
(NI or IN) with the site of Can Pedrerol, both of which were located in the
Llobregat river valley. These deposits are significant, as they provide terrestrial
evidence that the products of different sites were assembled near the harbour
before being loaded onto ships.
The second example is the Madrague de Giens wreck. This well-known
Republican wreck (70–50 BC) carried approximately 350–450 tons of cargo,
including some 6,000–7,000 amphorae, the majority of which are Dressel 1B
wine amphorae.12 Recent reappraisal of the stowage arrangements of the cargo
of the Madrague de Giens wreck clearly shows that not only did the wine
amphorae come from multiple production sites (identified through stamps on
the amphorae) but that no effort was made to keep the products separated by
their origin, thus implying that their production location was no longer
important. Indeed, the amphorae were not even arranged by the merchants’
stamps, which were seemingly applied by merchants after they had purchased
the wine from the producers.13 This strongly suggests that, by the time the
cargo was loaded onto the ship itself, it was under the ownership of a single
merchant and intended for a single destination.14
Evidence such as the Baetulo deposits and the cargo of the Madrague de
Giens wreck suggests that, while the number of producers represented by a

10 11
Comas and Padrós (2008: 83). Comas and Padrós (2008).
12 13
For the most complete publication, see Tchernia et al. (1978). Hesnard (2012).
14
Similar cargoes composed of collections of amphorae-borne products from a variety of
local producers are found on the Cap del Vol, La Giraglia, Grand Ribaud D (wine cargoes), and
the Plemmirio B and Chiessi wrecks (olive oil and salted-fish products). It is important to note
that the composition of the cargoes from these wrecks suggests that these ships were primarily
carrying cargoes on the first stage of their maritime export. This type of interpretation is
obviously not possible for the more heterogenous wrecks, where cargoes have potentially been
bought and sold multiple times.
102 Candace Rice

single shipload may be substantial, the control of the product, and often the
ownership of the product, had been drastically consolidated by the time the
product was placed on the ship.15 Indeed in some cases, the naukleros of
the ship and the owner of the goods were one and the same, which leads us to
the third example.16
The final example is that of the second-century AD P. Bingen 77, which
provides strong evidence that cargoes were owned by a limited number of
merchants and naukleroi.17 This papyrus records eleven ships entering the
port of what is probably Alexandria. All of the ships originated at a single
place, and most have either one or two types of cargo. The names of the
naukleroi are preserved for all eleven ships (though in one case only the latter
part of the name is preserved), and the names of the owner of the cargo are
preserved for nine of the ships. No cargo has more than two owners, and six of
the ships have only one owner attested for the cargo. For example, the final
ship recorded on the register originated from Anemurium and was carrying
2,500 jars of wine owned by Ninos, son of Tounis, also the naukleros.
This evidence suggests a separation between producer and maritime
exporter, a phenomenon that is also supported by several textual sources.18
Merchant middlemen often provided the link between the producers and the
maritime traders. A system in which maritime traders bought their wares from
middlemen would perhaps have created a simplified network of communica-
tion and exchange where the locally or regionally based merchant coordinated
transactions through fostered local contacts, while the merchant specializing
in long-distance trade established and maintained a different set of contacts.
Our understanding of the maritime phases of transport depends largely on
shipwrecks, which preserve cargoes as they were en route. Negotiatores dealt
broadly with three main cargo types: export cargoes, re-export cargoes assem-
bled at emporia, and return cargoes. It is, however, important to realize that
these are not mutually exclusive categories. Export cargoes are often regarded
as homogeneous shiploads that were exported from an assembly port in their
production region to an emporion or a centre of mass consumption.19 Cargoes
originating from a localized production zone, such as the previously discussed
Madrague de Giens, are often indicative of products in their initial stage of
maritime export. The term homogeneous is, however, misleading, particularly
as it obscures the relationships between a multiplicity of producers and
merchants; while it is often the case that cargoes are homogeneous in terms

15
For discussion of the potential processes involved in the movement of goods from the farm
to the market as interpreted from the literary sources, see Erdkamp (2005: 118–30).
16
A striking exception appears to be the Port-Vendres II wreck, which contained the
products of at least nine merchants (Colls et al. 1977).
17
Heilporn (2000).
18
e.g. Cato, On Agriculture 144–8, Pliny, Letters 8.2.1 and the Digest 18.6.1, 18.6.6, 18.6.15.
19
e.g. Boetto (2012: 156).
 Mercantile Specialization and Trading Communities 103

of the type of product (for example, wine or salted-fish products), they are
rarely homogenous in terms of the estate or farm where the product was
produced (stone cargoes are an important exception).20 We can imagine that
the negotiatores who specialized in the trade of a particular product (for
example, negotiatores vinarii) were particularly associated with export cargoes.
The second main cargo type consisted of shiploads of products that had
been assembled at emporia and were being re-exported. Emporia acted as both
regional and inter-regional collection points of products, and cargoes assem-
bled in these places frequently included a mix of products from various places
of origin. Cargoes loaded at emporia might contain products on their original
stage of maritime export, products that had previously arrived at an emporium
and were being re-exported, or a combination of the two.
A third regular cargo type is the return cargo.21 Return cargoes are closely
related to trade between emporia: after a merchant has moved a cargo from
emporium A to emporium B, he ideally returns to emporium A with a new
cargo. Economically speaking, return cargoes ensured that both ends of a
journey were potentially profitable: given the choice between sailing with only
ballast or sailing with a cargo, it often made financial sense to take on a cargo.
Return cargoes are thought to have been responsible for the widespread
distribution of many low-value products across the Roman world: Italian
bricks were used in the construction of bath complexes and cisterns in
numerous North African port sites, and Cilician roof tiles were used in Syria
and Phoenicia. 22 The argument behind this idea is that the profits from high-
value products in effect subsidize the shipment of low-value products, and,
while this is certainly true, we should not automatically equate return cargoes
with low-value or afterthought cargoes. After all, it is often the return cargoes
that are responsible for the import products that are locally needed at the
home port and its environs; this could be bricks for a specific building project
or a selection of foreign wines. As a result, it is extremely difficult, if not
impossible, to identify return cargoes among shipwrecks, despite the ample
evidence of return cargoes visible through terrestrial archaeology.

20
When amphorae were produced on location with the oil, wine, etc., as was often the case,
the association between the actual product and its container is clear. Amphorae were, however,
also produced at kiln sites that are unattached to a particular farm or estate. These manufactur-
ing sites served as bottling centres for producers or merchants who did not have kilns themselves.
In these cases, apparent homogeneity can be deceptive. Given that these kiln sites were inde-
pendent from production, however, it is probable that they were used to bottle the products of
many agricultural production units. The San Anastasia amphorae kiln, which produced two-
thirds of the Dressel 1B amphorae aboard the Madrague de Giens wreck, was just such a kiln.
Many of the amphorae from the San Anastasia kiln, however, had stoppers with merchants’
marks, indicating that they were meant to be distinguishable from each other during the period
between the bottling and the point at which they were loaded onto the ship.
21
On return cargoes, see Tchernia (2011: 124, 344–5).
22
For the use of Italian bricks in North African port sites, see Wilson (2001); Wilson, Schörle,
and Rice (2012). On Cilician roof tiles in Syria and Phoenicia, see Mills (2005).
104 Candace Rice

Significantly, the trading process as described above presupposes a stable

system in which there was relatively reliable information available about
the market situation in remote ports; at least, the process was based on the
expectation that one can sell particular cargoes in places with which one
maintains regular trading connections. By establishing themselves as mer-
chants of known routes, cities, or products, these negotiatores and merchants
improved the efficiency of the trading process. Actual selling strategies may
have varied: a merchant who specialized in long-distance trade might regularly
sell his products to a local merchant in a destination city—the counterpart to
the merchant middleman on the production end. As with the merchant
middleman on the production end, this would allow a simplification of the
communication network. However, merchants could also sell their products
on the local market themselves, setting up shop in the forum, macellum, or
agora, or making use of more transient fairs and markets.23

COORDINATIN G TRADE

Roman maritime trade thus operated within a complex and sophisticated

system where several groups of actors played several specific roles in different
places; the modalities of trade suggest reliance on some degree of information
about distant markets, rather than casual opportunistic tramping from one
port to another.24 This did, however, complicate the coordination of various
stages of the trading process. Pre-industrial maritime trade was limited by a
number of factors, the two most important being a dependence on wind for
sailing and the difficulties of long-distance communication. As early modern
parallels suggest, these factors were not wholly prohibitive; the Venetians, the
Portuguese, and the Dutch all had successful maritime trading empires; while
none of these empires overcame the dependence on the sail, they all found
ways around the problem of communication, whose resolution did not really
begin until the invention of deep-sea cables in the 1860s.25
The primary way in which the Romans dealt with communication issues in
the context of trade was the establishment of diaspora communities in major
trading centres across the Roman world.26 The term ‘trading diasporas’ was
coined in 1971 by anthropologist Abner Cohen to describe ‘dispersed, but
highly interrelated communities’. In his view, a trading diaspora is established
to ‘co-ordinate the co-operation of its members in the common cause and
establish channels of communication and mutual support with members from

23
On markets and fairs, see De Ligt (1993).
24 25
Arnaud (2011); Wilson (2011), contra Bang (2008: 141). Stopford (2009: 27).
26
For an overview of Roman trading communities, see now Terpstra (2013).
 Mercantile Specialization and Trading Communities 105

communities of the same ethnic group in neighbouring localities who are

engaged in the trading network’.27 Historical diaspora communities include
the Venetians in Constantinople and the Dutch in Venice.28 Excellent evi-
dence for the workings of medieval merchant communities comes from the
documents of the Cairo Geniza.29
One category of Latin inscriptions mentions specific ethnic groups of
people who trade in specific locations. The construction used in these inscrip-
tions is typically very formulaic: ‘ethnic group-qui location-negotiantur’. In
contrast with other inscriptions referring to traders, these inscriptions are also
largely concentrated in the eastern Mediterranean and are occasionally bilin-
gual. Thirty-nine inscriptions belonging to this category are known
(Fig. 4.3),30 twenty-seven of which (73 per cent) are from the eastern Medi-
terranean, primarily from the provinces of Asia and Achaia. Some of the
earliest inscriptions come from Delos, where four inscriptions from the first
century BC have been found that refer to the ‘Italians and Greeks who trade at
Delos’.31

luliobriga

Bracara Augusta

Prymnessus
Mytilene Apamea Mallos
Sardes
Ephesus
Hispalis Panormus Kos
Aegio Delos
Argos Salamis
Rhodos
Agrigentum
Paphos
Thinissut
Gortyn

Cyrene
Alexandria

Fig. 4.3. Places where ‘qui negotiantur’ inscriptions have been found.
Map: Miko Flohr.

27
Cohen (1971: 267).
28 29
Constantinople: Dursteler (2008); Venice: Van Gelder (2007). Goitein (1967).
30
EDCS <db.edcs.eu> (accessed 3 December 2013): CIL 1.746, 830, 831, 836, 2955, 2958,
2970; CIL 2.242, 1168, 1169, 2423; CIL 3.532, 5230, 6051, 7043, 7237, 7240, 12101, 12266, 13542,
13617, 14195.39; CIL 6.40903; CIL 10.1797; CIL 11.7288; CIL 13.11825; AE 1912, 51; AE 1924,
69, 72; AE 1968, 480; AE 1974, 671; AE 1975, 803; AE 1990, 938; AE 1996, 1453; InscCret-4,
Gortyn 290, 291; IK 12. 409.
31
e.g. CIL 3.7240.
106 Candace Rice

Many of the inscriptions mention cives Romani, who traded with various
locations across the Mediterranean.32 At Ephesos, numerous inscriptions
show the presence of associations of Italian traders. Around 60 BC, the Italicei
quei Ephesi negotiantur dedicated a statue of a certain L. Agrius Publeianus,
and in 37 BC they dedicated an honorary inscription placed in the western hall
of the agora to the consul M. Cocceius Nerva.33 In 36 BC, the conventus civum
Romanorum quei Ephesi negotiantur dedicated another inscription to
M. Cocceius Nerva.34 In the AD 40s, the conventus civum Romanorum qui in
Asia negotiantur made two dedications to Claudius, one of which was an
equestrian statue.35
Groups of cives Romani also appear on Crete. Several inscriptions attest
groups of Roman traders at Gortyn—the cives Romani qui Gortynae negotian-
tur.36 In the first century BC, these people set up an honorific inscription to
Doia Procilla, probably a member of a duumviral family from Knossos.37 In
the second century AD, the cives Romani—who instead of being identified
as negotiatores are now described as those who settled in Gortyn (qui
consistunt)—along with two priests of the imperial cult dedicate an honorary
inscription to Septimius Severus.38
Groups of merchants (probably Italian) doing business in Alexandria, Asia,
and Syria are also identified in an inscription dedicated to L. Calpurnius
Capitolinus:
Calpurnio L(uci) f(ilio) | Capitolino || C(aio) Calpurnio L(uci) f(ilio) | [[6]] ||
mercatores qui Alexandr[iai] Asiai Syriai negotiantu[r].39
Aside from the above inscriptions, some of our best examples of trading
diasporas come from Ostia and Puteoli. So-called stationes are attested in
several harbour cities and were probably present in many others. The best
attested are those in the Piazzale delle Corporazioni at Ostia. The Piazzale
consists of sixty-one individual stationes, many of which preserve mosaics
reflecting various economic activities. Surviving inscriptions attest navicularii
and negotiatores from fourteen different cities: ten are in North Africa, two in
Gaul, and two in Sardinia (Fig. 4.4).40 Offices are also indicated for traders of

32
There are twelve overall that refer to the cives Romani. They come from Ephesos (3), Kos
(1), Mytilene (1), Prymnessus (1), Gortyn (1), Limnia (2), Bracara (1), Terracina (1), Thinissut,
Africa Proconsularis (1).
33 34 35
IvE 6, 2058. IvE 3, 658. IvE 2, 409; IvE 7.1, 3019.
36 37
I.Cret. 4, 291. I.Cret. 4, 290; Baldwin Bowsky (1999: 310).
38 39
I.Cret. 4, 278, CIL 3.4. CIL 10.1797.
40
North Africa: Misua (CIL 14.4549.10), Musluvium (CIL 14.4549.11), Hippo Diarrytus (CIL
14.4549.12), Sabratha (CIL 14.4549.14), Gummi (CIL 14.4549.17), Carthage (CIL 14.4549.18),
Sullecthum (CIL 14.4549.23), Colonia Iulia Curubis (Kurba, Tunisia) (CIL 14.4549.34), Alexan-
dria (CIL 14.4549.40), and Mauretania Caesariensis (CIL 14.4549.48). Gaul: Arelate (Statio 27),
Narbo Martius (CIL 14.4549.32). Sardinia: Turris Libisonis (Porto Torres) (CIL 14.4549.19) and
Carales (Cagliari) (CIL 14.4549.21).
 Mercantile Specialization and Trading Communities 107

Arelate
Narbo

ROME
OSTIA
Turris

Carales

Hippo Diarrhytus
Missua
Musluvium Carthago Curubis
Sullecthum
Gummi

Alexandria
Sabratha

Fig. 4.4. Cities with a statio in the Piazzale delle Corporazioni at Ostia.
Map: Miko Flohr.

flax and rope, leather traders, and wood traders. The individual offices in the
building are relatively small; they were certainly not for the storage of any
goods and were most likely offices where one could coordinate trading agree-
ments with members of various represented areas of the empire. The building
thus facilitated commercial transactions at Ostia by creating a centrally defined
location where one could readily find representatives from key emporia.
A well-known inscription from Puteoli dated to AD 174 records the request
of an association of Tyrians resident in Puteoli for assistance from Tyre in
paying the annual rent for their statio.41 Other foreign groups were also active
at Puteoli, though none is attested as explicitly as the Tyrians.42 The presence
of residents from Berytus is indicated by an inscription set up by the Berytian
worshippers of Jupiter Heliopolitanus dedicated to Trajan.43 Another inscrip-
tion mentioning Jupiter Heliopolitanus was set up by a corpus of worshippers
from Berytus or Heliopolis that apparently owned a seven-iugera necropolis
with a cistern and tabernae.44
It is likely, on the basis of several cultic inscriptions to the principal god of
the Nabataeans, Dusares, that there was also a community of Nabateans at
Puteoli.45 The evidence suggests there was a temple to this deity, possibly

41
Editions: Mommsen (1843: 57–62); CIG 3.5853; IG 14.830. Discussed by D’Arms (1974:
105); Duncan-Jones (1974: 210, 236); Sosin (1999).
42
On foreign communities in Puteoli, see particularly D’Arms (1970, 1974) and Terpstra
(2013: 51–94).
43 44
CIL 10.1634. CIL 10.1579.
45
AE 1971, 86; AE 1994, 422–3; AE 2001, 843; CIL 10.1556.
108 Candace Rice

located in the vicus Lartidianus, an area of Puteoli believed to have been

inhabited largely by foreigners, on the basis of an inscription dedicated to
Hadrian by the inquilini of the vicus Lartidianus.46 The numerous references
to this cult, and the probable location of the temple in an area associated with
foreign groups coupled with the general scarcity of the worship of Dusares in
Italy, certainly point to a resident community of Nabateans in Puteoli. The
existence of foreign cults in harbour cities can be a useful marker of groups of
foreign residents.47
The presence of long-standing diaspora communities and particularly of
resident foreign merchants links directly with a sophisticated and organized
model of trade; that is, the idea of premeditated harbour-to-harbour trade
linked with trade through emporia. It was the recurring trade among specific
parties that made the establishment of diaspora communities of merchants a
desirable strategy. The amounts of money invested in maintaining such
communities, whether for the maintenance of a statio or the installation of a
new cult, suggests regular and sustained contact between regions.

THE S OCIAL S TATUS OF BUSINESSMEN

The key role of trading communities makes it relevant briefly to consider the
social situation of the people discussed. The social status of merchants and
traders, and indeed of businessmen in general, has generated much debate
among scholars. This is unsurprising given not only the general debate over
the nature of the Roman economy, but also because it seems to have been a
somewhat controversial topic among authors of the Roman period. In the
trading scenarios outlined, it is quite possible for the elites to avoid the actual
marketing processes altogether—at least visibly. The elites often owned the
estates on which agricultural products were grown, yet, as seen in passages
from Cato and Pliny, as well as the Digest, they did not market the products
personally, but sold them to merchants.48 D’Arms and Tchernia have argued
that the separation of the producers from the traders allowed a way for the
landed elites to maintain at least a symbolic separation from trade.49

46 47
AE 1977, 200. See Wilson, Schörle, and Rice (2012).
48
e.g. Cato, On Agriculture 144–8, Pliny, Letters 8.2.1 and the Digest 18.6.1, 18.6.6, 18.6.15.
See Erdkamp (2005: 118–34) for a thorough discussion of the advance sale of agricultural
products.
49
D’Arms (1981); Tchernia (2011). There were, of course, exceptions. The senatorial Sestius
family at Cosa, well known for the production and trade of wine during the latter half of the
second century through the first century BC, is a prime example. They owned vineyards,
produced amphorae, owned ships, and seemingly expanded into the fish-salting industry. Cf.
McCann (2002). Numerous amphora sherds discovered around Cosa stamped with Sestii family
 Mercantile Specialization and Trading Communities 109

The most obvious way in which elites could maintain a discreet role in the
world of commerce was through the agency of freedmen.50 There are numerous
ways in which freedmen, or even slaves, could have been involved in commercial
partnerships with their patron, from selling the products of their patron
(as already discussed) to handling specific stages of transactions or coordinating
supply.51 This idea has been most prominently put forward by Frier and
Kehoe, who discuss the idea of freedman agency within the framework of
New Institutional Economics. They conclude as follows:
This method of managing businesses through friends or social dependants had
significant implications for the organization of the Roman economy. For one, it
tended to reinforce the strict social hierarchy that helped to preserve the eco-
nomic and social privileges of the landowning elite: there was little capacity for
developing a class of artisans let alone entrepreneurs who were fully independent
of elite patronage or control. Successful freedmen who gained wealth as artisans
or business managers were ultimately dependent on a master or patron for an
initial investment in skills and capital, and they often remained socially bound at
least to some degree to their patron.52
Frier and Kehoe do, however, argue that slaves ‘could operate with a great deal
of independence’ despite being socially bound to their patrons.53
The ideas presented by Frier and Kehoe are largely credible—there is
abundant evidence that freedmen and even slaves often acted as agents for
their elite owner or former owner’s business. Wim Broekaert, who analyses the
issue much more closely elsewhere in this volume, concludes that tracing
the relationship between patron and freedmen is extremely difficult with the
evidence at hand, and he agrees with Frier and Kehoe, largely because of the
significant advantages that a freedmen with start-up capital from a patron
would have over an independent freedmen.54 But is this really to say that
behind every freedman negotiator there was an elite patron in control?
It is important not to underestimate the independence of freedmen busi-
nessmen, particularly within certain environments.55 This is not to say that
they did not get their beginnings in trade through their patrons, or that many
of them did not continue to work with the same families following manumis-
sion. What must be kept in mind, however, are the environments in which
merchants and traders were working. Within a busy port where merchants,

stamps indicate the production of amphorae, which were exported on a significant scale, with
distribution extending primarily to Gaul and Spain, but also to Athens and Delos. Cf. Manacorda
(1978). Cicero refers to the navigia luculenta that Sestius contributed to Brutus’ forces during the
civil wars (Cic. Att. 16.4.4).
50
For the most recent work on freedmen, see Mouritsen (2011).
51
Broekaert (2012). Broekaert (Chapter 10, this volume) identifies six basic levels of agency.
52 53
Frier and Kehoe (2007: 133–4). Frier and Kehoe (2007: 134).
54 55
Broekaert, Chapter 10, this volume. See also Tran (2014).
110 Candace Rice

traders, and shippers interacted, it cannot always have been the familial
connections that dictated social and economic interactions. Instead, the evi-
dence presented above seems to suggest that merchants—freedmen or not—
had a considerable say in the ways in which products moved after they had
been purchased from estates. It is also evident from epigraphy that self-
identifying as a merchant was acceptable and, in fact, quite common in
some locales. This strongly suggests that there was a recognizable group of
entrepreneurs in the Roman world who were responsible for negotiating their
own reputation and were not relying on the reputation of their patrons.
In a controversial book published in 2012, Mayer argues for the existence of a
Roman middle class with not just shared economic standing, but also ‘a class-
specific set of values’.56 His methodology, and in particular his assertions
regarding shared cultural norms between the supposed ancient and modern
middle class, have been justifiably questioned.57 While Mayer’s arguments for
the existence of a ‘middle class’ stretch the evidence to overreaching conclu-
sions, Verboven has argued in several articles that there was in fact a Roman
‘business class’.58 His approach is more circumspect than that of Mayer and
he makes a very strong argument for the emergence of a business class of
merchants, producers, and traders in the north-west provinces. He argues that
the strong military presence in these provinces provided a large emerging market
unlike anything previously in the region and that this created an environment for
a business class to prosper.59 While the disproportionately large amount of
epigraphy related to merchants and traders in Gaul and Germany must depend
on factors other than the military—otherwise the same pattern would occur in
every militarized frontier zone—the point that regions with large markets create
extended opportunities for businessmen is valid.60
Port cities create a similar type of environment; commerce was their raison
d’être, and the people involved in this commerce formed the largest and in
some respects most important social class present within them.61 Without
imposing modern ideas of class and value on the commercial actors of the
Roman world, the fact remains that in some environments, such as port cities
and other large market zones, people involved in commerce, particularly at its
upper levels, would have had a substantial income and elevated status. Whether
or not they should be termed a ‘class’ is debatable, but, as the complexity of the
Roman economy continues to be explored, it is certainly worth pursuing
further the societal place of the commercial agents of the Roman world.

56
Mayer (2012: 6).
57
On the reception of the book, see, e.g. reviews by Flohr (2013), Noreña (2013), and
Wallace-Hadrill (2013).
58 59
Verboven (2007b, 2009, 2011a, b). See also Broekaert (2012). Verboven (2007b).
60
I would like to thank Tyler Franconi for his discussion of this phenomenon.
61
Cf. Mouritsen (2001); Heinzelmann (2010: 7).
 Mercantile Specialization and Trading Communities 111

CO NCLUSION

It has been suggested by Whittaker that the widespread distribution of wine

amphorae around the Mediterranean was due not to market trade, but to elites
transporting their own products among their familial estates.62 While I would
not deny that this practice sometimes occurred, the evidence presented here
illustrates a scale and sophistication across all aspects of the trading process
that would not have been necessary outside a market system. Facilitated by
physical, infrastructural, and social networks, Roman maritime trade involved
advanced commercial processes that operated within a market economy. The
discussion of ports, shipwrecks, merchants, traders, and producers in this
chapter has outlined the broad mechanisms of maritime trade. By thinking
of trade as a composite process in which production, transport, and marketing
processes are intrinsically linked, the level of organization and development
that were required to succeed in commerce becomes readily apparent. Each
category demonstrates a specific part of the processes of maritime trade; when
considered together, they produce a detailed picture of how maritime trade
was carried out during the Roman period. The evidence of the people involved
in maritime trade further brings to light the complexity of managing success-
ful trading ventures: trading diasporas reveal a high level of sophistication in
dealing with problems of communication inherent in pre-industrial trade; the
obvious level of specialization among merchants and traders further speaks to
the advanced structure and scale of maritime trade. It also indicates that trade
could make a profitable career.

REFERENCES

Andreau, J. (2000). ‘Negotiator’, Brill’s Neue Pauly, Brill Online.

Arnaud, P. (2011). ‘Ancient Sailing-Routes and Trade Patterns: The Impact of Human
Factors’, in Robinson and Wilson (2011), 61–80.
Baldwin Bowsky, M. W. (1999). ‘The Business of Being Roman’, in A. Chaniotis (ed.),
From Minoan Farmers to Roman Traders. Stuttgart, 305–47.
Bang, P. F. (2008). The Roman Bazaar: A Comparative Study of Trade and Markets in a
Tributary Empire. Cambridge.
Boetto, G. (2012). ‘Les Épaves comme sources pour l’étude de la navigations et des
routes commerciales: Une approche methodologique’, in S. Keay (ed.), Rome, Portus
and the Mediterranean. Rome, 153–73.
Broekaert, W. (2012). ‘Joining forces. Commercial partnerships or societates in the
early Roman Empire’, Historia 61: 221–53.

62
Whittaker (1985, 1989).
112 Candace Rice
Cohen, A. (1971). ‘Cultural Strategies in the Organization of Trading Diasporas’, in
C. Meillassoux (ed.), The Development of Indigenous Trade and Markets in West
Africa. Oxford, 266–78.
Colls, D., Etienne, R. Lequément, R., Liou, B. and Mayet, F. (1977). L’épave Port-
Vendres II et le commerce de la Bétique à l’époque de Claude (Archaeonautica, 1).
Paris.
Comas, M., and Padrós, P. (2008). ‘Deux grands dépotoirs d’amphores Léétaniennes,
Bétiques et Gauloises, hors les murs de la ville de Baetulo (Badalona)’, SFECAG,
Actes du Congrès de L’Escala-Empúries. Marseilles, 75–86.
D’Arms, J. H. (1970). Romans on the Bay of Naples: A Social and Cultural Study of the
Villas and their Owners from 150 BC to AD 400. Cambridge, MA.
D’Arms, J. H. (1974). ‘Puteoli in the Second Century of the Roman Empire: A Social
and Economic Study’, JRS 64: 104–24.
D’Arms, J. H. (1981). Commerce and Social Standing in Ancient Rome. Cambridge, MA.
De Ligt, L. (1993). Fairs and Markets in the Roman Empire: Economic and Social
Aspects of Periodic Trade in a Pre-Industrial Society (Dutch Monographs on Ancient
History and Archaeology). Amsterdam.
Doerflinger, T. M. (1983). ‘Commercial Specialization in Philadelphia’s Merchant
Community, 1750–1791’, Business History Review 57: 20–49.
Duncan-Jones, R. P. (1974). The Economy of the Roman Empire. Cambridge.
Dursteler, E. (2008). Venetians in Constantinople: Nation, Identity, and Coexistence
in the Early Modern Mediterranean (The Johns Hopkins University Studies in
Historical and Political Science). Baltimore.
Erdkamp, P. (2005). The Grain Market of the Roman Empire: A Social, Political and
Economic Study. Cambridge.
Feuvrier-Prevotat, C. (1981). ‘Negotiator et mercator dans le discours cicéronien: Essai
de définition’, Dialogues d’histoire ancienne 7: 367–405.
Flohr, M. (2013). Review of E. Mayer, ‘The Ancient Middle Classes: Roman
Life and Aesthetics in the Roman Empire’, JRS 103: 308–9.
Frier, B., and Kehoe, D. P. (2007). ‘Law and Economic Institutions’, in W. Scheidel,
I. Morris, and R. P. Saller (eds), The Cambridge Economic History of the
Greco-Roman World. Cambridge, 113–43.
Goitein, S. D. (1967). A Mediterranean Society: The Jewish Communities of the Arab
World as Portrayed in the Documents of the Cairo Geniza, i. Economic Foundations.
Berkeley (repr. 1999).
Heilporn, P. (2000). ‘Registre de navires marchands’, in H. Melaerts (ed.), Papyri in
honorem Johannis Bingen Octogenarii (P. Bingen). Leuven, 339–59.
Heinzelmann, M. (2010). ‘Supplier of Rome or Mediterranean marketplace? The
Changing Economic Role of Ostia after the Construction of Portus in the Light of
New Archaeological Evidence’, Bollettino di Archeologia Online, Volume Spe-
ciale: 5–10 <http://www.bollettinodiarcheologiaonline.beniculturali.it/documenti/
generale/2_Heinzelmann_paper.pdf> (accessed 14 July 2015).
Hesnard, A. (2012). ‘L’épave La Madrague de Giens (Var) et la plaine de Fondi
(Latium)’, Archaeonautica 17: 71–93.
Kneissl (1983). ‘Mercator-negotiator. Römischer Geschäftsleute und die Terminologie
ihrer Berufe’, Münstersche Beiträge zur Antiken Handelsgeschichte 2.1: 73–90.
 Mercantile Specialization and Trading Communities 113
McCann, A. M. (2002). The Roman Port and Fishery of Cosa: A Short Guide = Il porto
romano e la peschiera di Cosa: Guida breve. Rome.
Manacorda, D. (1978). ‘The Ager Cosanus and the Production of the Amphorae of
Sestius: New Evidence and a Reassessment’, JRS 68: 122–31.
Mayer, E. (2012). The Ancient Middle Classes : Urban Life and Aesthetics in the Roman
Empire, 100 BCE–250 CE. Cambridge, MA, and London.
Mommsen, T. (1843). De collegiis et sodaliciis Romanorum. Kiel.
Mouritsen, H. (2001). ‘Roman Freedmen and the Urban Economy: Pompeii in the
First Century AD’, in F. Senatore (ed.), Pompeii tra Sorrento e Sarno. Roma, 1–28.
Mouritsen, H. (2011). The Freedman in the Roman World. Cambridge.
Noreña, C. F. (2013). ‘Emanuel Mayer: The Ancient Middle Classes: Urban Life and
Aesthetics in the Roman Empire, 100 BCE–250 CE’, American Historical Review
118/5: 1576–7.
Robinson, D. and Wilson, A. (eds) (2011). Maritime Archaeology and Ancient Trade in
the Mediterranean (Oxford Centre for Maritime Archaeology, Monograph 6).
Oxford.
Russell, B. (2011). ‘Lapis Transmarinus: Stone Carrying Ships and the Maritime
Distribution of Stone in the Roman Empire’, in Robinson and Wilson (2011),
139–55.
Russell, B. (2013). The Economics of the Roman Stone Trade. Oxford.
Sosin, J. D. (1999). ‘Tyrian stationarii at Puteoli’, Tyche 14: 275–84.
Stopford, M. (2009). Maritime Economics. Abingdon.
Tchernia, A. (2011). Les Romains et le commerce. Naples.
Tchernia, A., Pomey, P., Hesnard, A., Girard, M., Hamon, E., Laubenheimer, F.,
Lecaille, F., Carrier-Guillomnet, A., Chene, A., Gassend, J.-M., Reveillac, G., and
Rival, M. (1978). L’Épave Romaine de la Madrague de Giens (Var) (Campagnes
1972–1975): Fouilles de l’Institut d’archéologie méditerranéenne. Paris.
Terpstra, T. (2013). Trading Communities in the Roman World: A Micro-Economic
and Institutional Perspective (Columbia Studies in the Classical Tradition, 37).
Leiden and Boston.
Tran, N. (2014). ‘Les Hommes d’affaires romains et l’expansion de l’Empire (70 av.
J.-C.–73 apr. J.-C.)’, Pallas 96: 111–26.
Van Gelder, M. (2007). ‘Trading Places: The Netherlandish Merchant Community in
Venice, 1590–1650’. Ph.D. thesis, University of Amsterdam.
Verboven, K. (2007a). ‘Ce que negotiari et ses dérivés veulent dire’, in J. Andreau,
J. France, and V. Chankowski (eds), Vocabulaire et expression de l’économie dans le
monde antique, Bordeaux, 89–118.
Verboven, K. (2007b). ‘Good for Business: The Roman Army and the Emergence of a
“Business Class” in the Northwestern Provinces of the Roman Empire (1st Century
BCE–3rd Century CE)’, in L. De Blois, E. Lo Cascio, O. Hekster, and G. de Kleijn (eds),
The Impact of the Roman Army (200 BC–AD 476): Economic, Social, Political,
Religious and Cultural Aspects: Proceedings of the Sixth Workshop of the Internation-
al Network Impact of Empire (Roman Empire, 200 BC–AD 476), Capri, March
29–April 2, 2005. Leiden and Boston, 295–314.
Verboven, K. (2009). ‘Magistrates, Patrons and Benefactors of Collegia: Status Building
and Romanisation in the Spanish, Gallic and German Provinces’, in B. Antela
114 Candace Rice
Bernárdez and T. Ñaco Del Hoyo (eds), Transforming Historical Landscapes in the
Ancient Empires: Proceedings of the First Workshop, December 16–19th 2007. Ox-
ford, 59–167.
Verboven, K. (2011a). ‘Resident Aliens and Translocal Merchant Collegia in the
Roman Empire’, in O. Hekster and T. Kaizer (eds), Frontiers in the Roman World:
Proceedings of the Ninth Workshop of the International Network Impact of Empire
(Durham, 16–19 April 2009). Leiden and Boston, 335–48.
Verboven, K. (2011b). ‘Professional Collegia: Guilds or Social Clubs?’, Ancient Society
41: 187–95.
Wallace-Hadrill, A. (2013). ‘Trying to Define and Identify the Roman “Middle Class” ’,
JRS 26: 605–9.
Whittaker, C. R. (1985). ‘Trade and the Aristocracy in the Roman Empire’, Opus 4:
49–75.
Whittaker, C. R. (1989). ‘Amphorae and Trade’, in D. Lenoir, D. Manacorda, and
C. Panella (eds), Amphores romaines et histoire économique: Dix ans de recherche.
Actes du colloque de Sienne, 22–24 mai 1986. Rome, 537–9.
Wilson, A. (2001). ‘Ti. Cl. Felix and the Date of the Second Phase of the East Baths’, in
L. M. Stirling, D. J. Mattingly, and N. Ben Lazreg (eds), Leptiminus (Lamta): The
East Baths, Cemeteries, Kilns, Venus Mosaic, Site Museum and Other Studies. Report
No. 2 (JRA Supplementary Series, 41). Portsmouth, RI, 25–8.
Wilson, A. I. (2011). ‘Developments in Mediterranean Shipping and Maritime Trade’,
in Robinson and Wilson (2011), 33–59.
Wilson, A., Schörle, K., and Rice, C. (2012). ‘Mediterranean Connectivity and Roman
Ports’, in S. J. Keay (ed.), Rome, Portus and the Mediterranean. Rome, 367–91.
 5

Driving Forces for Specialization:

Market, Location Factors,
Productivity Improvements
Kai Ruffing

The high degree of specialization of crafts and trades in the Greco-Roman world
has long been considered in modern research as a negative sign of the respective
underlying economies.1 This was especially true for the study of the economy of
the Roman Empire in the time of the ‘primitivist orthodoxy’, which prevailed in
modern historiography on the ancient economy from the 1970s until the 1990s.2
In the primitivist view and in research influenced by primitivist or substantivist
positions, the city of Rome of the early and high empire was seen as an example of
an absurd fragmentation of occupations in crafts and trades. Taking the evidence
of literary and epigraphic sources together, some 160 terms for crafts are attested
in the city of Rome alone, whereas one can find 225 designations for occupations
in the whole western part of the empire. These figures were thought to reflect an
extravagant division of labour, which was caused by the absence of specialized,
skilled craftsmen. The presence of large masses of slaves caused a strong seg-
mentation in production: in this way, unskilled labourers were able to produce at
a relative high level of quality. This, in turn, led to an increase in medium-sized
and large workshops, where the connection between master and apprentice was
lost. To support this interpretation, comparisons were made with early modern
metropoleis, such as eighteenth-century Paris or fifteenth-century Florence,
where only about a hundred crafts are attested. In his masterpiece on the history
of the Roman Empire, Karl Christ expressed a similarly negative view: he
believed that the Roman Empire was characterized by an archaic mode of
production. This was caused by extensive specialization, which led to small-

1
See, e.g. Morel (1991: 256–8). See also De Robertis (1946: 53–4), who had already taken such
positions.
2
Named in this way by Hopkins (1983: p. ix). But see Flohr and Wilson, Chapter 1, this volume.
116 Kai Ruffing

scale workshops combining manufacturing with retail. Owing to specialization,

competition, and a lack of capital, only an atomized mode of production was
possible—one that was based on the family as workforce.3
Harry Pleket, who compared the economic history of the Roman Empire
with that of modern Europe, developed a more positive interpretation of the
specialization of crafts in Rome. He identified some 500 Latin terms for crafts or
associations of craftsmen in the Roman Empire and some 200 for Rome itself. He
compared this evidence with mid-eighteenth-century London, where some 350
occupations are to be found. This led him to the conclusion that, in terms of craft
specialization, the Roman Empire was not very different from early modern
Europe.4 A more negative view on specialization was again formulated by Hans
Kloft, who saw the extensive craft specialization as a consequence of the humble
social status of craftsmen, though he also acknowledged that economic demand
and location factors were incentives for it.5 In general, most modern research is
now inclined to see economic conditions rather than social structures or a
dependence on slave labour as causes for extensive specialization. Increasingly,
this is also now judged as a positive indicator for the underlying economy.6
Thus, modern research has discussed specialization in the light of the
presumed character of the economy of the Roman Empire in general, using
occupational terms for crafts and trades in Latin literary texts and inscriptions
as evidence.7 This is not without methodological problems. First of all, in the
Latin epigraphic record, the city of Rome is heavily over-represented: only a
minority of texts comes from elsewhere in Italy or from the provinces.8
Secondly, modern research often focused on crafts, but to apply the modern
terminological differentiation between ‘craft’ and ‘trade’ to Latin (and Greek)
texts involves imposing a modern distinction on ancient evidence that is not
entirely unproblematic. To address these issues, some years ago I used the
evidence of Greek inscriptions and documentary papyri in an attempt to trace
chronological developments in the use of Greek occupational terms and to
analyse the development of specialization itself, especially in the Roman Em-
pire. In this context, causes for labour specialization were analysed, such as
location factors, apprenticeship, and competition.9 A key aim of this research

3
Christ (1995: 495–6).
4 5
Pleket (1990: 121). Kloft (1992: 214).
6
See, e.g. Kolb (1995: 464–9) regarding the city of Rome; Landmesser (2002: 76–8) regarding
classical Athens. In his innovative study of the ancient economy, which was based on the
theoretical framework of the German national economy, Heichelheim had already taken the
number of attested occupational terms as an indicator for the level of development of an
economy: Heichelheim (1938: 335–6, 373–4, 735, 819–21).
7
For the Latin occupational terms, see Petrikovits (1981a, b); Wissemann (1984). For modern
research on Latin occupational terms, see the excellent discussion in Cristofori (2004: 18–23).
8
See Petrikovits (1981a: 79); for the chronological and spatial distribution of Latin inscrip-
tions, see Eck (1997: 98–100).
9
See Ruffing (2008).
 Driving Forces for Specialization 117

was to show that specialization cannot be taken as an argument in favour of a

primitivist interpretation of the economy of the Roman Empire. The approach
made it possible to observe terminological changes during antiquity, which
meant that the sheer existence of designations for crafts and trades in itself
cannot be taken as evidence for an extravagant specialization. Moreover, it
became clear that even seemingly odd specializations were caused by the
market and its structural framework and that there was no linear development
of specialization in antiquity. Starting from the results of this project, this
chapter will discuss three conditions and incentives for the development of
different occupations in the Roman Empire: the demands of the market, the
desire to increase productivity, and location factors.

SPECIALIZATION

Before a discussion of driving forces for specialization, it is necessary to make

some remarks on the term itself and on the concept of ‘profession’ used to
identify specialized trades and crafts in the documentary Greek sources. The
concepts used in economic theory and in sociology in describing specialization
are discussed by Edward Harris in his groundbreaking paper regarding the
economy of classical Athens and the occupations attested there.10 Thus, in the
present context a few words may suffice.
Economic theory distinguishes between ‘horizontal’ and ‘vertical’ specializa-
tion. Horizontal specialization describes the diversity of goods and services
produced in a society by using different professional formations or work roles.
Thus, for example, the demand for skills for the production of amphorae is
different from that for the production of shoes or textiles, and so on. The number
of goods and services produced in an economy in this way is proportional to the
number of specializations. Vertical specialization, on the other hand, describes
the number of separate work roles and skills used in manufacturing a single
product. A good example is the building of an ancient ship, which requires a
set of different skills: carpentry, ironwork (for nails), rope-making, as well as
textile production (for the sails). Moreover, both the building process itself and
the supply of building materials and finished products need to be coordinated.
Sociologists use the concepts of Max Weber, who differentiated between ‘speci-
fication of function’ and ‘specialization of function’.11 In the first case, one worker
performs all steps in the production process for finishing a product; in the second
case, the product is made in different steps by different specialized workers.12

10
Harris (2001).
11
Weber (1947: 225); but see Tran (2011: 8-10) with a different approach focusing on technologies
used for the production of different goods and negating the usefulness of the concept of ‘crafts’.
12
Harris (2001: 70–1).
118 Kai Ruffing

Using Greek and Latin inscriptions and documentary papyri as a source for
analysing specialization in crafts and trades creates a certain problem regarding
the status of occupational or professional terms in these types of sources. First
of all, in many cases it is impossible to decide whether a specific term designates
a craft or a trade. This is, for example, true for all Greek and Latin substantives
ending with the suffixes -άριος or -arius, and for Greek substantives ending
in -ᾶς. For example, a κιναρᾶς may be a producer or seller of artichokes, or
both. At the same time, some terms that at first sight seem to be associated
with trade are used to designate crafts: thus, a ζυτοπώλης is a brewer, not a
trader of beer, although he also distributed his product.13 As a necessary
consequence, crafts and trades must be treated together; drawing a sharp
distinction between them is often impossible.14 Yet such a distinction is
also unnecessary, given the fact that presumably most artisanal products
reached the consumer through direct retail, rather than through networks
of trade: as a rule, craftsmen themselves sold their products in their shops,
and producers of agricultural goods sold these at markets themselves.
Both practices are well attested for—for example, the Sarapeion market of
Oxyrhynchos.15
Finally, some words need to be devoted to the use of terms such as
‘occupation’ and ‘profession’. In using documentary sources we deal with a
variety of contexts for naming crafts and trades in a written document. It is
sometimes very difficult to judge whether a term used to refer to a trade or
craft in a document is really an occupational term in Max Weber’s sense. For
Weber, a ‘profession’ is characterized through specification and through a
typical combination of the different capacities of an individual, which is the
basis of the ability to gain a livelihood continuously.16

T H E MA R K E T

Since the days of Adam Smith, it has generally been assumed that there is a
direct relation between specialization or division of labour and the market.17
For Smith, the size of the market was the factor determining the degree of
specialization. As the size of a market is defined by the number of persons who
interact by means of it, the size of the population of settlements and the degree

13
Drexhage (1997).
14
This is not only a problem for antiquity; it also exists in analysing occupations and
specialization in the nineteenth and twentieth centuries: see Leeuwen, Maas, and Miles (2002:
26–9).
15
SB 16.12695. On this document, see Parsons (2007: 103–4). For Rome itself, however, the
system of the retail trade is much more sophisticated: see Holleran (2012: 62–98, 194–231).
16 17
Hofbauer and Stooß (1977: 468). e.g. for the Roman world, Wilson (2008).
 Driving Forces for Specialization 119

of specialization are directly proportional.18 This basic assumption is also used

in the New Institutional Economics, at least by Douglass North.19 Applying it
to ancient economies in general and to the economy of the Roman Empire in
particular is legitimate and useful for two reasons: first, the Roman economy
without any doubt is to be characterized as a market economy; secondly,
specialization is a necessary consequence of economies beyond autarky.20
The correlation between the size of the population and the number of profes-
sions can be demonstrated, for example, for Rome. Certainly, it is due to the
epigraphic habit and certain social conditions that about three-quarters of
the occupational terms attested in Latin inscriptions come from Rome. On the
other hand, Rome with its population of about a million people was also the
largest market of the Roman Empire and depended heavily on commerce and
importation of goods of every kind from all over the empire and beyond. What
becomes visible here is a close connection between the size of population,
commerce, and importation, and the degree of specialization, which fits well
with the hypotheses of Adam Smith and Douglass North already mentioned.
While Rome is without parallels, a similar picture emerges if we compare the
evidence for Ephesos and other cities of Asia Minor.21 It is very interesting to
see that these principles, to a certain degree, were already observed by Plato, as
has been convincingly demonstrated by Edward Harris.22
If we take the proportionality between size of market and degree of special-
ization as a given, we must ask: what are the driving forces that cause this
interdependence? For Adam Smith the predisposition of human beings for the
exchange of goods was caused either by nature or by the human capacity to
speak and to think, which, in consequence, is the main driving force for
specialization. In his view, the discovery of individuals that they have certain
skills by which they can gain their livelihood with the exchange of goods and
services on the market causes specialization and consequently the division of
labour in a society.23 For Plato, on the other hand, specialization was caused by
a desire for greater efficiency and higher quality. In this way he believed a
general increase of wealth was possible.24 These explanations are certainly all
but satisfying. A more convincing explanation for the way in which the market
fostered specialization is the existence of competition between craftsmen,
retailers, and traders.
Competition as a phenomenon of ancient business life has not attracted a
lot of attention in modern research, although presumably it is of great
importance for specialization as well as for other aspects of the economic

18 19
Smith (1789: 16–22). North (1992: 141) with a focus on commerce.
20
See Temin (2001). The same is true for the Greek economies: see Loomis (1998: 251–4);
Bresson (2008).
21 22 23
Ruffing (2008: 230–69). Harris (2001: 72–7). Smith (1789: 16–18).
24
Plat,. Rep. 369–71; Harris (2001: 72).
120 Kai Ruffing

history of antiquity.25 The question of how competition influenced specializa-

tion in crafts and trades is connected with two other issues. First, the problem
of living standards, here understood as the possible extent of costs over
earnings in relation to purchasing power, and, secondly—as a result of
probably low living standards—the presumably high economic pressure on
individuals.26 Unfortunately we cannot go much beyond impressions, but it
seems that for the majority of craftsmen and retailers the earning of a
livelihood was not without difficulties.27 This means that there was strong
economic pressure on most individuals acting on the market to guarantee
their subsistence.28 At the same time, actors on the market also wanted to
make a profit. Two mosaics from Pompeii demonstrate this clearly—one is a
salute to profit, the other one to the pleasure caused by it.29 In a similar way,
the self-representation of merchants and craftsmen, which shows their pride
in their abilities and a presumed superiority over other individuals in the same
business, also must be seen as an indicator of competition.30
Thus, most craftsmen and retailers for one reason or another are likely to
have been exposed to intense competition on the market. There are also some
direct hints for this in our sources. A very interesting group of texts in this
respect are the curse tablets in which rivals in business life are bound or
cursed.31 Most of those tablets date to the Hellenistic period, but a couple of
texts date to the High Empire.32 For example, there is the binding of the profit
of a competitor in a curse tablet from Nomentum: ‘in these tablets I bind [his]
business profits and health [sc. of a certain Malcius, son or servant of
Nicona].’33
Other, and to modern eyes more efficient, strategies for avoiding competi-
tion on the market are attested in the literary sources. In the Metamorphoses of
Apuleius, a certain Aristodemus, a seller of honey, cheese, and other such
things, attempts to buy up all the cheese in a certain region, but fails to do so

25
See Wacke (1982); Kudlien (1994).
26
On living standards, see Rathbone (2009: 299); also Allen (2009: 327–8), with a discussion
of how ‘living standards’ can be assessed. The assumption of ‘living standards’ that can be
defined by an objective set of data is not without problems, since ‘living standards’ are defined
also by cultural practice and the subjective behaviour of individuals: see Drexhage, Konen, and
Ruffing (2002: 161).
27
See Drexhage, Konen, and Ruffing (2002: 181–3); see also Rathbone (2009: 322); Droß-
Krüpe (2011: 41–102).
28
But see Rathbone (2006: 108), who maintains that poverty in Roman Egypt was relatively
absent. I am grateful to Alan Bowman, who called my attention to this article. Against the view
formulated by Rathbone, see Harris (2011: 36, 54), who argues ‘that most parts of the Roman
Empire harboured a population of fluctuating size that was constrained to struggle for life below
the level of subsistence’.
29
CIL 10.874: salve lucru(m); CIL 10.875: lucrum gaudium.
30
On self-representation, see Drexhage (1990: 27–32); Ruffing (2004).
31 32
Gager (1992: 151–4); Kudlien (1994: 28–9). Gager (1992), nos 79, 82.
33
Gager (1992), no. 80 (undated). The English version is Gager (1992: 172).
 Driving Forces for Specialization 121

because another trader was faster than him and had already bought it up.34
A similar case is documented in a letter from Roman Egypt: a certain Ammo-
nios, his brother Apion, and a relative called Dionysius the Deaf intended to
buy all the peaches on the market to have a private monopoly, so that they
could sell peaches on the market without competition from others.35 A further
sign of competition on the market is the advertising of products by the
craftsmen themselves by means of tituli picti on amphorae or through the
use of shop signs.36
Without any doubt, competition on the market was further aggravated by
the concentration of crafts and trades in specific areas of towns, which is hinted
at by the existence of toponyms related to certain crafts and trade.37 In the city
of Rome toponyms such as the Street of the Sickle-Makers or the Street of the
Glassmakers are known.38 Other examples can be found in the east of the
Empire—a διῶρυξ γναφικός (fullers’ channel) at Antioch and a Cobblers’ Street
(Σκυτικὴ Πλατεῖα) at Apamea in Phrygia.39 Other examples are documented
in Roman Egypt, like the Great Street of the Goldsmiths (ῥύμη μεγάλη
χρυσοχόων) in Ptolemais Euergetis.40 Spatial concentration of crafts also
shows in the name of some professional associations, such as the purple
dyers ‘of the eighteenth street’ at Thessalonike (ἡ συνήθεια τῶ|ν πορφυροβάφ|ων
τῆς ὀκτω|καιδ ̣εκά ̣τη ̣ς [sc. πλατείας] κτλ).41 Further evidence for the concen-
tration of crafts in local areas can be found in the archaeological record.42 On
the one hand, these local concentrations of crafts and trades lowered transac-
tion costs, because consumers knew where they could buy certain products.
Yet, on the other hand, one can certainly assume that as a result competition
among craftsmen and traders became even fiercer.
Under these circumstances, further specialization could provide an ‘eco-
nomic niche’ for individuals. The concept of the ‘niche’ has its origins in
ecology, but has for a long time been used in economics too. Although it is not
clearly defined by economists, the term can be applied to ancient economic
history using a broad definition.43 In modern economic theory, an ‘economic
niche’ is a subsection of the market; focusing on a niche improves the level of
need satisfaction, lowers the intensity of competition, or protects against it

34 35
Apul. Met. 1.5.3–5 P. Ross. Georg III 3 (third century AD).
36
Curtis (1984–6); Gassner (1986: 43–5); Kruschwitz (1999); Berdowski (2003).
37
More elaborately on this issue, see Goodman, Chapter 13, this volume; ß-Krüpe,
Chapter 14, this volume.
38
Sickle-makers: inter falcarios, Cic. Cat. 1.8; glassmakers: vicus vitrarius, Curios. Urb. Rom.
73.12 (ed. Nordh); von Petrikovits (1981a: 70–1); Morel (1987); Kolb (1995: 496–507); Holleran
(2012: 51-60).
39
Antioch: SEG 35.1483 = AE 1986, 694 (AD 73–4); Apamea: IGR 4.790.
40
P. Berl. Leihg. II 42 A 12 (AD 176–9).
41
IG 10.2,1 291. On this inscription, see Robert (1937: 535 n. 3).
42
See Goodman, Chapter 13, this volume, on the spatial concentration of crafts and trades.
43
Trachsel (2007: 6–9, esp. 8).
122 Kai Ruffing

altogether, and it may foster a willingness to pay on the side of the customer.
Concentrating on a subsection of the market means that a producer or a trader
sells his products to a narrower group of consumers—for example, by con-
sciously choosing to focus on certain products and/or regions. Because of this
concentration on a market segment, a better satisfaction of consumer needs
becomes possible. This leads to a protection against competition or to a lower
intensity of competition, as competitors offering a wider spectrum of products
on the market may be less able to satisfy more specific needs. The ‘willingness
to pay’ is a consequence of this in the sense that consumers are willing to pay
more to have their needs satisfied more specifically or closely.44
While this is not the place to discuss these criteria and their application at
length, the use of applying the concept of the ‘economic niche’ to ancient
economic history becomes clear in the following example. Since clothes are a
basic need of human beings, their production and distribution were a sector of
major importance in the economy of the Roman Empire. Thus, many occupa-
tional terms dealing with the production and distribution of clothing are
documented in the literary and documentary sources, suggesting that many
entrepreneurs sought and found economic niches.45 One telling example of an
economic niche in this sector is the commerce in silk. From a dedicatory
inscription from Gabii we know a silk-trader called Aulus Plutius Epaphroditus,
who evidently had great economic success in this town near Rome. He was able
to make generous donations to the temple of Venus Vera Felix Gabina. On the
occasion of the dedication of these offerings he gave five denarii to each decurio,
three denarii to each sevir Augustalis, and one denarius to each tabernarius inside
the walls of Gabii. Furthermore he made another donation of HS 10,000 to the
city of Gabii in order to provide a public dinner for the local elite each year on the
occasion of his daughter’s birthday.46 Apparently, Aulus Plutius chose to limit
himself to selling only one product to his clientele, which consisted of a group of
consumers willing to buy silk for—presumably—a high price. The high price and
the nature of the product lowered the intensity of competition, and partially
protected Aulus Plutius, since he sold a product that was available only through
long-distance commerce via the harbours of the Levant or Egypt.47
Thus, we may conclude that the high economic pressure on individuals acting
on the market—either to gain enough for supplying basic needs or to make more
substantial profit—together with fierce competition could stimulate them to look

44
Trachsel (2007: 41–7).
45
Von Petrikovits (1981a: 123–4); Wissemann (1984); Labarre and Le Dinahet (1996); Vicari
(2001: 37–69, 94–115); Droß-Krüpe (2011: 47–102).
46
CIL 14.2793.
47
The import of silk via Palmyra is documented through silk draperies found in Palmyra:
Stauffer (1996: 426–7) and Falkenhausen (2000). The import of silk via Egypt is documented in
the Periplus Maris Erythraei: see Periplus Maris Erythraei, 49.56. Other evidence for the import
of silk is provided by Dig. 39.4.16.7.
 Driving Forces for Specialization 123

for an economic niche. In practice, limiting oneself to a segment of the market of

course is nothing other than specialization. Indeed, there are some hints in the
sources of people in search of such a niche. A good example is a certain
Onesimus, whose activities in various years are documented by five papyri.
Onesimus was acting as a trader of young pigs (χοιριδέμπορος), as a butcher
(μάγειρος), and as a seller of meat (κρεοπώλης καὶ ταριχευτής).48 Another
possible example is provided by a graffito from Pompeii, with a long list of
occupations that had allegedly been exercised by one man:
[cum] de[d]uxisti octies tibi superest ut (h)abeas sedecies coponium fecisti
cretaria fecisti salsamentaria fecisti pistorium fexisti agricola fuisti aere minutaria
fecisti propola fuisti languncularia nunc facis si cunnu(m) linx<s>e<e>ris, con-
summaris omnia.
Since you’ve held eight jobs all that’s left for you now is to have sixteen. You
worked as an innkeeper, you worked as a clay man, you worked as a pickler, you
worked as a baker, you were a farmer, you worked as a maker of bronze trinkets,
you were a retailer, now you work as a jug man. If you perform cunnilingus you’ll
have done everything.49
To sum up: specialization in one economic niche was a response to economic
pressure and competition on the market enabling people to earn a livelihood
or even to be very successful on the market and make substantial profit. The
consequence of this is that the market is to be seen as one of the main driving
forces for professional specialization.

IN CREA SE O F PRO DUCTIVI TY

Specializing in one product had further advantages for craftsmen: it made it

possible to increase both the quality and the quantity of their output. The
quantity of goods produced per day rises notably through specialization,
because, as has already been noticed by Adam Smith, it allows workers to
focus completely on one specific production process.50 A good example of that
is the nailsmith. The occupational term (ἡλοκόπος) emerges in the Greek
documentary sources in the second century AD.51 In Latin inscriptions, too,
there is evidence for the profession of the nailsmith (clavarius) from the
imperial period onwards.52 Nails were also produced by normal smiths, but,
if one looks at evidence from the medieval and early modern period, one gets

48
Drexhage (1991: 34–6).
49
CIL 4.10150. Translation: after Peña and McCallum (2009: 63), corrected.
50 51
Smith (1789: 12). Drexhage (2008: 169–72).
52
Petrikovits (1981a: 91) with the relevant inscriptions. As far as I see there is no further
evidence.
124 Kai Ruffing

the impression that the productivity of someone specialized in producing nails

was considerably higher than that of a ‘normal’ smith. Similarly, Adam Smith
in the chapter in his Wealth of Nations on the division of labour took the
example of the production of nails when he discussed the increase of prod-
uctivity and quality by specialization. According to him, a smith not accus-
tomed to making nails can produce only 200 or 300 pieces a day, whereas a
smith frequently making nails can produce between 800 and 1,000 pieces a
day. A specialized nail-smith, on the contrary, was able to make more than
2,300 nails a day.53 In today’s research one finds that in the later Middle Ages
and in the modern era production of between 800 and 4,000 pieces per day by
nailsmiths was possible.54 Since the method of producing nails did not change
between antiquity and the early modern era, we have every reason to believe
that a similar increase of productivity was possible for nailsmiths in the
Roman Empire.55 If this is true, the emerging profession of the nailsmith is
at least a hint at the possibility of increasing productivity by specialization.56
While the case of the nailsmith can be used as an example to show
increasing productivity by means of horizontal specialization, it is also pos-
sible to do this by means of vertical specialization.57 Although evidence for
vertical specialization in antiquity is rare, some examples can be mentioned.
A good starting point, although seemingly rather unimpressive at first sight, is
a dedicatory inscription from third-century Side by a certain Aurelius Ken-
deas, commemorating the harmony (ὁμόνοια) between the ‘sifters of flour’
(ἀλευροκαθάρται) and the ‘formers of bread’ (ἀβακῖται).58 The importance of
this inscription, which has been dated between AD 220 and 250, becomes clear
if we read it in the light of a famous funerary monument in Rome: that of
Marcus Vergileius Eurysaces, a baking contractor, as he is styled in his
inscription.59 On the monument, which dates from the first decade of the
Augustan period, there is a frieze illustrating the individual steps of labour in
the mass production of bread.60 The sifters of flour can be found on the
northern part of the frieze, while the formers of bread are displayed on the
southern side. In this light, and together with other evidence for the existence
of huge bakeries in Side, the inscription of Aurelius Kendeas becomes a clear
indication for the production of bread in third-century Side, with a division of

53 54
Smith (1789: 12). Stahlschmidt (1991: 179).
55
On the production method, see Hug (1935: 1581–2). For lower estimates, see Drexhage
(2008: 178).
56
Wilson (2008).
57
On this topic, see also Silver (2009); Broekaert (2012).
58
SEG 33.1165; IK 43.30, pp. 296–7. See the rich commentary to this inscription by Nollé
(1983). See also Zimmermann (2002: 167–8).
59
ILS 7460 a–d: hoc est monimentum Marcei Vergileius Eurysacis pistoris redemptoris appa-
ret. See also discussion by Wilson (2008).
60
On the date of the grave monument, see Ciancio Rossetto (1973: 66–7); Zimmer
(1982: 108).
 Driving Forces for Specialization 125

labour in the production process apparently similar to that in Rome in the

time of Eurysaces.61 In the case of Eurysaces, the division of labour meant an
increase of productivity, because every step in the production process was
reduced to a simple operation. This allowed Eurysaces to mass-produce bread.
The whole production process is vertically specialized, which can be inter-
preted as an indication that Eurysaces made the most of the opportunities
provided by the emerging new market of his era.62 A similar practice seems to
have developed in Roman imperial Side.
In a certain way, specialization aimed at increasing productivity is also fostered
by the market. For example, there are several indications that there was a high
demand for nails in the Roman Empire. In the Roman legionary fortress of
Inchtuthil in Scotland over 875,000 nails were found, buried unused when the
unfinished fort was abandoned.63 Papyrological evidence from Roman Egypt
also points in this direction.64 Further, that there was already a very huge market
in Rome for bread in Augustan times is beyond doubt, in view of the number of
inhabitants (about a million people).65 The fact that even the medium-sized town
of Pompeii had at least thirty-seven specialized workshops dedicated to the
production of bread is all the more illuminating in this respect.66 Thus we may
conclude that high demand fostered an increase of productivity, which was made
possible through horizontal as well as vertical specialization.

LOC ATION FACTORS

All aspects of the production, distribution, and consumption of goods are

closely linked to the realities of economic geography. In modern economic
geography, this connection is described using the theory of location factors.
These location factors are, on the one hand, determined by the geomorph-
ology of a given region and the available natural resources and, on the other
hand, by anthropogenic factors. Anthropogenic factors include, among others,
the specific roles of a location on the economic, political, and religious level,
the changes to its natural environment caused by human activity, and its place
in the geographies of transport.67 All factors are interlinked, which results in a
very complex framework for the economy of individual localities, where
changes in one factor influence all others. Discussing this in great detail is
clearly beyond the scope of this chapter, but some observations may serve to

61 62 63
Nollé (1983: 135–9). Silver (2009: 177–8). Wright (1961: 160)
64 65
Drexhage (2008: 173–8). Kolb (1995: 457).
66
Mayeske (1972: 167). See Monteix, Chapter 10, this volume.
67
For a catalogue of location factors, see Fellmeth (2002: 7–9).
126 Kai Ruffing

illustrate how location factors could be a major driving force for professional
specialization.
A good example is Hierapolis in Phrygia with its hot springs. These springs
were the raison d’être of the town, because, together with the Ploutonion, they
were the centre of the sanctuary of Apollo. The Ploutonion was believed to be
an entrance to Hades, because every creature entering this cave died quickly.
Strabo tells the story that he himself visited the Ploutonion and threw sparrows
into the cave, which died as soon as they breathed its air.68 The deadliness of
the Ploutonion, which is prominent in ancient literature, was the result of the
chemical composition of the water of the hot springs: it contained a high
percentage of carbonic acid gas.69 Furthermore, the water contained a high
concentration of lime, which was responsible for the other natural spectacle of
Hierapolis: the lime deposits.70 Thanks to the water and its natural wonders,
Hierapolis and its sanctuary were an attractive destination for travellers in
Roman times.71
Yet the water of Hierapolis also had another important feature. According to
Strabo, it was very useful for dyeing wool. Using a plant not clearly named by
the geographer, the craftsmen of the town were able to produce red clothes of
high quality.72 As a consequence, textile production was of great importance for
the economy of the city. This also shows in the city’s epigraphic record, in which
craftsmen related to the production of clothes are rather prominent, suggesting
that producers and traders of textiles had an elevated social position, which
must have been the result of their economic success.73 The inscriptions also
reveal a strong horizontal specialization in textile production. Inscriptions refer
to wool-cleaners (ἐριοπλύται), dyers (βαφεῖς), who honoured the boule with a
statue, and red-dyers. Furthermore there are sellers of red clothes.74
The following scenario may help to clarify the situation at Hierapolis. The
natural location factors—especially the hot springs and their chemical
properties—are fundamental features for the sanctuary as well as for the produc-
tion of clothes. The sanctuary of Apollo and the Ploutonion were a reason for the

68 69 70
Strabo, 13.4.14. Ritti (1985: 7–15). Humann (1898: 2–4).
71
Casson (1974: 232); Kreitzer (1998: 382).
72
On red dyeing in Hierapolis, see Herz (1985: 98).
73
See also Arnaoutoglou, Chapter 12, this volume.
74
Wool-cleaners: Labarre and Le Dinahet (1996: no. 64; cf. IGR 4.821), dated to the third
century AD. Dyers: SEG 41.1201; cf. Labarre and Le Dinahet (1996: no. 56); IK 36.1/6*; cf. Labarre
and Le Dinahet (1996: no. 67); Labarre and Le Dinahet (1996: no. 68). Red-dyers: Labarre and Le
Dinahet (1996: no. 66; cf. IGR 4.816); Labarre and Le Dinahet (1996: no. 59), second century AD;
Labarre and Le Dinahet (1996: no. 57), second/third century AD; Labarre and Le Dinahet (1996:
no. 62; see also SEG 46.1656; AE 1994, 1660; Miranda (1999: 131, 23)), second/third century AD;
Labarre and Le Dinahet (1996: no. 65), AD 206/208–9, see also Ritti (1985: 108, b); Labarre and Le
Dinahet (1996: no. 63; cf. IGR 4.822), third century AD. Sellers of red clothes: Labarre and Le
Dinahet (1996: no. 58); Labarre and Le Dinahet (1996: no. 69; cf. BE (1971: 650–37).
 Driving Forces for Specialization 127

supra-regional importance of the cult place: the cult evidently attracted visitors
from at least Asia Minor and maybe from the entire eastern part of the empire.75
In this way, the special red clothes of Hierapolis came to the attention of people in
Asia Minor, which increased demand for this product. As a consequence, pro-
duction increased, which fostered horizontal specialization in crafts and trade.
This appears to have generated wealth for the Hierapolitan producers and traders
of clothes: they become visible in the epigraphic record as an important social
group. For this reason, Pleket considered Hierapolis a Weberian ‘producer city’.76
The case of Hierapolis demonstrates the importance of interlinked location
factors such as natural resources, religious centrality, and transport geography
for the specialization in crafts and trades.

CO NCLUSION

The market was the main driving force for specialization in crafts and trades, and
the degree of specialization was proportional to the extent of the market—that is,
the number of individuals who were involved in the market as buyers and sellers.
The sheer number of people acting on the market created competition among
producers and sellers of goods. Evidently such competition in the Roman Empire
was very fierce, and it was often further aggravated by the concentration of
certain crafts in specific areas of towns. This competition also meant that there
was a high economic pressure on most individuals to make ends meet. One
response to this pressure was to seek an economic niche, which allowed people to
avoid or minimize competition. Operating within an economic niche meant
specialization, and this specialization—in its horizontal as well as in its vertical
form—in certain sectors also brought with it an increase in productivity. This
increase in productivity, in turn, made it easier for people to avoid competition.
Yet, throughout the Roman world, the market and its niches were influenced to a
considerable extent by location factors, which in this way provided the frame-
work for the economic performance of a given location or region in general and
for specialization in crafts and trades in particular.

REFERENCES

Allen, R. C. (2009). ‘How Prosperous were the Romans?: Evidence from Diocletian’s
Price Edict (AD 301)’, in A. K. Bowman and A. I. Wilson (eds), Quantifying the

75
On the supra-regional importance of the cult of Hierapolis and its consequences for the
economy, see Ruffing (2009). See also Ruffing (2008: 256–65).
76
Pleket (1984: 22–6; 1988: 33–5). See also Ritti (1995).
128 Kai Ruffing
Roman Economy: Methods and Problems (Oxford Studies on the Roman Economy).
Oxford, 327–45.
Berdowski P. (2003). ‘Tituli picti und die antike Werbesprache für Fischprodukte’,
MBAH 22/2: 18–55.
Bresson, A. (2008). L’Économie de la Grèce des cités (fin VIe–Ier siècle a.C.), ii. Les
Espaces de l’échange. Paris.
Broekaert, W. (2012). ‘Vertical Integration in the Roman Economy: A Response to
Morris Silver’, AncSoc 42: 109–25.
Casson, L. (1974). Travel in the Ancient World. London.
Christ, K. (1995). Geschichte der römischen Kaiserzeit: Von Augustus bis zu Konstantin,
3rd edn. Munich.
Ciancio Rossetto, P. (1973). Il sepolcro del fornaio Marco Virgilio Eurisace a Porta Maggiore.
Rome.
Cristofori, A. (2004). Non arma virumque: Le occupazioni nell’epigrafia del Piceno, 2nd
edn (Tarsie, 2). Bologna.
Curtis, R. I. (1984–6). ‘Product Identification and Advertising on Roman Commercial
Amphorae’, AncSoc 15–17: 209–10.
de Robertis, F. M. (1946). La organizzazione e la tecnica produttiva: Le forze di lavoro e
i salari nel mondo romano. Naples and Bari.
Drexhage, H.-J. (1990). ‘Zum Selbstverständnis arbeitender Menschen im Imperium
Romanum’, Humanistische Bildung 14: 7–40.
Drexhage, H.-J. (1991). ‘Einige Bemerkungen zu den ἔμποροι und κάπηλοι im
römischen Ägypten (1.–3. Jh. n.)’, MBAH 10/2: 28–46.
Drexhage, H.-J. (1997). ‘Bierproduzenten und Bierhändler in der papyrologischen
Überlieferung’, MBAH 16/2: 32–9.
Drexhage, H.-J. (2008). ‘Einige Bemerkungen zu den Nagelschmieden (ἡλοκόποι) nach
dem papyrologischen und epigraphischen Befund’, MBAH 25/2: 169–82.
Drexhage, H.-J., Konen, H., and Ruffing, K. (2002). Die Wirtschaft des Römischen
Reiches (1.–3. Jahrhundert): Eine Einführung. Berlin.
Droß-Krüpe, K. (2011). Wolle—Weber—Wirtschaft: Die Textilproduktion der römischen
Kaiserzeit im Spiegel der papyrologischen Überlieferung (Philippika, 46). Wiesbaden.
Eck, W. (1997). ‘Lateinische Epigraphik’, in F. Graf (ed.), Einleitung in die lateinische
Philologie. Stuttgart and Leipzig. 92–111.
Falkenhausen, L. von (2000). ‘Die Seiden mit chinesischen Inschriften’, in A. Schmidt-
Colinet and A. Stauffer (eds), Die Textilien aus Palmyra. Mainz, 58–81.
Fellmeth, U. (2002). ‘Eine wohlhabende Stadt sei nahe . . .’: Die Standortfaktoren in der
römischen Agrarökonomie im Zusammenhang mit den Verkehrs- und Raumord-
nungsstrukturen im römischen Italien. St Katharinen.
Gager, J. G. (1992). Curse Tablets and Binding Spells from the Ancient World. Oxford.
Gassner, V. (1986). Die Kaufläden in Pompeji. Vienna.
Harris, E. M. (2001). ‘Workshop, Marketplace and Household: The Nature of
Technical Specialization in Classical Athens and its Influence on Economy and
Society’, in P. Cartledge, E. E. Cohen, and L. Foxhall (eds), Money, Labour and
Land: Approaches to the Economy of Ancient Greece. London and New York, 67–99.
Harris, W. V. (2011). ‘Poverty and Destitution in the Empire’, in W. V. Harris (ed.),
Rome’s Imperial Economy: Twelve Essays. Oxford, 27–54.
 Driving Forces for Specialization 129
Heichelheim, F. M. (1938). Wirtschaftsgeschichte des Altertums vom Paläolithikum bis zur
Völkerwanderung der Germanen, Slaven und Araber, vols 1–3. Leiden (repr. Leiden,
1969).
Herz, P. (1985). ‘Parthicarius und Babyloniarius: Produktion und Handel feiner
orientalischer Lederwaren’, MBAH 4/2: 89–106.
Hofbauer, H., and Stooß, F. (1977). ‘Beruf und Berufsbildung’, in W. Albers, K. E. Born,
E. Dürr, H. Hesse, A. Kraft, H. Lampert, K. Rose, H. H. Rupp, H. Scherf, K. Schmidt,
and W. Wittmann (eds), Handwörterbuch der Wirtschaftswissenschaft, i. 468–78.
Holleran, C. (2012). Shopping in Ancient Rome: The Retail Trade in the Late Republic
and the Principate, Oxford.
Hopkins, K. (1983). ‘Introduction’, in P. Garnsey, K. Hopkins, and C. R. Whittaker
(eds), Trade in the Ancient Economy. London, pp. ix–xxv.
Hug, A. (1935). s.v. Nagel, RE XVI.2: 1576–82.
Humann, C. (1898). ‘Topographie und Bauten’, in C. Humann, C. Cichorius,
W. Judeich, and F. Winter (eds), Altertümer von Hierapolis (Jahrbuch des Kaiserlich
Deutschen Archäologischen Instituts Ergänzungsheft, 4). Berlin, 1–17.
Kloft, H. (1992). Die Wirtschaft der griechisch-römischen Welt: Eine Einführung.
Darmstadt.
Kolb, F. (1995). Rom: Die Geschichte der Stadt in der Antike. Munich.
Kreitzer, L. J. (1998). ‘The Ploutonion of Hierapolis and the Descent of Christ into the
“Lowermost Parts of the Earth” (Ephesians 4,9)’, Biblica 79: 381–93.
Kruschwitz, P. (1999). ‘Römische Werbeinschriften’, Gymnasium 106: 231–53.
Kudlien, F. (1994). ‘Die Rolle der Konkurrenz im antiken Geschäftsleben’, MBAH 13/1:
1–39.
Labarre, G., and Le Dinahet, M.-Th. (1996). ‘Les Métiers du textile en Asie Mineure de
l’époque hellenistique à l’époque imperiale’, in Aspects de l’artisanat du textile dans
le monde mediterranéen (Égypte, Grece, monde Romain) (Collection de l’Institut
d’archéologie et d’histoire de l’Antiquité, 2). Paris, 49–115.
Landmesser, D. (2002). Wirtschaftsstil und wirtschaftliche Entwicklung im klassischen
Athen. Munich.
Leeuwen, M. H. D. van, Maas, I., and Miles, A. (2002). HISCO: Historical International
Standard Classification of Occupations. Löwen.
Loomis, W. T. (1998). Wages, Welfare Costs and Inflation in Classical Athens. Ann
Arbor.
Luik, M. (2002). ‘Handwerk in den Vici des Rhein-Maas-Gebietes’, in K. Strobel (ed.),
Die Ökonomie des Imperium Romanum: Strukturen, Modelle und Wertungen im
Spannungsfeld von Modernismus und Neoprimitivismus. Akten des 3. Trierer Sym-
posiums zur Antiken Wirtschaftsgeschichte (Pharos, 17). St Katharinen, 169–209.
Mayeske, B. J. B. (1972). ‘Bakeries, Bakers, and Bread at Pompeii: A Study in Social and
Economic History’. Ph.D. dissertation, Ann Arbor.
Miranda, E. (1999). ‘La comunità giudaica di Hierapolis di Frigia’, EA 31: 109–55.
Morel, J.-P. (1987). ‘La Topographie de l’artisanat et du commerce dans la Rome
antique’, in L’Urbs: Espace urbain et histoire (Ier siècle av. J.-C.–IIIe siècle ap. J.-C.):
Actes du colloque international organisé par le Centre national de la recherche
scientifique et l’École française de Rome (Rome, 8–12 mai 1985) (Collection de l’École
française de Rome, 98). Rome, 127–55.
130 Kai Ruffing
Morel, J.-P. (1991). ‘Der Handwerker’, in A. Giardina (ed.), Der Mensch in der
römischen Antike. Frankfurt–New York–Paris: 243–75.
Nollé, J. (1983). ‘Die Eintracht der Mehlsieber und Brotformer in Side’, EA 1: 131–40.
North, D. (1992). Institutionen, institutioneller Wandel und Wirtschaftsleistung (Die
Einheit der Gesellschaftswissenschaften, 76). Tübingen.
Parsons, P. (2007). City of the Sharp-Nosed Fish: Greek Lives in Roman Egypt. London.
Peña, J. T., and McCallum, M. (2009). ‘The Production and Distribution of Pottery at
Pompeii: A Review of the Evidence; Part 1, Production’, American Journal of
Archaeology 113/1: 57–79.
Petrikovits, H. von (1981a). ‘Die Spezialisierung des römischen Handwerks’, in
H. Jankuhn, W. Janssen, R. Schmidt-Wiegand, and H. Tiefenbach (eds), Das
Handwerk in vor- und frühgeschichtlicher Zeit, Teil I: Historische und rechtshistor-
ische Beiträge und Untersuchungen zur Frühgeschichte der Gilde: Bericht über die
Kolloquien der Kommission für die Altertumskunde Mittel- und Nordeuropas in den
Jahren 1977 bis 1980 (Abhandlungen der Akademie der Wissenschaften zu Göttingen,
Philologisch-Historische Klasse III 122). Göttingen, 63–132.
Petrikovits, H. Von (1981b). ‘Die Spezialisierung des römischen Handwerks II
(Spätantike)’, ZPE 43: 285–306.
Pleket, H. W. (1984). ‘Urban Elites and the Economy in the Greek Cities of the Roman
Empire’, MBAH 3/1: 3–36.
Pleket, H. W. (1988). ‘Greek Epigraphy and Comparative Ancient History: Two Case
Studies’, EA 12: 25–37.
Pleket, H. W. (1990). ‘Wirtschaft’, in W. Fischer, J. A. van Houtte, H. Kellenbenz,
I. Mieck, and F. Vittinghoff (eds), Handbuch der europäischen Wirtschafts- und
Sozialgeschichte, i. Stuttgart, 25–160.
Rathbone, D. (2006). ‘Poverty and Population in Roman Egypt’, in M. Atkins and
R. Osborne (eds), Poverty in the Roman World. Cambridge, 100–14.
Rathbone, D. (2009). ‘Earnings and Costs: Living Standards and the Roman Economy
(First to Third Centuries AD)’, in A. K. Bowman and A. I. Wilson (eds), Quantifying
the Roman Economy: Methods and Problems (Oxford Studies on the Roman Econ-
omy). Oxford, 299–326.
Ritti, T. (1985). Fonti letterarie ed epigrafiche (Hierapolis. Scavi e Ricerche, 1). Roma.
Ritti, T. (1995). ‘Associazioni di mestiere à Hierapolis di Frigia’, in B. M. Giannattasio
(ed.), Atti. VII Giornata Archeologica: Viaggi e commerci nell’antchità. Genova, 65–84.
Robert, L. (1937). Études anatoliennes: Recherches sur les inscriptions grecques de l’Asie
Mineure. Paris (repr. Amsterdam, 1970).
Ruffing, K. (2004). ‘Die Selbstdarstellung von Händlern und Handwerkern in den
griechischen Inschriften’, MBAH 23.2: 85–101.
Ruffing, K. (2008). Die berufliche Spezialisierung in Handel und Handwerk: Untersu-
chungen zu ihrer Entwicklung und zu ihren Bedingungen in der römischen Kaiserzeit
im östlichen Mittelmeerraum auf der Grundlage der griechischen Inschriften und
Papyri (Pharos, 24). Rahden/Westfalen.
Ruffing, K. (2009). ‘Thermalquellen und Kult: Das Beispiel Hierapolis’, in E. Olshausen
and V. Sauer (eds), Die Landschaft und die Religion: Stuttgarter Kolloquium zur
Historischen Geographie des Altertums 9, 2005 (Geographica Historica, 26). Stutt-
gart: 287–300.
 Driving Forces for Specialization 131
Silver, M. (2009). ‘Glimpses of Vertical Integration/Disintegration in Ancient Rome’,
AncSoc 39: 171–84.
Smith, A. (1789). The Wealth of Nations: An Enquiry into their Nature and Origins. 5th
edn. London (German translation: Der Wohlstand der Nationen: Eine Untersuchung
seiner Natur und seiner Ursachen. Munich, 1974).
Stahlschmidt, R. (1991). ‘Nagelschmied’, in R. Reith (ed.), Lexikon des alten Handwerks.
Vom Spätmittelalter bis in das 20. Jahrhundert. Munich, 176–81.
Stauffer, A. (1996). ‘Textiles from Palmyra: Local Production and the Import and
Imitation of Chinese Silk Weavings’, AAAS 42: 425–30.
Temin, P. (2001). ‘A Market Economy in the Early Roman Empire’, JRS 91: 169–81.
Trachsel, M. (2007). Nischenstrategien und ihre Bedeutung für den Unternehmenserfolg.
Wiesbaden.
Tran, N. (2011). ‘De “l’artisanat” aux métiers: Quelques réflexions sur les savoir-faire
du monde romain à partir de l’exemple pompéien’, in N. Monteix and N. Tran
(eds), Les Savoirs professionnels des gens de métier: Études sur le monde du travail
dans les sociétés urbaines de l’empire romain (Collection du Centre Jean Bérard, 37.
Archéologie de l’artisanat antique, 5). Naples, 7–26.
Vicari, F. (2001). Produzione e commercio dei tessuti nell’Occidente romano (BAR
International Series, 916). Oxford.
Wacke, A. (1982). ‘Wettbewerbsfreiheit und Konkurrenzverbotsklauseln im antiken
und modernen Recht’, ZRG 99: 188–215.
Weber, M. (1947). The Theory of Social and Economic Organization. New York.
Wilson, A. I. (2008). ‘Large-Scale Manufacturing, Standardization, and Trade’, in
J. P. Oleson (ed.), Handbook of Engineering and Technology in the Classical
World. Oxford, 393–417.
Wissemann, M. (1984). ‘Die Spezialisierung des römischen Handels’, MBAH 3/1:
116–24.
Wright, R. P. (1961). ‘Roman Britain in 1960’. JRS 51: 157–98.
Zimmer, G. (1982). Römische Berufsdarstellungen (Archäologische Forschungen, 12).
Berlin.
Zimmermann, C. (2002). Handwerkervereine im griechischen Osten des Imperium
Romanum (Römisch-Germanisches Zentralmuseum, Forschungsinstitut für Vor-
und Frühgeschichte, 57). Mainz.
 6

Fashionable Footwear: Craftsmen

and Consumers in the North-West
Provinces of the Roman Empire
Carol van Driel-Murray

In the Roman world, leather was used for a multitude of purposes, and it
specifically played a key role in footwear. This chapter will focus on the
economy of footwear and will analyse the remarkable similarities, throughout
the empire, in both the final products and the technologies used in making
them. Globalization is too fraught a concept, but there does seem to be
integration to the extent that travellers would be familiar with most of the
footwear displayed on market stalls wherever they went (Figs 6.1 and 6.7).1
Regional differences may have persisted to a degree—certain forms of sandals
in Egypt, for instance, or a preference for single-piece shoes (so-called carba-
tinae) in Germany—but craftsmen everywhere seem to have been able to
participate in an empire-wide dialogue that comes close to our concept of
‘fashion’, in the sense of the contemporaneous stylistic development of cloth-
ing accessories across a wide geographical area and available to almost all
social classes.2
The archaeological record is somewhat uneven in that finds tend to be con-
centrated at military settlements and in the frontier regions, while comparative
material from the civilian hinterland is distressingly scarce. Nevertheless, in
addition to a number of huge complexes, there are sufficient scattered finds to

1
Woolf (1998: 169 ff.) also emphasizes the role of common ‘goods’ in the development of
provincial culture.
2
For dating and development, see van Driel-Murray (2001b). The term carbatinae is one of
convenience used by archaeologists to identify footwear made of a single piece of leather, seamed at
the back, of type without modern equivalents (Fig. 6.1, nos 10, 20, 25). For details of manufacture
and shoe styles, see van Driel-Murray (2001a) and Volken (2014). See also Goldman (1994) and
Leguilloux (2004), which includes Egyptian material,
 Fashionable Footwear 133

1 2

6 7
5
4
9

10

11
8
12

14
13
18
17

16
15

20 19
21

24 22
23
25

27
26 28

Fig. 6.1. Selection of Roman footwear, from first century AD (top left) to fourth
century (bottom right). Image: Carol van Driel-Murray.

confirm that footwear was easily available, widely used, and readily discarded by
both urban and rural populations (see Table 6.1). It was not a scarce and
luxurious good, available only to the rich: indeed, footwear gave people of modest
means a way of exhibiting individuality, taste, and style. As such it is not only the
craftsmen but also the consumers who need to be taken into account when
attempting to explain the remarkable similarities in the surviving material.
134 Carol van Driel-Murray

INTEGRATION

Images on coinage may have introduced the provincial population to elite

hairstyles, but how did shoemakers acquire the knowledge of the latest models
and the special technologies used to make them? That the caligae worn by
Roman soldiers in Augustan Mainz are similar to those left by legionaries at
Qasr Ibrim or Didymoi in Egypt is perhaps explicable in terms of military
manufacturing conventions (Fig. 6.1, no. 4), but other parallels are more
intriguing.3 With remarkable perspicacity, Martine Leguilloux recognized
that a small fragment from Didymoi resembles a child’s sandal from a barrack
room in Vindolanda (Fig. 6.2).4 Identical shoes, of an infrequent and compli-
cated pattern, appearing at the extreme north and the extreme south of the
empire might suggest the movement of an individual, but this instance is by no
means unique. Also recorded from Didymoi are front-laced shoes with frilled
edges (style designation: Irthing, Fig. 6.3), well known in north Britain in the
Antonine period, as well as the slightly later shoes with integral laces of style
Geltsdale (Fig. 6.1, no. 16).5 Many more examples from Egypt or the Near East
that happen to find their way into the published record can be paralleled in
northern Europe, suggesting a considerable degree of uniformity in footwear

Fig. 6.2. Child’s sandal from Vindolanda period IV (scale 1:3). Image: Carol van
Driel-Murray.

3
Van Driel-Murray (1985; 1999a; 2001a: 362–4).
4
Leguilloux (2006: 31, fig. 9); van Driel-Murray (1998: fig. 5).
5
For ease of identification, common shoe styles are named after a settlement site or a location
near a characteristic find spot: van Driel-Murray (2001b); Volken (2014: 78-9). Irthing: Leguilloux
(2006: pl. 26, no. 135), cf. Charlesworth and Thornton (1973: no. 5). Geltsdale: Leguilloux (2006:
pl. 26, no. 134) (laces snapped); cf. van Driel-Murray (2001a: 366–7). In all these cases, as with the
caligae, there are sufficient differences for local manufacture to be certain.
 Fashionable Footwear 135

design and technology throughout the Roman period. Connections over vast
regions are even more pronounced in the case of sandals. Not only do the
shapes, decorative motifs, and quite complex details of manufacturing tech-
nology appear in every province, but the entire trajectory of stylistic develop-
ment also runs parallel throughout the empire. Thus the exaggerated shape of
sandals after the second quarter of the third century can be paralleled at
numerous sites all over the empire, from remote villas in the Yorkshire
Wolds to North Africa (see Fig. 6.8).6
Remarkable expressions of what might be called Zeitgeist are the patterns in
which shoe nails are arranged. Basically, the nails serve to hold the shoe
construction together and to reinforce the sole, but shoemakers regularly
embellished their work with elaborate, non-essential decorative patterns
(Fig. 6.5). Some, like tridents and swastikas, occur fairly regularly, but others
are quite restricted in their popularity. Thus diamonds tend to occur after
AD 150, elaborate tendrils appear after AD 170, S’s after c.190, big asymmetrical

Fig. 6.3. Shoe style Irthing from Didymoi, Egypt. Image: Carol van Driel-Murray after
Leguilloux (2006: pl. 26).

Fig. 6.4. Shoe style Irthing from Hardknott UK. Image: Charlesworth and Thornton
(1973: fig. 5). Not to scale.

6
Dunbabin (1990); Mould (1990), van Driel-Murray (2001b: fig. 3).
136 Carol van Driel-Murray

a b c d

e f g h

Fig. 6.5. Diagrammatic selection of nailing patterns. Image: Carol van Driel-Murray.

S-shapes c. AD 220–30, while groups of triple nails are characteristic of the

second half of the third and early fourth centuries.7 These are not random
momentary inspirations: people evidently drew on common experiences that
could be expressed in a tangible form. Whatever the actual meaning of the
symbolism, or the methods by which this knowledge was communicated,
the choices made clearly involve the mechanisms of production as well as
the desires of consumers.

LEATHER P RODUCTION AS AN ALIEN TECHNOLOGY

In the northern provinces, the Roman Conquest marked a distinct change in

leatherworking practices. Not only was the method of vegetable-tanning intro-
duced as a new procedure, but an entire package of novel products appeared,
together with the specific technologies associated with their manufacture.8 There
was a considerable increase in the scale of leather production, and procedures

7
Van Driel-Murray (1999b: fig. 1).
8
Groenman-van Waateringe (1967); van Driel-Murray (2001a: 345–8; 2008); Winterbottom
(2009).
 Fashionable Footwear 137

Fig. 6.6. Vindolanda: off-cut stamped with C(oh). IX Bat(avorum) (scale 1:3). Image:
Carol van Driel-Murray.

became more complex and time-consuming. Newly introduced, alien technolo-

gies often remain divorced from the pre-existing economic structures, and retain
their separate organization primarily on account of capital investment and the
restriction of technical knowledge.9 In the case of tanning, we may also be seeing
a shift from female-controlled skin-processing to male-dominated large-scale
manufacturing, which would further dislocate native practices.
Although tannery structures have been identified in the cities of Roman Italy,
north of the Alps tanning was not an urban industry: as a newly introduced
technology, operations could be freely sited in the most appropriate location,
and the economic rationale for organization may well differ from what we
would expect on the basis of medieval and early modern practice.10 For instance,
separating operations such as flaying and initial cleansing from the actual
tanning process may have made economic sense, in order to maximize the
extraction of what we would now deem ‘waste products’. Recuperating horn,
sinew, hair, and glue in addition to the hides would affect not only the location
of the industry but also the nature of the waste left over. This is perhaps why it is
so very difficult to identify tannery locations in the archaeological record.11
The military demand for leather goods—footwear, tents, covers of all kinds,
horse gear—was vast, and the introduction of tanning was initially geared exclu-
sively to military supply.12 The imposition of taxes in kind, such as the cowhide
tax levied on the Frisians, drew raw materials from underdeveloped regions for
processing in the more stable hinterland, and, though soldiers may have been

9
Bennett (1996: 82–4).
10
On tannery structures in Roman Italy, see Leguilloux (2002).
11
Vanderhoeven and Ervynck (2007); van Driel-Murray (2011).
12
Groenman-van Waateringe (1967); van Driel-Murray (1993); Winterbottom (2009).
138 Carol van Driel-Murray

involved in preparing the hides for transport, tanning (in contrast to shoemaking)
was not carried out at the forts themselves.13 Tanning seems always to have
remained in the hands of large operators—it is a capital-intensive industry, where
investments are tied up for several years, and it also requires a well-organized
infrastructure with secure and continuous supplies of hides and other raw
materials. The organization of the process and the installations involved also
imply an increase in the scale of production and the availability of leather, thus
further stimulating its use. Presumably only major entrepreneurs would have the
managerial skills and financial resources to do this, but details of tannery organ-
ization are obscure. It may be significant that the so-called tanners’ stamps on
hides are generally in the tria nomina form, but examples are relatively infrequent,
and there is so little duplication that extreme consolidation is unlikely.14 Indeed, it
is strange that such a major industry has left so little epigraphic or iconographic
evidence, and I sometimes wonder whether major estate-owners controlled the
entire chain of activities, from slaughter to glue making, using their own person-
nel, but perhaps putting certain activities, like tanning itself, out to specialists.
Likewise, the Cohors IX Batavorum, stationed at Vindolanda, seems to have
retained administrative control of hides taken from its own food animals, putting
them out to be tanned by non-military entrepreneurs (Fig. 6.6).15 Exactly how a
new industry is established, and how the transition from military to civilian
control occurs, are difficult to track archaeologically. However, here an analogy
with the introduction of vegetable tanning in sub-Saharan Africa may be instruct-
ive. In this region, the spread of leather technology was closely associated with that
of Islam, and the first tanners were brought to Niger in the 1850s as slaves of the
sultan. As freed men, they later worked in the households of the great pre-colonial
merchants; as a result, leatherworkers and tanners still continue to be regarded as
outsiders.16 The restriction of knowledge to select groups might account for the
loss of tanning technology in the northern provinces, as Roman influence waned
in the fourth century.

CRAFTSMEN AND PRODUCERS

Archaeologically, shoemakers’ workplaces are practically invisible: the tools

are simple, and the products are sold and dispersed—all that is left is the
characteristic refuse, small snippets of leather that reveal the presence of shoe-
makers at settlements of all kinds: urban, military, and roadside settlements
(vici), and even some rural locations. The mass market in Rome may have
encouraged considerable specialization, but modern re-enactors have

13 14
Van Driel-Murray (1985, 2008). Van Driel-Murray (1977).
15
Van Driel-Murray (1993: 64–5, figs 26–7). 16
Arnould (1984: 137).
 Fashionable Footwear 139

convincingly demonstrated that a competent craftsman is quite capable of

producing the full range of footwear, from cork-soled slippers to nailed boots,
using the same basic toolkit.17 Shoemakers in the mining community at Vipasca
were expected to supply varied footwear styles as well as their own hobnails as
part of their concession, and failure to do so meant clients were free to take their
custom elsewhere.18 The small provincial town at Nijmegen (Ulpia Noviomagus)
was able to support a professional association of shoemakers, which suggests the
presence of a number of independent craftsmen, while rather enigmatic, official-
looking stamps on late-second-century AD sandals and slippers in the Rhineland
also hint at craftsmen working in some sort of corporate organization (see
Fig. 6.10, right).19 Details of organizational structures are elusive, and it is here
that one could perhaps expect the differences between the northern provinces
and the Mediterranean areas to be most marked.
This all tends to suggest a lack of product specialization, but whether
these individuals worked alone or were collected into larger workplaces must
remain open. Perhaps too much conditioned by medieval analogies and literary
conventions, we frequently envisage the shoemaker as a humble, poverty-
stricken, yet independent and argumentative figure.20 Nevertheless, the appar-
ent personal links recorded between wealthy families and freedmen involved in
shoemaking in Pizzone and Ostia are a salutary reminder that, in a society with
slaves, freedmen, and—in the northern provinces—tribal dependants, larger-
scale organizations are certainly conceivable, particularly when the technology
had been introduced to new regions in the wake of military conquest.21 In many
cases, tombstones with depictions of craftsmen in all probability commemorate
entrepreneurs controlling the activity rather than the workers themselves, with
symbolic vignettes illustrating the trade concerned.22 Bringing craftsmen together
could lead to specialization and economies of scale, but individual shoemakers
would have the advantage of flexibility, and, perhaps, innovation. Nevertheless,
the contrast between the ubiquity of shoemakers’ off-cuts, the scarcity of shoe-
makers’ tombstones, and the remarkable lack of memorials to tanners suggests
considerable caution in the interpretation of this sort of evidence.
Without documentation, degrees of dependency cannot be reconstructed
from finds alone. Moreover, shoemaking and cobbling are mobile trades relying
more on the skill of the craftsman than a complex toolkit or fixed installations,
and are therefore easily taken up by individuals seeking means of support. On the

17
As Ivor Lawton demonstrated vividly in 2007 in the archaeological park at Xanten. See also
Volken (2008, 2014). On specialization in Rome, see Leguilloux (2004: tableau IIA).
18
Domergue (1983). 19
Van Driel-Murray (1977; 2001a: 33, 38–9, figs 2–3 and 26c).
20 21
Lau (1967). De’ Spagnolis (2000: 71–3).
22
Such as the well-dressed Septimia Stratonice, who flourishes what seems to be a last or a
shoemaker’s anvil as a trade symbol (CIL 14, suppl. 4698), while the sarcophagus dedicated to
T. Flavius Trophimus by friends with links to Pizzone depicts both shoemaking and the twining
of flax or tendons. Zimmer (1982: nos 47–52); van Driel-Murray (2008: 491).
140 Carol van Driel-Murray

other hand, dependent craftsmen answering to a distant master, and the personal
networks of slaves and freedmen acting as agents abroad, would enable the rapid
dissemination of new ideas and technological information.23 Furthermore, the
widespread use of papyrus patterns in the Egyptian textile industry raises the
intriguing possibility of the more general existence of craftsmen’s workbooks
with details of manufacturing technology.24 Parallel modes of transmission,
applicable to both free and dependent labour, may be found in apprenticeships.25
Apprenticeships outside the family, coupled to the mobility of trained craftsmen
in order to reduce competition in their home town, provide a dynamic context
for transmission of ideas. There is likely to be considerable diversity in actual
practice, with operators acting at different levels of complexity. Regional markets
also provide an environment for communication and competition, with crafts-
men from different areas coming together to offer their wares for sale, interacting
with each other as well as with clients.

CONSUMERISM

All military camps on the frontiers—legionary as well as auxiliary—were self-

sufficient in footwear manufacture, and it is a small step to seeing ex-
servicemen as the initiators of this craft in the new provinces. The role of
military shoemakers may also explain the overwhelming preference for nailed
footwear in the northern provinces (Table 6.1). Though evidence is not par-
ticularly abundant, in the Mediterranean and Egypt civilian footwear is

Table 6.1. Shoe types per site

Site Carbatina Sewn Sandal Nailed Slipper Total no.
(%) (%) (%) (%) (%)

Vindolanda VI 5 5 5 83 2 333
Vindolanda 7/8 8 5 12 72 3 365
London NFW 2 11 26 59 2 129
Voorburg 6 4 15 75 0 111
Valkenburg Marktveld 1.5 1.5 13 83 1 304
Welzheim 24 6 8 57 5 177
Saalburg 18 12 10 60 ? 262

Sources: Vindolanda: van Driel-Murray (2001b: 187); Saalburg: Busch (1965); Welzheim: van Driel-Murray
(1999c); London: MacConnoran (1986); Valkenburg: Hoevenberg (1993); Voorburg: van Driel-Murray,
Pollmann, and Richter (2014). The sites are not all quantified in the same way, though it may be assumed
that each author is consistent. The totals for the Saalburg have been corrected to account for fragmentation.

23 24
Broekaert, Chapter 10, this volume. Stauffer (2008).
25
Freu, Chapter 8, this volume.
 Fashionable Footwear 141

generally soft soled, and only certain forms of heavy-duty footwear—as worn
by soldiers or labourers—had the additional hobnails. But north of the Alps,
shoes and sandals with nailed soles are the norm, even for very young children,
and remain so until the fourth century.
Civilian settlements founded in Britain soon after conquest already reveal a
rapid uptake of new footwear types, all radically different in style and tech-
nology from the footwear of the Iron Age communities and by the mid-second
century even quite remote areas were obtaining Roman-style shoes and boots.
On the continent the process is completed rather earlier, and it seems as
though it takes about three generations from the earliest introduction in
towns and military settlements to general availability in rural areas.
The population at large was being offered an unprecedented choice of
desirable novelties, a situation not repeated till the ‘consumer revolution’ of
the seventeenth and eighteenth centuries (Fig. 6.7). Kevin Greene has shown
the potential of applying the concept of ‘consumerism’ to Roman material
culture, drawing the wishes of the consumer into explanations for techno-
logical innovation in craft production.26
Roman footwear of the first three centuries AD answers all the consumerist
criteria. In the northern provinces, for the first time, a simple desire for
novelties becomes visible in the basic apparel of a wide segment of the
population. The adoption is not just a matter of improved quality: shoes are
indeed better made and longer lasting than previously, but they are above all
extremely attractive and varied (Figs 6.1 and 6.7). Hallmarks of consumerism

Fig. 6.7. A choice of fourth-century shoes from Cuijk (reproductions made by Olaf
Goubitz 2004). Image: Carol van Driel-Murray.

26
Greene (2007, 2008).
142 Carol van Driel-Murray

date
AD

250

121

18

16
225 13

8
9

200

14

7
150

100

Fig. 6.8. Dated sequence of changing sandal shapes from Voorburg (The Netherlands).
Image: Carol van Driel-Murray.
 Fashionable Footwear 143

are rapidity of change and inbuilt redundancy: and this is exactly what we see
in footwear. Change is often radical, following classic fashion cycles—high to
low, narrow to broad, rounded toes to pointed—the extremes often being
followed by the disappearance of the style and the emergence of something
completely new. Within the general trend there is also ample room for personal
expression. At Voorburg (Forum Hadriani), a small and not very successful
town in the territory of the Canannefates in Germania Inferior, most men wore
fashionable wide sandals in the 220s and 230s, but a few trend-setters went for
really exaggerated forms—forms that also appear elsewhere across the empire,
as though defining a particular lifestyle (Figs. 6.5 and 6.9). These exaggerated
soles are more simply made than earlier sandals: perhaps they were meant to be
worn only briefly during the summer, but a simple construction also allows for
more experimentation and a quicker turnover. Renewal of clothing in the
spring (as is suggested by the appearance of springtime astrological symbolism
on third-century sandal soles) may also lie behind the rather abrupt changes in
shape in this period.27
Among the issues emerging in the debate surrounding pre-industrial con-
sumerist behaviour is functional differentiation in material culture. In contrast
to both earlier and later periods, the provincial Roman population had access

Fig. 6.9. Sandal soles from Voorburg (mid-second century—c. AD 230). Image:
A. Dekker, Amsterdam Archaeological Centre, University of Amsterdam.

27
Van Driel-Murray (1999b: 134, figs 2–3); van Driel-Murray, Pollmann, and Richter (2014:
723–5, fig. II-10.20)
144 Carol van Driel-Murray

Fig. 6.10. De Meern: the boatman’s shoes. One of each pair: left, insole of the nailed
shoe; right, sandal (scale 1:3). Image: Carol van Driel-Murray.

to footwear made for different purposes, and individuals evidently possessed

more than one set of shoes.28 Unsurprisingly, the family of the commander at
Vindolanda possessed a wide range of footwear, but even the boatmen on two
capsized barges in the Rhine aspired to sandals for best wear in addition to
their closed, work shoes (Figs. 6.10–6.11).29 Simply differentiating footwear
from a number of late second- and early third-century sites on the basis of
construction method reveals the extent of consumer choice (Table 6.1). With-
in each technological group further variation in style, fastening method, and
decoration was available, allowing the consumer to vary apparel according to
the occasion. The high proportion of single-piece shoes in Germania Superior
contrasts markedly with the scarcity of the form elsewhere at this time, while
sandals increase sharply in popularity in the third century, especially in urban
centres like London. Military sites are characterized by very high (<70 per

28
Closed shoes, low-cut shoes, boots, sandals, shoes with soft soles or with hobnails, slippers,
wooden pattens, single-piece shoes (carbatinae): for the range available at a particular point in
time at a single site, see Van Driel-Murray (1999c) (Welzheim).
29
Vindolanda: Van Driel-Murray (1993: 45–7, fig. 16); Woerden: Van Driel-Murray (1996:
fig 11–12).
 Fashionable Footwear 145

Fig. 6.11. Woerden: the boatman’s shoes. One of each pair: left, nailed shoe; right,
sandal (scale 1:3). Image: Carol van Driel-Murray.

cent) percentages of closed, nailed footwear, but the extent of the accompany-
ing civilian population at some forts, like the Saalburg, is revealed by the more
even distribution across the different types.
The importance of being correctly dressed resonates with studies stressing the
role of gentility, style, and good manners in the consumer revolution of the
eighteenth century. Leaving aside the ongoing discussion around ‘Romanization’,
such factors evidently affect any assessment of the depth of Roman influence on
daily life as experienced by the provincial population. At present the provincial
Roman debate is dominated by the idea of ‘elites’ stimulating the uptake of Roman
goods, with acquisition being seen in terms of competition and political status.
But imitation is not necessarily emulation, and the varying proportions illustrate
rather a sense of community, within which personal display was negotiable.30 In
its recognition of the cumulative power of small inexpensive purchases, consumer
theory is particularly relevant for items like footwear. As Spufford shows for the
immediately pre-industrial period, such novelties gave people of modest means
the opportunity of expressing both taste and individuality within their own circles.
Even the workers at a remote upland cattle yard at Pontefract (Yorkshire) aspired
to fashionably pointed hobnailed boots, contrasting with a distinct lack of interest
in Roman pottery.31 The appearance of Roman-style footwear on rural sites

30 31
Woolf (1998: 170–1). Van Driel-Murray (2001d: pl. 9).
146 Carol van Driel-Murray

especially reveals a very real diversion of resources towards items of readily

discarded apparel, something hardly seen at all either in the Iron Age or for
much of the Middle Ages.
While it is widely recognized that clothing and accessories are culturally
constructed commodities, with complex symbolic properties, there is little
research into exactly why people purchase some items but not others. Was
choice dependent on availability (that is, production driven) or does the
consumer actually activate the producer?

CONSUMERS

In some circumstances the makers were certainly free to respond to specific

client demand. At both the Saalburg (late second or early third century) and
Cuijk (Netherlands, fourth century) a substantial proportion of the footwear
collection is composed of distinctive styles with parallels in north Germany
and Denmark, which were made for the Germanic mercenaries (and their
families) serving at the fort (Fig. 6.1, no. 19; Fig. 6.7, top right).32 Style and
design are foreign, but the materials and technology follow the established
traditions of provincial Roman footwear manufacture. Here, the desire of
clients to maintain their tribal identity stimulated local craftsmen to add to
their repertoire.33
The shoe sizes themselves can tell us something about who was making the
purchases. The size distribution of footwear in most Roman settlements reveals
fairly equal proportions of men, women, and children, a pattern comparable to
that seen in Late Medieval towns such as Lübeck or London.34 Here, the
increasing prominence of children’s shoes throughout the Middle Ages is an
indicator of prosperity levels and contrasts strongly with prehistoric and early
medieval footwear, which is largely confined to male sizes. Footwear is not
actually an essential item of clothing: until quite recent times, women and
children went barefoot in rural Ireland and Scotland, while men were shod.35
Even though the very poorest remain invisible to us, in the Roman provinces
shoes seem to have been accessible to much of the population: men, women,
and children alike. This seems to me to be a fairly obvious measure of
increasing standards of living and attention to the comfort of the entire family.
Indeed, the obviously Mediterranean fashion of wearing sandals seems to have
been seized upon by women already in the first century AD, while men begin
to wear sandals regularly only from the later second century AD onwards. In

32 33
Busch (1965); van Driel-Murray (2007). Van Driel-Murray (2009: 815–6).
34
Vons-Comis (1982: fig. 90); Grew and Neergard (1988: 104); van Driel-Murray (2001a: 360–1).
35
Lucas (1956: 353).
 Fashionable Footwear 147

footwear, we can to some extent see gendered responses to Roman consumer

goods. The feminization of consumption is a significant factor in the develop-
ment of seventeenth- and eighteenth-century production and retail distribu-
tion, and it would certainly be worth investigating other Roman consumption
patterns from this perspective.

M A R K E T ING

Did people simply buy what was on offer? Studies of early modern consumer-
ism emphasize the importance of developing communication networks and
retail outlets in making novelties accessible to a wide clientele. Margaret
Spufford, in her study The Great Reclothing of Rural England, emphasized
the role of chapmen and peddlers in the quite rapid diffusion of novelties in
apparel from urban centres to rural areas and the mechanisms she describes
for the seventeenth century would certainly be applicable to the Roman world.
The interesting thing here is that the whole system of suppliers, distributors,
and retailers mediated between producers and consumers. Knowledge of the
most recent developments in fashion and accessories was passed via personal
networks to the country towns and into the rural hinterlands. But, ultimately,
it was the consumer who actually made the choices, thereby affecting the
selection of goods on offer on the chapman’s next journey out.
For the Mediterranean area De Ligt has discussed the networks of local and
regional markets and annual festivals that brought consumers and producers
together.36 This model of periodic gatherings for commercial exchange has never
been fully exploited to explain the spread of low-value consumer goods in the
northern provinces, though it would reinforce the significance of rural cult
centres and roadside settlements in the social and economic structures of poorly
urbanized regions. In conjunction with travelling peddlers, even quite remote
areas could connect with international trends. Markets also form the arena for
negotiation between consumer and producer: leatherworkers from different
ethnic groups congregating at weekly markets in West Africa exchange ideas,
and, though distinct ethnic traditions are maintained, Frank observed that
craftsmen are prepared to make anything according to the wishes of the
customer.37 The Germanic-style shoes and fibulae manufactured on the Saalburg
are products of similar processes of negotiation, where customer demand pre-
vailed. Shoes may have been made in series and hawked around, but sandals at
any rate seem to have been made individually, and the regular finds of small

36 37
De Ligt (1993). Frank (1998: 148–9).
148 Carol van Driel-Murray

amounts of offcuts in rural sites from the mid-second century onwards may even
suggest itinerant shoemakers making to order.
However, the interpretation of rural consumer economies is not so straight-
forward. Indeed, it may be questioned whether people had free access to
consumer goods at all. Many possibly lived in various forms of dependency
and may have been constrained, receiving goods only on special occasions or
directly through their patron, and landowners may have controlled not only
travel to fairs and markets, but also the goods offered there, if these were on
their land—as was the case on Russian and Polish serf estates.38 This makes
some find contexts hard to interpret. Large numbers of worn-out shoes
dumped in wells in, for example, Welzheim or Tollgate Farm (Staffordshire)
may represent the welcome visit of an itinerant shoemaker, but could just as
well be the result of a great re-clothing of estate (or military) personnel.39
Complicating the interpretation of such unusual complexes still further is the
widespread use of footwear in ritual activities, which makes finds from wells
and water sources somewhat suspect for simple economic explanations.40

MOBILITY

The mobility of the customers themselves is an issue hardly raised in connec-

tion with consumption patterns, despite the evidence for regular private
travel.41 Indeed, the worn-out boots left by travellers could well have served
as models for local craftsmen. Occasionally individual travellers may be
suspected, and the Germanic-style footwear on military sites such as the
Saalburg and Cuijk (Netherlands) has already been mentioned. Familiarity
with Roman footwear, and some actual imports, seem to have influenced shoes
outside the frontier, in Scotland and northern Germany.42 Especially in the
frontier provinces, the influence of the highly mobile military population—
soldiers as well as their families—should not be underestimated in the spread
of new ideas. Fashion change in the sixteenth century is strongly related to
military movements, and, though the army cannot be the only instigator, the
general impression of the appearance of new stimuli at times of high military
mobility does remain. The elaborate nailing and the spread of sandal-wearing
among males at the time of the Marcomannic wars may not be coincidence,
while the profusion of new forms in the first half of the third century is

38
Domanial markets, cf. de Ligt (1993: 156–8).
39
Van Driel-Murray (1999c); Tollgate Farm: Hollis (2011); Thomas and Thomas (2012).
40 41
Van Driel-Murray (1999b). Handley (2011: 11).
42
Hald (1972: 56–7) was the first to draw attention to this, and work in progress on finds in
various northern museums confirms the extent of influence on native footwear, particularly in
decorative details.
 Fashionable Footwear 149

suggestive of social as well as political insecurity. But such renewal in fashions

can also signal times of social change and increasing competition, and the lack
of comparative evidence from the Mediterranean regions makes me wary of
such simplistic explanations. Nevertheless, personal mobility, both military
and civilian, must be a factor in the homogenization of fashion consciousness
and the development of a communal aesthetic sense.

CO NCLUSION

Many scholars dispute whether it is possible to reach the mass of the popu-
lation of the Roman Empire, and, indeed, the poorest will no doubt remain
ever invisible. But consumer theory, developed in relation to the material
legacy of the period immediately before the Industrial Revolution, does offer
the possibility of understanding the motivation of the non-elite and the
influence their individual choices brought to bear on the manufacturing
process. Items of dress such as shoes (but also fibulae, hairpins, and buckles),
widely distributed and easily discarded, offer an insight into the way quite
ordinary people (the ‘middling sort’ of consumer studies) accessed new idioms
and developed a taste for goods that made them recognizably part of the wider
world. They were prepared to divert resources from basic subsistence needs to
obtain ephemeral goods that added to the comfort of themselves and their
family as well as enhancing their feelings of self-worth. That international
styles of footwear are so widespread and are found in such large quantities
reveals, I believe, that the Roman impact on technology, craft production, and
personal bearing went far deeper than is often allowed.

REFERENCES

Arnould, E. J. (1984). ‘Marketing and Social Reproduction in Zinder, Niger Republic’,

in R. M. Netting, R. R. Wilk, and E .J. Arnould (eds), Households. Comparative and
Historical Studies of the Domestic Group. Berkeley, 130–62.
Bennett, J. M. (1996). Ale, Beer, and Brewsters in England: Women’s Work in a
Changing World 1300–1600. Oxford.
Busch, A. L. (1965). ‘Die römerzeitlichen Schuh- und Lederfunde der Kastelle Saalburg,
Zugmantel und Kleiner Feldberg’, Saalburg-Jahrbuch 22: 158–210.
Charlesworth, D., and Thornton, J. H. (1973). ‘Leather Found in Mediobogdum, the
Roman fort of Hardknott’, Britannia 4: 141–52.
De’ Spagnolis, M. (2000). La Tomba del Calzolaio: Dalla necropolis monumentale
romana di Nocera Superiore. Rome.
150 Carol van Driel-Murray
Domergue, C. (1983). La Mine antique d’Aljustrel (Portugal) et les tables de bronze de
Vipasca. Paris.
Driel-Murray, C. van (1977). ‘Stamped Leatherwork from Zwammerdam’, in B. L. van
Beek, R. W. Brandt, and W. Groenman-van Waateringe (eds), Ex Horreo, IPP
1951–1976. Amsterdam, 151–64.
Driel-Murray, C. van (1985). ‘The Production and Supply of Military Leatherwork in
the First and Second Centuries AD: A Review of the Archaeological Evidence’, in
M. C. Bishop (ed.), The Production and Distribution of Roman Military Equipment
(BAR International Series, 275). Oxford, 43–81.
Driel-Murray, C. van (1993). ‘The Leatherwork’, in R. Birley (ed.), Vindolanda Research
Reports, iii. The Early Wooden Forts. Hexham, 1–75.
Driel-Murray, C. van (1996). ‘Die Schuhe aus Schiff I und ein lederner Schildüberzug’,
in J. K. Haalebos, ‘Ein römisches Getreideschiff in Woerden’, Jahrbuch des Römisch-
Germanischen Zentralmuseums Mainz 43: 493–8.
Driel-Murray, C. van (1998). ‘Women in Forts?’, Gesellschaft Pro Vindonissa. Jahresbericht
1997: 55–61.
Driel-Murray, C. van (1999a). ‘Dead Men’s Shoes’, in W. Schlüter and R. Wiegels
(eds), Rom, Germanien und die Ausgrabungen von Kalkriese. Osnabrück, 169–89.
Driel-Murray, C. van (1999b). ‘And did those feet in ancient time . . . Feet and Shoes as
a Material Projection of the Self ’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher
(eds), TRAC 98: Proceedings of the Eighth Annual Theoretical Roman Archaeology
Conference, Leicester 1998. Oxford, 131–40.
Driel-Murray, C. van (1999c). Das Ostkastell von Welzheim, Rems-Murr-Kreis: Die
römischen Lederfunde (Forschungen und Berichte zur Vor- und Frühgeschichte in
Baden-Württemberg, 42). Stuttgart.
Driel-Murray, C. van (2001a). ‘Footwear in the North-Western Provinces of the Roman
Empire’, in O. Goubitz, W. Groenman van Waateringe, and C. van Driel-Murray,
Stepping through Time: Archaeological Footwear from Prehistoric Times until 1800.
Zwolle, 337–75.
Driel-Murray, C. van (2001b). ‘Vindolanda and the Dating of Roman Footwear’,
Britannia 32: 185–97.
Driel-Murray, C. van (2001c). ‘Technology Transfer: The Introduction and Loss of
Tanning Technology during the Roman Period’, in M. Polfer (ed.), L’Artisanat
romain: Évolutions, continuités et ruptures (Italie et provinces occidentales). Mon-
tagnac, 55–68.
Driel-Murray, C. van (2001d). ‘Hobnails’, in S. Wrathmell, Romano-British Enclosures
at Apple Tree Close, Pontefract, West Yorkshire. Wakefield, 20–23.
Driel-Murray, C. van (2007). ‘Mode in de nadagen van het Keizerrijk: De schoenen van
Cuijk’, Westerheem 56: 133–41.
Driel-Murray, C. van (2008). ‘Tanning and Leather’, in J. P. Oleson (ed.), Oxford
Handbook of Engineering and Technology in the Classical World. Oxford, 483–95.
Driel-Murray, C. van (2009). ‘Ethnic Recruitment and Military Mobility’, in A. Morillo,
N. Hanel, and E. Martín (eds), Limes XX: Roman Frontier Studies. Madrid, 813–23.
Driel-Murray, C. van (2011). ‘Are we Missing Something? The Elusive Tanneries of
the Roman Period,’ in R. Thomson and Q. Mould (eds), Leather Tanneries: The
Archaeological Evidence. London, 69–83.
 Fashionable Footwear 151
Driel-Murray, C. van, Pollmann, K., and Richter, E. (2014). ‘De leervondsten van
Voorburg-Arentsburg’, in M. Driessen and E. Besselsen (eds), Voorburg-Arentsburg.
Een Romeinse havenstad tussen Rijn en Maas (University of Amsterdam, Themata 7).
Amsterdam, 717–39.
Dunbabin, K. M. D. (1990). ‘Ipsa deae vestigia . . . Footprints Divine and Human on
Graeco-Roman Monuments’, JRA 3: 85–109.
Frank, B. E. (1998). Mande Potters and Leatherworkers: Art and Heritage in West
Africa. Washington.
Goldman, N. (1994). ‘Roman Footwear’, in J. L. Sebesta and L. Bonfante (eds), The
World of Roman Costume. Madison, 101–29.
Greene, K. (2007). ‘Late Hellenistic and Early Roman Invention and Innovation: The
Case of Lead-Glazed Pottery’, AJA 111: 653–71.
Greene, K. (2008). ‘Learning to Consume: Consumption and Consumerism in the
Roman Empire’, JRA 21: 64–82.
Grew, F., and Neergard, M. de (1988). Shoes and Pattens: Medieval Finds from
Excavations in London 2. London.
Groenman-van Waateringe, W. (1967). Romeins lederwerk uit Valkenburg Z.H.
Groningen.
Hald, M. (1972). Primitive Shoes: An Archaeological–Ethnological Study Based upon
Shoe Finds from the Jutland Peninsula. Copenhagen.
Handley, M. (2011). Dying on Foreign Shores: Travel and Mobility in the Late-Antique
West (JRA, Supplementary Series, 86). Portsmouth, RI.
Himmler, F. (2008). ‘Testing the ‘Ramshaw’ Boot—Experimental Calceology on the
March’, Journal of Roman Military Equipment Studies 16: 347–57.
Hoevenberg, J. (1993). ‘Leather Artefacts’, in R. M. Dierendonck, D. P. Hallewas, and
K. E. Waugh (eds), The Valkenburg Excavations 1985–1988. Amersfoort, 217–340.
Hollis, W. (2011). ‘Rubbish or Ritual? Roman Well Yields its Booty’, Current Archae-
ology 253: 36–41.
Lau, O. (1967). Schuster und Schusterhandwerk in de griechisch-römischen Literatur und
Kunst. Inaugural-Dissertation Rheinischen Friedrich-Wilhems-Universität, Bonn.
Leguilloux, M. (2002). ‘Techniques et équipements de la tannerie romaine: L’Example
de l’officina coriaria de Pompéi’, in F. Audoin-Rouzeau and S. Beyries (eds), Le
Travail du cuir de la préhistoire à nos jours (XXIIe rencontres internationales
d’archéologie et d’histoire d’Antibes Octobre 2001). Antibes, 267–81.
Leguilloux, M. (2004). Le Cuir et la pelleterie à l’époque romaine. Paris.
Leguilloux, M. (2006). Les Objects en cuir de Didymoi, praesidium de la route carava-
nière Coptos-Bérénice. Le Caire.
Ligt, L. de (1993). Fairs and Markets in the Roman Empire. Amsterdam.
Lucas, A. T. (1956). ‘Footwear in Ireland’, County Louth Archaeological Journal 13: 309–94.
MacConnoran, P. (1986). ‘Footwear’, in L. Millar, J. Schofield, and M. Rhodes (eds),
The Roman Quay at St Magnus House, London: Excavations at New Fresh Wharf,
Lower Thames Street, London 1974–78 (London & Middlesex Archaeological Soci-
ety, Special Paper, no. 8). London, 218–26.
Mould, Q. (1990). ‘The Leather Objects’, in S. Wrathmell and A. Nicholson (eds),
Dalton Parlours: Iron Age Settlement and Roman Villa (Yorkshire Archaeology, 3).
Wakefield, 231–5.
152 Carol van Driel-Murray
Overton, H., Whittle, J., Dean, D., and Hann, A. (2004). Production and Consumption
in English Households 1600–1750. London.
Spufford, M. (1984). The Great Reclothing of Rural England: Petty Chapmen and their
Wares in the Seventeenth Century. London.
Stauffer, A. (2008). Antike Musterblätter: Wirkkartons aus dem spätantiken und
frühbyzantinschen Ägypten. Wiesbaden.
Thomas, D., and Thomas, M. (2012). ‘Roman Shoes Report’, Stoke-on-Trent Museum,
Archaeological Society <http://www.stokearchaeologysociety.org.uk/html/tgfm_
shoes.htm>.
Vanderhoeven, A., and Ervynck A. (2007). ‘Not in my Backyard? The Industry of
Secondary Animal Products within the Roman Civitas Capital of Tongeren, Belgium’,
in R. Hingley and S. Willis (eds), Roman Finds: Context and Theory: Proceedings of a
Conference Held at the University of Durham 2002. Oxford, 156–75.
Volken, M. (2008). ‘Making the Ramshaw Boot, an Exercise in Experimental Archaeology’,
Journal of Roman Military Equipment Studies 16: 359–66.
Volken, M. (2014). Archaeological Footwear: Development of Shoe Patterns and Styles
from Prehistory till the 1600s. Zwolle.
Vons-Comis, S. Y. (1982). ‘Das Leder von Lübeck, Grabung Heiligen-Geist-Hospital,
Koberg 9–11’, Lübecker Schriften zur Archäologie und Kulturgeschichte 6: 239–50.
Weatherill, L. (1996). Consumer Behaviour and Material Culture in Britain 1660–1760.
2nd edn. London.
Winterbottom, S. (2009). ‘The Leather and Other Organic Artefacts’, in C. Howard-
Davis (ed.), The Carlisle Millennium Project: Excavations in Carlisle 1998–2001.
Lancaster, 817–41.
Woolf, G. (1998). Becoming Roman: The Origins of Provincial Civilization in Gaul.
Cambridge.
Zimmer, G. (1982). Römische Berufsdarstellungen. Berlin.
 7

Contextualizing the Operational

Sequence: Pompeian Bakeries as a
Case Study
Nicolas Monteix

One way of approaching the material remains of Roman workshops is to

investigate how they were designed specifically to accommodate the activity
that took place within them. Such an approach may be based on the analysis,
from a technological point of view, of a specific kind of workshop whose
elements are all considered in their spatial context. In terms of theoretical
background, such an approach relies on the notion of the chaîne opératoire
(operational sequence), which was first defined by A. Leroi-Gourhan, was
gradually refined by his successors, and was then applied to archaeology.1 The
idea is that a close study of the different elements that characterize the
operational sequence of a workshop, including materials, tools, gesture, know-
ledge, energy, and actors, combined with an analysis of their interaction, leads
to further understanding of social and economic issues, particularly if one tries
to take a more systemic point of view aimed at understanding the système
technique or the complexe techno-économique.2
In order to test how this approach works in practice, this chapter presents a
case study of the bakeries of Pompeii: although they are described in every book
on Pompeii, previous studies have always portrayed them as an unsatisfactorily
random assemblage of equipment. Early descriptions, written by Sir William
Gell and François Mazois, include precious observations and interpretations of
the bakeries but remain far too general to be completely useful.3 August Mau’s

1
Leroi-Gourhan (1964: 26–64); Balfet (1991); Creswell (2003); Pernot (2006).
2
First defined by Gille (1978: 19), as ‘l’ensemble [en équilibre] de[s] cohérences [techniques]
aux différents niveaux de toutes les structures de tous les ensembles de toutes les filières’, the
système technique was then questioned by Lemonnier (1983). Garçon (1997: 24–5) defined the
complexe techno-économique on a smaller scale.
3
Gell and Gandy (1817–19: 189–91, pl. XXXVII); Mazois (1824: 56–61, pls XVI–XVII).
154 Nicolas Monteix

description of bakeries does follow the operational sequence from milling to

baking, but lacks precision, owing to his high degree of generalization.4 Betty Jo
Mayeske failed to go much further than presenting a mere update of these data
based on the excavation reports published since Mau’s time.5 Since then,
Pompeian bakeries have appeared only as a starting point—even if often
implicitly—for larger syntheses about ancient food technology.6
Consequently, there is still much to be learned from Pompeian bakeries, by
studying their equipment closely and trying to interconnect the individual
installations through an analysis of their spatial disposition. It is the aim of this
chapter to discuss the different sets of equipment found in Pompeian bakeries
in order to reconstruct the operational sequence and then evaluate the ration-
alization of their organization. Two preliminary points need to be made. First
of all, as far as the number of bakeries within the city walls is concerned, some
forty-one bread ovens have been recognized to date (Fig. 7.1). These include
those with a baking chamber with an internal diameter of 1 metre or more,
regardless of the destination of the bread—commercial or otherwise.7 The
second point is that not all the thirty-nine identified bakeries were in oper-
ation in AD 79: some were partially or totally inactive, either because of damage
that occurred shortly before the eruption, during the ‘ongoing seismic activity’
in the region, or due to some cessation of activity.8

RECONSTRUCTING THE OPERATIONAL S EQUENCE:

FROM GRAIN TO BREAD

The main outlines of bread-making have not changed radically since the
Neolithic, except for the slow—and even nowadays still incomplete—
generalization of leavening: (1) flour, usually obtained by milling grain, is
(2) mixed with water—and possibly yeast—in order to obtain dough, which,
often after some time of fermentation, has (3) to be baked in a purpose-built

4 5
Mau (1899: 380–4). Mayeske (1972).
6
e.g. Curtis (2001: 341–7, 358–70; 2008: 373–9); Thurmond (2006: 14–72).
7
Another one could be added (I 8, 15–16), although the oven is only 78 cm wide: situated in
the back room of a food outlet (caupona), this oven is embedded in a smoking shed most
probably associated with cheese production. The presence of a human-driven mill and the whole
commercial character of this installation make bread a minor but real production in this food
workshop and retail shop. In the present chapter, this one will be ignored, as none of the other
ovens studied is smaller than 1.25 metres in diameter.
8
For ‘ongoing seismic activity’, see Allison (1995). The bakery installed in VII 1, 25.46–7 was
partially dismantled in AD 79, as shown by the removal of the three mills, whose surrounding
pavement has been partly covered by a wall—constructed and ?—painted during the first years of
the AD 70s; cf. Esposito (2009: 57, 77–80). During the last years of Pompeii, though the oven
seems to have been functioning.
 5
IV,
12
4 III,
IV,
11
3 III,
IV,
10
2 III,
IV,
9
1 III,
IV,
5 8 7
V, III, III,
4
V, 5
4 II,
,1
3 IX
15 V,
VI, ,1
0 6
IX III,
11
VI, 5 4
,9 III, II,
2 IX
9 V,
VI, 16
VI, ,8 4
IX III, 3
II,
,7 1
VI V, 3
III, 2
,5 II,
,5 IX
VI 4 2
I, 1 III, 1
2 V II,
VI, 1
13 ,4 ,6 III, 3
VI, IX IX I, 1
,1
VI ,1
3
12 IX 2
VI, I, 1
,3 7
,3 IX ,1
2 II,
10 VII IX 11
VI, I,
9
,7 11 II,
I, 8 IX I X, 9
V I,
,2 4
IX I, 1
6 I, 8
17 VI, VII,
2
15
VI, 4 I,
3 VII, 7
VI, I,
,1 6 8
IX I, 1 II,
,1 I, 6
VII 0
,5 VII,
12 I, 2
VII 17
4 I,
VI, 1
8 I, 2
11 4 I, 1
N I,
6 VII, I,
VI I, 1
0 22
10 I,
VII,
4
VII,
9 VII, 1
,1
6 I, 4 ,3
VII I 9
VII 13 I, 1
VII,
0 50 100 200 m
I, 7
VII I, 2
15

I, 5
VII
I,
VI

7 I, 5
VII, 6
VIII,

1
I, 3

I,
damaged mills in AD 79
VII
16
VII,
levelled mills in AD 79

5 manual mills

0 I, 2
VII
I, 1
5 ‘Pompeian’ mills VII

Bakery with mills and shop

Bakery and shop, no mills

Bakery with mills

Bakery without mills or shop

Fig. 7.1. Location of the bakeries in Pompeii. Map: Nicolas Monteix.

156 Nicolas Monteix

oven. What has changed from one culture to another are the tools used to
perform these broad operations and their subtle subdivisions.

Milling

In the Roman world, as far as we actually know from archaeological remains,

milling was performed independently of baking only in cases where hydraulic
energy was used, despite the etymology of the Latin term for baker—pistor
comes from pinsere, pounding. In Italy outside Rome, contrary to what Pliny
suggests, known water-mills are relatively few, and all are later than the
destruction of Pompeii and Herculaneum.9 Around Pompeii, no water-mill
has been identified, and we have to assume, until further discovery, that the
grain was milled in villas and urban bakeries. Yet the primary preparation of
grain—threshing and winnowing—was completed before its arrival in the
bakery, as is suggested by the lack of any traces of chaff, not even phytoliths,
in excavated contexts in some recently studied workshops.10 That the grain
arrived clean does not mean that the first operation was the milling: prepar-
ation might have occurred beforehand. According to Pliny, tempering the grain
in salt water before milling would have given whiter flour.11 It is not clear why
salt should be added to water, but soaking the grain, as is still done nowadays,
enables better separation between the bran—which has gained elasticity—and
the kernel, and thus produces, after milling and sifting, flour that has a whiter
colour than that produced with a non-tempered grain which contains many
small bran fragments.
Until now, only three or four examples of tempering facilities have been
identified in Pompeian bakeries (Fig. 7.2). The first one, initially interpreted
in this way by Fiorelli, is located in bakery IX 3, 19–20, where recent cleaning
has improved the understanding of its functioning.12 In room 5, where the
pavement sloped slightly towards a cistern, a symmetrical structure was built:
two containers—a reused millstone set upside down and a dolium—were placed
in the south-west and north-west corners. Both have a hole for emptying.
Corresponding to each of these basins, a row of tiles was inserted obliquely into
the pavement, most probably to support some wooden element. A connection to
the piped water network shows that at least the reused millstone was filled up with

9
Plin., HN 18.97. On the few examples of known water-mills in Italy, cf. Brun (2007); on
water-mills in general, see Wikander (2008).
10
The carpological study has been conducted by M. Derreumaux (Centre de recherches arché-
ologiques de la vallée de l’Oise) and V. Matterne (CNRS – UMR 7209 Muséum National d’Histoire
Naturelle). Phytolith analysis was conducted by C. Hamon (CNRS – UMR 7041 ArScAn).
11
Plin., HN 18.87: Nam quae sicca moluntur, plus farinae reddunt, quae salsa aqua sparsa,
candidiorem medullam, uerum plus retinent in furfure.
12
Fiorelli (1873: 53–4).
 Pompeian Bakeries as a Case Study 157

0 1 2 3 4 5 6 7 8 9 10 m

Water distribution network Restored water distribution network

Fig. 7.2. Tempering structure in bakery IX 3, 19–20. Image: Nicolas Monteix.

water continuously. In bakery V 3, 8, low walls were built on the edge of the
impluvium, which itself was divided into two parts with a wall that could let
water pass through, either to the cistern mouth or to a channel that dis-
charged into the street in front of the building. A similar construction might
have occurred in bakery VII 1, 36–7, albeit without a dividing wall in the
impluvium. A small basin, seen in the impluvium of bakery VII 2, 3.6–7,
might have been used for the same purpose. Although the water supply was
158 Nicolas Monteix

different in volume, both structures allowed grain to be tempered for a few

hours before the next step, milling.
Some information about the operational sequence might be derived from
the basalt pavement that was generally set around the mills in order to prevent
the floor being worn by the repeated circular movement of the working
animals. This pavement does not appear in all cases: one bakery in Pompeii
(I 12, 1–2) has a mill without a basalt pavement, as did both bakeries in
Herculaneum. The absence of this pavement, especially in Herculaneum, is
difficult to explain. However, observations made in several bakeries give an
indication of the construction of this pavement. In some cases, a masonry base
is set in a shallow cylindrical foundation. The basalt blocks, whose thickness
might vary greatly, from 15 to 30 centimetres, are placed around it, in a pit
whose lower part is filled with plaster fragments, most probably in order to
facilitate the setting of the blocks. If a bakery was extended with further mills,
strikingly, the paving blocks already installed—which might be in common
with the new ones—were partly dismantled in a way that allowed the new mill
to have a new pavement laid completely around it (Fig. 7.3). Once the paving
had been installed, the meta—the bed stone—was either just laid directly on it
or inserted in the circular pit. A small wall was then built around it and was

a b

50 cm
50 cm

a b

0 1 2 3 4 5 6 7 8 9 10 m

Fig. 7.3. Extension of the basalt pavement in bakery I 12, 1-2. The filling for the
second pavement, seen under excavation (a), cuts the one used for the first pavement
(b) that abutted the blocks. As the plan shows, the first pavement has been partly
dismantled to install the second. Image: Nicolas Monteix.
 Pompeian Bakeries as a Case Study 159

most likely rebuilt each time the meta was replaced when it had become too
worn to be used. Once the catillus—the running stone—had also been set up, a
lead sheet was placed at the joint between the meta and the small wall around
it to facilitate flour recovery.
It is not necessary to discuss here, at length, the general functioning of the
most typical tool—at least if one looks at ancient representations—used in
mill-bakeries, the animal-driven ‘Pompeian’ mill.13 Yet, as has already been
mentioned in the introduction to this chapter, it has always been assumed that
all bakeries and thus their components were functioning in AD 79. A careful
study of the state of wear of the mill elements shows a different picture.14 In
the thirty-nine bakeries, ninety-one animal-driven mill pedestals have been
identified: only twenty-eight (31 per cent) of these were functional at the
moment of the eruption, of which only three were almost new, not worn.15
Only further research could indicate—and not without difficulty—whether
this disrupted image is to be linked with ongoing earthquakes or if it indicates
some other changes.
Once the grain had been ground, the result had to be cleaned, in order to
separate the flour from the undesired parts it contained—mostly bran and germ.
The use of bolters is attested by both literary and iconographic sources.16
According to Pliny, three successive siftings, corresponding to three qualities of
flour, would have helped to increase grain output.17 Although none of these tools
has been observed or preserved, some recurring structures in Pompeian bakeries
can be linked with this operation: in some bakeries, there are L-shaped walls
about 80 centimetres high; these are often in the milling-room and have generally
been built against a load-bearing wall. As such a structure seems unfit for storage,
it is likely that they were used for sifting, and were once covered by some wooden
installation.18

13
For a general account, see Moritz (1958: 74–90) and now D. Peacock (2013: 77–97). The
provenience of mills has been studied by D. P. S. Peacock (1989), with updates in Buffone et al.
(1999).
14
The study of the mills has been conducted by S. Longepierre (Post-doc. CNRS – UMR 5140).
15
These statistics have been partly revised to include information from the unpublished and
published excavation reports: a few mills appear non-functional nowadays owing to their
progressive destruction since their unearthing; in some cases, the reports mention that mills
suffered damage during the eruption and in post-eruptive disruption. For instance, one of the
catilli from bakery I 12, 1–2 was discovered 2.5 m above the ancient ground level. Cf. Giornale
degli Scavi di Pompei, 13–19 February 1961.
16
Cat., Agr. 76.3; Plin., HN 18.108. Bolters are carved either on the Bologna relief, on
P. Nonius Zethus’s urn holder, or on a fragmented relief from Ostia. Cf. Zimmer (1982: 114).
Only on the latter might it be a sieve rather than a pre-grinding bolter.
17
Plin., HN 18.89–90.
18
The excavation of this structure in bakery I 12, 1–2 did not show any beaten soil inside the
double L-shape. Only the central part, which permitted access to it, seemed to reveal signs of a
prolonged stay; hence some planks might have prevented bakers from treading on it.
160 Nicolas Monteix

Kneading and Forming Loaves

The next step is technically the first in bread-making—what happens prior to

this can be seen as ‘milling’, and could potentially have been separated from
bread-making, occurring in a different workshop. The procedure is to mix
flour with water, and also yeast if the bread is to leaven, which, according to
carbonized material remains of loaves, was mostly the case in first-century AD
Pompeii.19 Traditionally—and until most recent times—kneading was per-
formed by hand, in a wooden chest.20 But from at least the very late Republic,
as the frieze on the tomb of M. Vergilius Eurysaces shows, an alternative to
hand kneading had been developed.21 This alternative, already identified by
August Mau, was a kneading machine, consisting of a hollowed cylindrical
stone fixed firmly in the pavement. Generally basalt was used, but occasionally
travertine was used as well. At Pompeii, most kneading machines were made
of local basalt, and not all of them were originally made as kneading machines:
three were reused metae. The main active part was a vertical shaft, fixed to the
stone through an iron pivot hole at the bottom of the vat and kept in a vertical
position thanks to its insertion in a horizontal beam that was fixed in the
surrounding walls (Fig. 7.4).22 At the lower end of the shaft, just above the
pivot, an iron blade, turned slightly upwards at its extremities, was fixed to it.
Further, on the inner side of the vat, two to six holes were carved; they allowed
wooden sticks to be inserted inside the vat at various heights. A horizontal arm
higher up the shaft made it possible for it to be turned around; the rotary
movement made the blade rotate and mix all the ingredients that were partly
blocked by the wooden sticks and thus better kneaded together.23 From the
lack of basalt pavement around the kneading machines, we might deduce that in
Pompeii they functioned with human energy, although, as discussed in the
section on milling, this argument has to be employed with caution.24 Throughout

19
None of the loaves discovered in Pompeii or Herculaneum has the flat shape of unleavened
bread; cf. Borgongino (2006: cat. nos 457–83). Most of the Pompeian loaves that were preserved
come from bakery VII 1, 36–7.
20
For hand-kneading in Roman times, see Wilson and Schörle (2009). This implement is
now attested in Pompeii in bakery VII 2, 3.6–7; see Monteix et al. (2014: 20–1). Such chests are
still depicted in Diderot and D’Alembert’s Encyclopédie. Kneading machines began to replace
them again at the beginning of the twentieth century.
21
See Ciancio Rossetto (1973); Wilson (2008a: 358).
22
In his first study of the feature, Mau (1886: 46) clearly mentions the horizontal beam, while
in his second less accurate description (Mau 1899: 384, fig. 214 p. 383), which is more often used
as a reference, he does not.
23
Contrary to what has been written by Mau (1899: 384) and his followers, nothing permits
the restoration of arms projecting from the vertical shaft, notably because the sockets fixed into
the side of the vat would have forbidden their rotation, as careful and exhaustive measurements
have demonstrated. Cf. e.g. Mayeske (1972: 22); Strocka (1991: 61); Thurmond (2006: 66).
24
Not only does the frieze on the tomb of M. Vergilius Eurysaces show a kneading machine
functioning with a donkey or a horse, but also bakeries in Ostia have a basalt pavement around
the kneading machines. See Bakker (1999).
 Pompeian Bakeries as a Case Study 161
25 cm

3 2

25 cm 0
7
25 cm
6

8
5

1
0 1m

1 Mixing basin (basalt) 5 Pivot and rivets 4

8
2 Vertical shaft 6 Iron blade
3 Horizontal beam 7 Iron collar 1
4 Iron pivot hole 8 Wooden teeth

Fig. 7.4. Cross-section and exploded isometric view of the kneading machine in the
House of the Chaste Lovers (IX 12, 6). Image: Nicolas Monteix.

the city, some twenty-three stone tubs once used as kneading machines have
been identified.25 However, in AD 79, only twenty bakeries might have had such a
tool, and in two of these, the only evidence found consists of holes in the walls
used to insert the horizontal beam that fixed the shaft.26
As soon as the different ingredients have been mixed and kneaded to form a
dough, the rising begins.27 The importance of this phase is inversely propor-
tional to the evidence related to it: there is no mention of it in ancient
literature, no detailed iconographic representation has been found, and it is
only very briefly alluded to in modern studies. This absence can be explained
by the numerous details given in ancient texts about yeast or by the difficulty
in representing—on stone or in painting—a rising loaf with enough details in
order to distinguish it from a baked one. Nevertheless, some features of
Pompeian bakeries must be linked with this phase. In seven bakeries terracotta
tubs have been found that are characterized by their wide opening (Fig. 7.5).
Depending on their size, those tubs are either embedded in masonry or fixed
on a reused catillus. Mostly, they are situated in the immediate vicinity of the
kneading machine or the oven. Two striking features need to be emphasized.

25
Two of them are to be found reused in bakeries as hydraulic implements, two others
outside and not necessarily in the vicinity of any bakery as kerb- or stepping-stones. Even noting
that ‘such basins are easily broken or removed’, Mayeske (1972: 169) counted only eight of them.
26
Seven bakeries still have their kneading machine in the place where it was used. In eleven it
has been moved either within the room or in other rooms.
27
In fact, there are two phenomena occurring at the same time: modification of the gluten
structure and carbon dioxide production. For detailed explanation of these changes, see Cauvain
(2007: 21–5).
162 Nicolas Monteix

IX 5, 4 V 3, 8

IX 1, 3.33 IX 1, 3.33

VI 11, 8–10 VI 11, 8–10

IX 3, 19–20 IX 3, 19–20

VI 3, 3.27–8 Inv. No. 52548

0 10 20 30 40 50 60 70 80 90 100 cm

Fig. 7.5. Vessels used for the first rising of the dough. The two on the first row had a
lead sheet as a bottom (scale: 1/10). Image: Nicolas Monteix.
 Pompeian Bakeries as a Case Study 163

First, they are to be found only in bakeries that are provided with a kneading
machine, though not all of these have these rising tubs. Secondly, these
recipients withstand any typological grouping: they do not seem to have
been produced for this specific purpose. Furthermore, two examples, one
being set in a disused catillus (Fig. 7.6), are in fact dolia truncated before firing
and hence exhibit a lead sheet as a bottom part.28 All of this suggests do-it-
yourself rather than standardized practices. These vessels, combining a wide
opening with shallow depth, and installed near the kneading machine, can be
understood to serve for the first rising of the whole batch of dough, which was
then remixed—by ‘punching it down’ or ‘folding it’ in modern bakery terms.29
It is after this first rising—when and if it has occurred—that loaves are
formed. In a rare correspondence with iconographic representations, this
operation was performed on tables whose feet are generally preserved in

0 1m

1 First rising vessel 2 Reused running stone

3 Lead sheet 4 Small wall

Fig. 7.6. Cross section of a runner stone (catillus) reused to support a first rising vessel
in bakery IX 5, 4. The fitting vessel was preserved on site in the same room. Image:
Nicolas Monteix.

28
The second example lies in bakery V 3, 8, embedded in a masonry structure in a small space
between the kneading machine room and the oven. During the cleaning of this room, the imprint
of a second tub was seen (Monteix et al. 2013: }7, fig. 3).
29
The main problem with this modern terminology is that it assigns a gesture to the process,
while even now both techniques coexist. The effect of this first rising followed by a remixing is to
restart yeast development.
164 Nicolas Monteix

Pompeian bakeries. Though their dimensions differ from one bakery to an-
other, their general shape is always the same: they are rectangular and have
usually been built in masonry—though some blocks of local limestone might
have been used. They are 60–70 centimetres high and were most likely covered
by wooden planks. Typically placed in the centre of the room, they allowed
workers to form loaves on both sides of the table. The dough had to be divided
into lumps; either these would have been weighed, or their weight estimated by
eye. The forming in itself must be inferred from the result: fourteen loaves,
partially or completely preserved, have been found in Pompeii and Hercula-
neum and can still be studied.30 They are all pre-divided into six or eight parts
on the upper face, and into two parts on their sides. Such a result could have
been obtained by scoring the loaves either with a knife, after a time of fermen-
tation, just before baking them in the oven, or—hypothetically—with a string
that was left around the loaf while it was leavening. In any case, none of the
Pompeian—or Herculanean—loaves of bread preserved after the eruption
shows any sign of moulding.31 Once the loaves had been shaped, with or
without the subdivision, the dough had to finish its leavening: the loaves were
left on shelves, which are almost always visible in the preparation room when
the wall preservation is sufficient. At least one example, known in a domestic
context (house VIII 2, 34), attests that loaves might have been covered by many
pieces of cloth—more than fifty in this case—in order to preserve humidity
while rising.32 That this use of textile covers was also practised in commercial
productive space remains conjectural. Nevertheless, it should be emphasized
that rising works better in a relatively high ambient temperature (35–40 C).

Baking

Once loaves have leavened sufficiently, the last step in bread-making is baking.
Although until now we have only considered leavened bread, the consumption
of a second type—called clibanus—is at least attested by an electoral notice.33
If this Latin term used in Pompeii has the same significance as in literary
sources, it should refer to flat bread, pancake-baked after slapping it onto the
internal side of a terracotta or metal oven in which a fire burns.34 Yet, while

30
Of eighty-one loaves discovered in VII 1, 36–7, only six are still preserved. The other eight
come from Herculaneum. Cf. Borgongino (2006: 140–5).
31
The alleged moulds supposedly found in bakery Or. II, 8 in Herculaneum, which have
always been used since then as an example to demonstrate that bread was moulded, were
discovered not in this workshop but in the flat above it, which lacks any connection with
productive spaces. See Monteix (2010: 151–2); contra Maiuri (1958: 457–8).
32
Médard, Borgard, and Moulhérat (2011: 84–6).
33
CIL 4.677: Trebium aed(ilium) o(ro) u(os) f(aciatis) | clibanari rog(ant).
34
André (1981: 67); Thurmond (2006: 64–5).
 Pompeian Bakeries as a Case Study 165

such earthenware has been found in southern France, it has not been dis-
covered in the AD 79 levels of Pompeii or Herculaneum, nor has any unleav-
ened bread been preserved.35 Hence, this chapter will consider only the
‘traditional’ bread oven built in masonry.
The principal parts of an oven consist of its base, which is generally square
and 70–80 centimetres high, and its baking chamber, which, besides some rare
exceptions, always has a domed shape (Fig. 7.7). The base is always deeper
than the chamber in order to create a shelf for the oven at the front. The way in
which those minimal though essential elements are built helps in understanding
how such ovens worked. The dome is built directly on the base. The oven floor,
where loaves are laid to be baked, is made of tiles—or much more rarely of square
bricks. These are set on a layer of sand, 10–40 centimetres thick, which covers the
masonry base of the oven.36 In Pompeii and Herculaneum, the first row of the
floor is always made of lava stones, the function of which has long been
questioned. While Luigi Fulvio believed that this material was chosen in order
to protect the oven against damage from the bread shovel, its actual purpose
might have been related to the characteristics of the lava: basalt has high thermal
effusivity, owing to both its high conductivity and its high volumetric heat
capacity.37 Thus, it can stock a huge amount of heat internally and then release
it slowly, either by radiative or by convective heat transfer. The sand underneath
it, on the contrary, particularly when dry, has low heat conductivity. Its presence
under the tiles or bricks thus avoids heat loss through the base and prevents heat
from being concentrated under the loaves.
Besides these general characteristics, which are common to all Pompeian
ovens, some optional features might have been added. A diaphragm wall,
which is characterized by an arch in its front part, might be built against the
shelf, thus creating a space 30–40 centimetres wide before the oven mouth. In
the upper part of this, a chimney might be inserted, designed with various
techniques, from reused amphorae to built-in masonry flue. This would allow
smoke to be vented out of the workshop.38 Within the space created, mostly

35
Barberan et al. (2006). The vessels found there differ extensively from Italian examples such
as those published by Cubberley, Lloyd, and Roberts (1988). Given the similarity between these
ovens and dolia when fragmented, they may yet still be found in Italy. A Latin label being needed,
I would use clibanus for tabouna-like ovens, while I would keep testa for baking pans. Clibani
would have been in use in Pompeii, as suggested by CIL 4.677; see also Monteix et al. (2015 }3g).
36
10 cm is the height of this sand layer as reconstructed in ovens that have lost their tiled floor
or in those where it has been possible to measure where it was preserved. A thickness of 40
centimetres was observed in Herculaneum, in Or. II, 1a, according to the excavation reports
(Giornale degli Scavi di Ercolano, 11 July 1935: ‘11 luglio. Un ambiente non ancora numerato
perché in corso di sterro, tiene il pavimento di grosse tegole di terracotta che poggiano sopra uno
spessore di metri 0.40 di sabbia marina . . .’).
37
Fulvio (1879: 285).
38
The height of these flues probably helped post-eruption looters to locate bakeries in order
to salvage items—mostly metallic ones, notably lead-boilers, but also mills. See n. 15.
166 Nicolas Monteix

1 6 6

7 3

2
2

5 2 3

3
9

0 1 2 3 4m

Lead boiler (spoiled,

1 Diaphragm wall 4 7 Mouth
dimensions unknown)
2 Landing 5 Water drawing hole 8 Bake chamber

3 Loaves pass-through 6 Chimney 9 Sole (tiles on sand layer)

Fig. 7.7. Oven in bakery V 4, 1. Image: Nicolas Monteix, based on survey and drawing
by François Fouriaux and Sandrine Mencarelli.

available on the two sides of the mouth, two implements might be added. First
is the lead-boiler, embedded between the diaphragm wall and the oven,
slightly above the landing, so that fuel could be inserted under it (Fig. 7.8).
Of the twenty-four ovens that had a diaphragm wall (or where one might
be reconstructed), twenty-two had a boiler, of which only two are still
 Pompeian Bakeries as a Case Study 167

0 1 2m

Fig. 7.8. Lead boiler embedded in the oven of bakery VII 12, 11 and its heating
surface. Image: Nicolas Monteix.

preserved.39 While their volume shows considerable variation, their function-

ing remains identical: one lead pipe fills the boiler, while a second drains it;
both are inserted into one of the sides and pass through the diaphragm wall. It
cannot be ascertained that these boilers were connected directly to the water
distribution network. Yet on the side or the front of the diaphragm wall, there
is always an inlet that received water and let it go to a drain.40 As for the use of
this warmed water, one should think of kneading: water that was warm, but
not too hot, could be of use in dough preparation.41 The second feature generally
linked to the diaphragm wall, first defined by August Mau, is a pass-through.
This is a window-like opening through which loaves could be passed directly
from the preparation room to the oven shelf.42 Twenty-three of these pass-

39
In I 12, 1–2 (volume: 109 L) and VII 12, 11 (volume: 26 L).
40
An intermediate vat might be used to stock warm water, as in I 12, 1–2. Most boilers seem
to empty directly in those vats that are directly linked to a sewer. Warm—if not hot—water
passed through, as limited calcareous deposits indicate in IX 12, 6. The form of these vats differs
greatly from one bakery to another reused Dressel 20 being quite frequent.
41
On temperatures in modern bakeries, see Cauvain (2007: 42). Thurmond (2006: 71) also
suggests that sprinkling water on loaves would have helped them rising while baking. If so, it
would not be the main purpose of this feature. Such a practice would have been counter-
productive though: opening the oven would generate a loss of temperature.
42
Mau (1899: 383).
168 Nicolas Monteix

throughs have been observed in Pompeii, some also in bakeries that do not have a
diaphragm. In these cases, it is an opening in the dividing wall between the
panificium and the oven. It is unlikely that these pass-throughs were also used to
bring baked bread back into the preparation room.
With or without the diaphragm wall, the functioning of the oven remains
the same: it is first heated, while its mouth remains opened in order for both
the basalt ring and the sand layer to reach the right temperature, which should
be slightly above 250 C. Once this has been done, all the fuel is removed (and
possibly stocked—for instance, in basalt recipient—or moved under a boiler),
and the loaves are placed on the oven floor. The iron door is shut, and the
baking starts. While this reconstruction is mainly based on an analysis of oven
architecture, it is confirmed by the situation in bakery VII 1, 36–7: when this
building was unearthed in 1862, eighty-one loaves were discovered in the
oven. Their dimensions—they had a diameter of 19–23 centimetres, about 20
centimetres on average—exclude the possibility that there was any remnant
fuel in the oven during the baking itself.43
Excavations especially in bakery I 12, 1–2 and IX 3, 19–20 in Pompeii led to
the discovery of one component of the fuel used for baking: some 99 per cent
of vegetal macro-remains found in these workshops consist of olive stones,
while beech is predominant among identified wood species (58 per cent).44
The choice of olive pits might be explained by their calorific power, which is
12 per cent higher than the wood species more commonly used in first-century
45
AD Pompeii. It must be emphasized that, even if the proportions of different
fuels used at the same time will never be quantified, the over-representation of
olive stones does represent a technical choice that must be linked with the
baking process: published domestic contexts show a total lack of such re-
mains.46 It remains quite difficult to estimate the length of the pre-heating

43
About this bakery, and the loaves discovered in it, see Giornale degli Scavi di Pompei,
9 August 1862; Fiorelli (1873: 16–17; 1875: 171–2).
44
The carpological identification was conducted by M. Derreumaux and V. Matterne (see n.
10). Charcoal analysis and determination were run by S. Coubray (Institut National d’Archéo-
logie Préventive). For complementary details on these results, see Coubray, Matterne, and
Monteix (forthcoming).
45
According to the sample from the Casa delle Vestali, beech (47%), Corylaceae (23%), and oak
(10%) are the most predominant species used in this domestic context in the first century AD; see
Veal and Thompson (2008). They have a higher heating value (HHV) respectively of 5.14 kWh/kg,
4.97 kWh/kg, and 5.04 kWh/kg (data <http://www.onf.fr> (accessed 10 July 2015)), while the HHV
for olive stones has been estimated recently as 5.69 kWh/kg (20,500 J/kg) (data <http://www.
afidoltek.org/index.php/Valorisation_énergétique_des_grignons> (accessed 10 July 2015)).
46
No olive stone has been found in the six samples studied in the Casa delle Vestali; see Veal
and Thompson (2008: 294). Olive stones, complete and fragmentary, discovered in several
bakeries contradict the explanation used by the two authors to justify their lack in domestic
context: ‘if used as kindling, olive would have been the most likely to burn completely due to its
oily character’ (2008: 294).
 Pompeian Bakeries as a Case Study 169

phase and of the baking procedure, given the multiple factors that might have
interfered, notably the quantity of fuel and the size of the oven.47
After baking, the loaves were ready to be sold. This final operation could
either take place in front of the bakery, if it had a shop—as nineteen bakeries
did—or through an intermediate retailer.

THE OPERATIONAL S EQUENCE I N I TS TECHNICAL,

SPATIAL, AND E CONOMICAL CONTEXT

Before I move on to discuss the spatial arrangement of Pompeian bakeries, a few

points concerning the operational sequence have to be discussed. The three main
steps described—grinding, kneading, and forming loaves, baking—and their
various subdivisions do reflect the technological knowledge in bread-making
used in the Pompeian bakeries. Furthermore, those observations remain fully
valid only for the third quarter of the first century AD; changes in practice, greater
or smaller, might have occurred before and after this period.
Nevertheless, I must emphasize that this reconstruction is a theoretical syn-
thesis: of the thirty-nine bakeries included in this study, not one presented all the
features discussed above; not one produced bread using all of the possible steps
mentioned. For instance, bakery IX 3, 19–20 contained the main three structures
used during grinding—a water system for soaking, four mills, and the L-shaped
small wall for sieving—and most of those necessary to knead and form loaves—
including a kneading machine, a boiler to warm water, embedded terracotta
vessels for the first rising, and a table with two masonry feet.48 However, there are
no clear traces of shelves, neither in the room where the kneading machine and
vessels were situated, nor in the space where the table on which loaves were
formed stood. In this particular bakery, the lack of shelves could be due to
preservation factors: although shelves identified by holes in walls are the most
common type of feature used for rising loaves in bakeries, they might not be the
only one. Indeed, as all known loaves found in Pompeii were leavened, allowing
the dough to rise would have been a crucial step in the operational sequence;
without it, loaves would have remained flat. Conversely, the practice of leavening
is attested in IX 3, 19–20, thanks to the two large terracotta vessels situated in the
kneading room. Thus, we need to consider in this case that, despite the lack of

47
Experiments conducted in the Saalburg, after the reconstruction of ovens found in a fort
occupied by the second Rhaetian cohort, led to a pre-heating phase that lasted twelve hours. Cf.
Knierriem and Löhnig (1997).
48
As it was not discovered during the excavation, the reconstruction of the boiler has had to
be based on the shape of the oven.
170 Nicolas Monteix

evidence in the standing remains, some space had been devoted to rising loaves,
most probably in the room where they had been formed just before.
Bakery I 12, 1–2, on the other hand, lacks two features that can be found in
IX 3, 19–20: no hydraulic system that would have enabled the grain to be
soaked before milling has been identified, nor was there any vessel that could
have been used for the first rising. The absence of these features does not seem to
have restrained production capacity, even over the long term if one accepts the
reconstructed evolution of this bakery, which will be discussed in more detail in a
later section. Thus, these two features, which are observed in only a few bakeries,
seem related to non-crucial steps whose presence is more significant than their
absence. Adding them to the operational sequence could have represented a
substantial initial cost, especially when the grain-tempering installation had
a connection to the water distribution network, as was the case in IX 3, 19–20.
All these non-crucial steps also imply the need for more manpower. An inter-
pretation of the textual evidence could lead one to assume that facilities were
added to enable a production process that, at least according to Roman taste, put
more emphasis on quality than on quantity. Tempering and sifting are both used
to whiten flour, which would have resulted in lower productivity, because of loss
of both flour and time, but at the same time would have produced a bread that
would have been greatly appreciated.49 But not all non-crucial steps served to
improve quality: the use of warm water while kneading would have permitted a
productivity increase, by shortening the time needed for kneading and rising.
Finally, a last technical point has to be made, despite its triviality: even at
Pompeii, only production phases that can be traced through material remains
can be reconstructed. Many, sometimes even crucial, operations cannot be
studied owing to lack of physical evidence. For instance, after the first rising—
if it occurred—the difference between ‘punching down’ or ‘folding’ the dough
has crucial repercussions for the general leavening of the loaf, but we cannot
reconstruct this for any ancient bakery. Some missing operations could per-
haps be restored from textual or iconographic evidence, but such evidence
should be treated cautiously in order to avoid over-interpretation. For
example, although the way in which the baker’s assistants have been carved
on the Bologna relief seems very ‘realistic’—notably, the hands forming the
loaves sink into the dough—it is highly improbable that one person ever
formed two loaves with both hands at the same time. In this manner, however
precise our reconstruction of the operational sequence may be, it will neces-
sarily always remain incomplete, although this does not invalidate it as a
heuristic tool for further analysis.

49
On the preference for white bread, see Thurmond (2006: 52–3).
 Pompeian Bakeries as a Case Study 171

SPATIAL I NSERTION AND P ATH

STANDARDIZATION

Almost all previous literature has failed to go further than providing a synthetically
designed description of the main features of bakeries; the first step beyond this
would be to insert each operation into the space of the workshop. This should lead
to an understanding of how technical challenges have been solved—or not—within
a peculiar spatial layout. Bakeries are not necessarily rationally organized or closely
built up. Of the three main stages, milling might be completely separated from
kneading and baking: eleven of the thirty-nine bakeries had no milling space in AD
79.50 If one discards the traditional interpretation of these bakeries as pistrina
dulciaria, this statement has two implications.51 First, some of the milling com-
plexes, which as far as is known in Pompeii and surroundings were always linked
with baking structures, had flour outputs beyond their baking capacities, allowing
the development of bakeries without mills.52 Secondly, this possible technical
separation has consequences for the spatial layout of the workshops: when install-
ing a bakery, it is not necessary to completely reshape the plan of the pre-existing
building in order to nucleate all the rooms linked with production. In particular,
milling space does not need to be directly adjacent to other production facilities.
However, probably because of their own building history, only two bakeries present
that kind of layout.53
On the other hand, the two last stages are always arranged in close prox-
imity. That is to say, most rooms dedicated to the forming and final rising of
loaves were deliberately sited directly adjacent to the oven. Two factors played
a role in this choice. In the first place, it would have facilitated the heating of
the room, which would have made rising easier. Moreover, the proximity
limited the distance between the rising spot and the oven, so that risen loaves
were less likely to be dropped or damaged during the transfer. For the same
reason, the kneading machine and the potential first rising vessels also needed
to be placed as close to the preparation tables as possible, though a longer path
would have had less consequence on bread quality, since the loaves still needed
to be formed at this step. Hence, even if clustering all structures linked to the
second and third stages would have shortened work paths and thus enhanced

50
A catillus, discarded and reused as a puteal, has been observed in VII 12, 11. As Fiorelli
(1873: 18) noticed, it might be a place where milling had ceased before the eruption. No sign of
milling has been revealed by a partial cleaning; see Monteix et (2015: 40–8).
51
Mayeske (1972: 171–2), but see Monteix (2010: 162–4).
52
Once chronological data for all bakeries have been obtained, it would be worth analysing
whether the simple bakeries were constructed after the mill-bakeries or after their progressive
enlargement. If the two phenomena do not appear to be linked, then milling complexes would
need to be looked for elsewhere—other places in the town, in villas, or even in water-mills.
53
Those two bakeries are situated in I 12, 1–2 and VI 5, 15.
172 Nicolas Monteix

productivity, there was no technical problem in setting kneading and/or first

rising apart from the forming of loaves.
As an example, bakery V 3, 8 illustrates such issues (Fig. 7.9). The first stages
are arranged around the atrium, in whose slightly transformed compluvium
the grain was soaked and then ground with the three mills disposed around it.
The ingredients were kneaded in the room south-east of the oven, where the
four feet of a table were also fixed in the ground. Once made, the dough was
left for a first rising in the small room between the oven mouth and the
preparation room, where it finally went back to be formed into loaves.

3
N

4
4

0 1 2 3 4 5 6 7 8 9 10 m

Fig. 7.9. From grain to bread in bakery V 3, 8. 1. from tempering to milling; 2. from
milling to kneading; 3. from kneading to first rising; 4. from first rising to loaves forming;
5. from loaves forming to bakery; 6. from baking to selling. Image: Nicolas Monteix.
 Pompeian Bakeries as a Case Study 173

In the bakery of the Casa dei casti amanti, the path is more rationalized. All
the productive structures were divided between two rooms, around the oven.
Four mills are clustered in the north-western part. South-east of those, in front
of the oven, are the kneading machine, close to the warm water, and two
reused catilli, transformed into the base of vessels for the first rising of the
dough. Once the dough had risen, it was brought to the room immediately
south of the oven, where there were three stone blocks supporting the bread-
shaping table. After the second rising, the loaves were transported to the oven,
thanks to the pass-through on its left. In this bakery, from grain to bread, the
product took a ‘circular’ clockwise path.
These two examples clearly exemplify how the apparent uniformity of
bakery structures is in fact misleading. There are no two identical bakeries:
the whole spatial structure depended both on the features used but also, and
more importantly, on the evolution of the workshop, and whether or not its
owner was able and or willing completely to restructure the existing layout of
the building.54

ECONO MIC INV ESTMEN T

Studying the spatial organization of bakeries in this way also brings up the
issue of investment, and the scale on which this took place. In Pompeii, no
house was built around a bakery: there is no trace of any workshop built
completely from scratch to incorporate the various rooms needed for the
technical processes of bread-making. Any installation thus had—at least to
some extent—to be adapted to a pre-existing general spatial layout. Thus there
was a tension between the technical requirements of the bakery and the
transformation works required to fit it into the pre-existing house. The
solution to this tension never seems to have favoured either a perfectly
rationalized bakery with a strictly linear or circular organization, or an un-
structured workshop thoroughly determined by the pre-existing building.
Varying degrees of spatial rationality emerge. The question is to determine
the point beyond which the spatial organization of a bakery had to be
rationalized in order to be functional, not to say profitable. Two examples
illustrate this issue.
Bakery I 12, 1–2 was inserted into a pre-existing house built on a narrow
strip plot (Fig. 7.10). The previous building on this site had a south/north
drain with a row of narrow rooms on its east side; the back part would have

54
This changing degree of rationalization in Pompeii can be compared with that observed in
Ostia, particularly in the Casseggiato dei Molini. See Bakker (1999) and Wilson (2008b: 406–8).
Pre-bakery phase 1st bakery phase 2nd bakery phase 3rd bakery phase
(terminus post quem: AD 22) (terminus post quem: AD 62/3) (terminus ante quem: c. AD 75)

.1
.1

.1

.1
MR_O
MR_O

MR_O

MR_O
M.1

S M.1

S M.1
SM.1
S

0 1 2 3 4 5 6 7 8 9 10 m

Fig. 7.10. The historical evolution of bakery I 12, 1–2. Image: Nicolas Monteix.
 Pompeian Bakeries as a Case Study 175

been used as a garden. Only minimal changes were undertaken in the building
of the initial bakery, which had only two animal-driven mills. In the front part,
the shop that opened on Via dell’Abbondanza was retained. A room was
connected to the mill room in order to install two L-shaped walls for sieving.
Just south of this, in another room that retained its original plan, a kneading
machine might have been installed.55 In the southern part of the house, a first
oven was built, which lacked the diaphragm wall. According to the excavations
conducted in this workshop, the first phase of the bakery has a terminus post
quem of AD 22.56 The second phase, which can be linked with the AD 62/63
earthquake, incorporated an extension to the bakery: a third mill was put in
place, with the surrounding pavement, the shop opening was closed, and the
mill room was unified by knocking down two dividing walls. The L-shaped
structure was transformed. Yet the most drastic changes took place in the
southern part of the house. By abolishing the garden, space was freed to install
a preparation room furnished with a kneading machine, two tables, and some
shelves. As part of this extension, a diaphragm wall was added to the oven,
allowing the insertion of a pass-through and a boiler. Because of these
transformations, the drain was slightly diverted. In the latest phase, dated to
the AD 70s, only a few changes occurred.57 Both the milling room and the
preparation room were extended south, at the expense of other spaces. Thus, a
fourth mill was added and the bread-shaping table was rotated, allowing more
people to form loaves.
In this bakery, a more rationalized production path emerged during the
second phase, when a third mill was installed. At this point, together with the
increase in production capacity, substantial building works were performed to
change the spatial layout. From this moment on, milling was strictly separated
from kneading, forming, and baking. A second point that emerges is a correlation
between the number of mills—the flour production capacity—and the number of
people theoretically involved in forming loaves.
Despite the lack of absolute dating, bakery IX 5, 4 offers a similar pattern of
expansion. From the investigation of its building history, it seems that the
creation of the bakery involved the knocking-down of half the plot. In the first
phase, three mills were installed in what was already a somewhat rationalized
plant: south of the milling room, a vast space was created for the kneading and
forming stages. The oven made the most of a pre-existing arched space. It is

55
The installation pit for this feature was observed during excavations. Amphorae fragments
had been used to wedge the kneading machine in place.
56
A Tiberian as (RIC I²: no. 81, p. 99) has been discovered in the backfill used to complete the
installation of the first basalt pavement.
57
A door, leading to a staircase and an upper-floor apartment, was opened in the north-east
part of the facade after the damages linked with the AD 62/63 earthquake had been repaired. This
door could have been closed and the staircase dismantled after the construction of the fourth
mill. After that, the apartment would have been accessible from the mill room.
176 Nicolas Monteix

possible that in this phase a pass-through already connected it with the

preparation room. The growth of the bakery in the next phase did not alter
the general layout: a fourth mill was added, and a first rising structure was
installed near the kneading machine. The main transformation linked with
these changes is the closing-off of the preparation room. During the last phase,
which most probably was unfinished in AD 79, the bakery seems to have been
dismantled, as would seem to be indicated by the raising of the floor level in
front of the oven by 20–30 centimetres.
This second case also suggests that three mills could be considered as the
point beyond which the spatial organization of the workshop had to be
rationalized, so that it followed the operational sequence as far as possible.
This spatial layout might be obtained either through substantial investment
during the initial construction of the workshop if it involved three or more
mills, or by subsequent changes if the original bakery had two or fewer mills.58

CONCLUSIONS

As far as can be reconstructed from archaeological remains, the bread-making

process in Pompeii was divided into three work clusters. The first one, grinding,
had three steps: tempering, milling, and sifting, the first and the last being in
some way optional: building a structure devoted to those steps was not compul-
sory. The second stage, dedicated to kneading and bread-shaping, shows fewer
optional steps: it was possible to use warm water to knead the dough, or to leave
out the first rising . The third and last stage seems, from a material point of view,
to have been quite uniform throughout the city: all thirty-nine bread ovens were
broadly built in the same way and most probably functioned similarly. Only
minor differences can be observed, but these relate to the presence of certain
optional features and are not directly linked with baking itself.
Focusing on the materialization of some technical clusters—such as the
eventual separation of milling from other operations, and the strong link
between rising and baking—helps in understanding the spatial layout of the
Pompeian bakeries. No one bakery is exactly identical to any of the others,
owing to the necessity of using pre-existing spaces to install or develop these
workshops: the spatial organization seems to have been a compromise
between technical needs and investment capacity. As they become bigger,
Pompeian bakeries tended to be laid out less in a do-it-yourself way, and to
gain in rationalization. Finally, besides shedding light on the bakeries of

58
In a forthcoming monograph on Pompeian bakeries, the author will test this hypothesis and
then connect it to a broader interpretative framework that discusses workshops as an economic
investment.
 Pompeian Bakeries as a Case Study 177

Pompeii, the approach presented in this chapter also suggests that analysing
the chaîne opératoire must be seen as a useful heuristic tool for a better
understanding of Roman urban crafts, even beyond technical matters.

ACKNOWLEDGEMENTS

This chapter would have been impossible to write without the kind permission of
the Soprintendenza Archeologica di Napoli e Pompei to study all Pompeian
bakeries, or, in the case of bakery IX 3, 19–20, without the opportunity to engage
in collaboration with the Expeditio Pompeiana Universitatis Helsingiensis
(EPUH—University of Helsinki). The majority of the research presented here is
derived from the ongoing project ‘Pistrina—Recherches sur les boulangeries de
l’Italie romaine’, funded by the École française de Rome and the French foreign
office through the Centre Jean-Bérard (Naples). Additional funding was also
provided by the Institut Européen d’Histoire et des Cultures de l’Alimentation
(IEHCA, Tours). All my thanks go to the team who have been involved in the
project since 2008, particularly to S. Aho, A. Coutelas, M. Derreumaux, L. Garnier,
C. Hartz, É. Letellier, S. Longepierre, V. Matterne, and S. Zanella. J. Andrews and
F. Trifilò kindly read and commented on this contribution, trying to enhance its
English. Any remaining mistakes are my own.

REFERENCES

Allison, P. M. (1995). ‘On-Going Seismic Activity and its Effects on the Living Conditions
in Pompeii in the Last Decades’, in T. Fröhlich and L. Jacobelli (eds), Archäologie und
Seismologie: La regione vesuviana dal 62 al 79 DC : Problemi archeologici e sismologici.
Munich, 183–90.
André, J. (1981). L’Alimentation et la cuisine à Rome (Collection d’études anciennes).
Paris.
Bakker, J. T. (1999). The Mills-Bakeries of Ostia: Description and Interpretation (Dutch
Monographs on Ancient History and Archaeology, 21). Amsterdam.
Balfet, H. (ed.) (1991). Observer l’action technique: Des chaînes opératoires, pour quoi
faire? Paris.
Barberan, S., Piquès, G., Raux, S., and Sanchez, C. (2006). ‘Un dispositif de cuisson
original en Languedoc dans l’antiquité: Les Fours à pain à cloche mobile en
céramique’, in SFECAG—Actes du Congrès de Pézenas. Marseilles, 257–71.
Borgongino, M. (2006). Archeobotanica: Reperti vegetali da Pompei e dal territorio
vesuviano (Studi della Soprintendenza archeologica di Pompei, 16). Rome.
Brun, J.-P. (2007). ‘Les Moulins hydrauliques en Italie romaine’, in J.-P. Brun and
J.-L. Fiches (eds), Énergie hydraulique et machines élévatrices d’eau dans l’Antiquité
(Collection du Centre Jean Bérard, 27). Naples, 201–14.
178 Nicolas Monteix
Buffone, L., Lorenzoni, S., Pallara, M., and Zantin, E. (1999). ‘Le macine rotatorie in
rocce vulcaniche di Pompei’, RSP 10: 117–30.
Cauvain, S. P. (2007). ‘Breadmaking Processes’, in S. P. Cauvain and L. S. Young (eds),
Technology of Breadmaking. New York, 21–49.
Ciancio Rossetto, P. (1973). Il sepolcro del fornaio Marco Virgilio Eurisace a Porta
Maggiore (I Monumenti romani, 5). Rome.
Coubray, S., Matterne, V., and Monteix, N. (forthcoming). ‘Of olives and wood: Baking
Bread in Pompeii’, in R. Veal and V. Leitch (eds), Fuel and Fire in the Ancient
Roman World: Towards an Integrated Economic Approach (McDonald Institute
Monographs). Cambridge.
Cresswell, R. (2003). ‘Geste technique, fait social total: Le Technique est-il dans le
social ou face à lui?’, Techniques & Culture 40: 125–51.
Cubberley, A. L., Lloyd, J. A., and Roberts, P. C. (1988). ‘Testa and Clibani: The Baking
Covers of Classical Italy’, PBSR, 56: 98–119.
Curtis, R. I. (2001). Ancient Food Technology (Technology and Change in History, 5).
Leiden and Boston.
Curtis, R. I. (2008). ‘Food Processing and Preparation’, in J. P. Oleson (ed.), The
Oxford Handbook of Engineering and Technology in the Classical World. Oxford and
New York, 369–92.
Esposito, D. (2009). Le officine pittoriche di IV stile a Pompei dinamiche produttive
ed economico-sociali (Studi della Soprintendenza archeologica di Pompei, 28).
Rome.
Fiorelli, G. (1873). Gli scavi di Pompei dal 1861 al 1872. Naples.
Fiorelli, G. (1875). Descrizione di Pompei. Naples.
Fulvio, L. (1879). ‘Delle fornaci e dei forni pompeiani’, in Pompei e la regione sotterrata
dal Vesuvio nell’ anno 79 Memorie e notizie publicate dall’ Ufficio tecnico degli scavi
delle province meridionali. Naples, 273–91.
Garçon, A.-F. (1997). ‘L’Innovation au regard du complexe technique minéro-
métallurgique armoricain’, Annales de Bretagne et des Pays de l’Ouest 104/3: 23–38.
Gell, W., and Gandy, J. P. (1817). Pompeiana: The Topography, Edifices, and Ornaments
of Pompeii. London.
Gille, B. (1978). ‘Prolégomènes à une histoire des techniques’, in B. Gille (ed.), Histoire
des techniques : Technique et civilisations, technique et sciences (Encyclopédie de la
Pléiade, 41). Paris, 1–118.
Knierriem, P., and Löhnig, E. (1997). ‘Panificium im experiment : Ein Erfahrungsbericht
aus dem Saalburgkastell’, in M. Junkelmann (ed.), Panis militaris : Die Ernährung des
römischen Soldaten oder der Grundstoff der Macht (Kulturgeschichte der antiken Welt,
75). Mainz, 134–6.
Lemonnier, P. (1983). ‘L’Etude des systèmes techniques, une urgence en technologie
culturelle’, Techniques & Culture 1: 11–34.
Leroi-Gourhan, A. (1965). Le Geste et la parole (Sciences d’aujourd’hui). Paris.
Maiuri, A. (1958). Ercolano. I nuovi scavi, 1927–1958. Rome.
Mau, A. (1899). Pompeii: Its Life and Art. London.
Mayeske, B. J. (1972). ‘Bakeries, Bakers and Bread at Pompeii: A Study in Social and
Economic History’, Ph.D. thesis, University of Maryland.
Mazois, F. (1824). Les Ruines de Pompéi: Seconde partie. Paris.
 Pompeian Bakeries as a Case Study 179
Médard, F., Borgard, P., and Moulhérat, C. (2011). ‘Le Travail du textile à Pompéi :
Ateliers et restes de tissus’, in C. Alfaro Giner, J.-P. Brun, and P. Borgard (eds),
Purpureae vestes III: Textiles y tintes en la ciudad antigua (Collection du Centre Jean
Bérard, 36). València and Naples, 83–90.
Monteix, N. (2010). Les Lieux de métier: Boutiques et ateliers d’Herculanum (BEFAR
344). Rome.
Monteix, N., Zanella, S., Aho, S., Macario, R., and Proudfoot, E. (2013). ‘Pompéi, Pistrina’,
Chronique des activités archéologiques de l’École française de Rome <http://cefr.revues.
org/954> (accessed 30 September 2014).
Monteix, N., Aho, S., Coutelas, A., and Zanella, S. (2014). ‘Pompéi, Pistrina’, Chronique
des activités archéologiques de l’École française de Rome <http://cefr.revues.org/1242>
(accessed 30 September 2014).
Monteix, N., Aho, S., Delvigne-Ryrko, A., and Watel, A. (2015). ‘Pompéi, Pistriana’,
Chronique des activités de l’école française de Rome <http://cefr.revues.org/1380>
(accessed 6 July 2015).
Moritz, L. A. (1958). Grain-Mills and Flour in Classical Antiquity. Oxford.
Oleson, J. P. (ed.) (2008). The Oxford Handbook of Engineering and Technology in the
Classical World. Oxford and New York.
Peacock, D. P. S. (1989). ‘The Mills of Pompeii’, Antiquity 73: 205–14.
Peacock, D. (2013). The Stone of Life: Querns, Mills and Flour Production in
Europe up to c.500 AD (Southampton Monographs in Archaeology, NS 1).
Southampton.
Pernot, M. (2006). ‘La Place de la technique dans les sociétés anciennes’, Cahiers
d’Épistémé 1: 7–25.
Strocka, V. M. (1991). Casa del Labirinto (VI 11, 8–10) (Häuser in Pompeji, 4). Munich
Thurmond, D. L. (2006). A Handbook of Food Processing in Classical Rome: For her
bounty no Winter (Technology and Change in History, 9). Leiden.
Veal, R., and Thompson, G. (2008). ‘Fuel Supplies for Pompeii: Pre-Roman and Roman
Charcoals of the Casa delle Vestali’, in G. Fiorentino and D. Magri (eds), Charcoals
from the Past: Cultural and Palaeoenvironmental Implications (BAR International
series, 1807). Oxford, 287–97.
Wikander, Ö. (2008). ‘Sources of Energy and Exploitation of Power’, in Oleson (2008),
136–57.
Wilson, A. I. (2008a). ‘Machines in Greek and Roman Technology’, in Oleson (2008),
337–66.
Wilson, A. I. (2008b). ‘Large-Scale Manufacturing, Standardization, and Trade’, in
Oleson (2008), 393–417.
Wilson, A. I., and Schörle, K. (2009). ‘A Baker’s Funerary Relief from Rome’, PBSR 77:
101–23.
Zimmer, G. (1982). Römische Berufsdarstellungen (Archäologische Forschungen, 12).
Berlin.
 III
People
 8

Disciplina, patrocinium, nomen:

The Benefits of Apprenticeship in the
Roman World
Christel Freu

Professional training in the ancient world has long been thought primitive, as
it was not formally practised in schools but depended only on the personal
relationship between the master and the apprentice.1 However, the absence of
schools does not necessarily mean that society did not consider ways to
institutionalize and regulate apprenticeship. Although informal training con-
tinued to play its role in the Roman Empire, we also have signs of the
formalization of technical training: when a family wanted to send a child
away to some friend or to some renowned master and expected good training,
the teaching could be an object of contract.
Because of the benefits it offered for masters as well as for their pupils,
formalized technical training was a much sought-after activity. Most scholar-
ship on the issue has focused on master craftsmen, who had reached the top of
their craft and were often also teachers, deriving authority and wealth from
their teaching, and taking advantage of a cheap labour force that over the years
acquired better skills, thus contributing to the good reputation of their work-
shop.2 Yet, for their pupils too, benefits were considerable. There is no doubt
1
Marrou (1948: 286–7); cf. Schulz-Falkenthal (1972: 194): ‘sich . . . die berufliche Ausbildung
im Handwerkerstand für unsere modernen Begriffe noch weitgehend unsystematisch, unorga-
nisiert, mehr spontan als geregelt vollzog.’ It is true that the choice of a master was certainly not
regulated; but the attention of the imperial cities towards their apprentices was not completely
non-existent, at least with regard to the taxes they raised. For Egypt, see Freu (2011: 28–35).
2
See Schulz-Falkenthal (1972: 196–8); Tran (2010; 2011: 128–30; 2013: 147-85) for magistri
and didaskaloi seen as honoured craftsmen in Greek and Latin inscriptions. Tran highlighted
well the existence of craftsmen specialized in training. I use the English term ‘craftsman’ to
qualify a professional involved, in one way or another, in the production process. See Monteix
(2011) for a reflection and better definition of the craftsmanship concept. He prefers not to
employ the French terms artisanat and artisan. But I have not found an English word to replace
‘craftsman’.
184 Christel Freu

that a qualified person earned more than a manual worker. As shown by

Stanislas Mrozek and Richard Duncan-Jones, qualified slaves were more
expensive than basic unskilled workers. These costs undoubtedly reflect their
scarcity but also the level of income that could be earned in the labour
market.3 This fact was stressed by Cicero in his defence of the actor Roscius,
when he claimed that educating a slave as an artist would increase his basic
value a hundredfold.4 This plea, the Pro Roscio Comoedo, incomplete and
difficult to date, deals with a conflict between a slave-owner and his associate
Fannius, who, as it seems, was granted half the ownership of the slave,
Panurgus. When the slave died, Cicero defended Roscius’ rights of ownership
and also the fact that he owed nothing to his former associate, Fannius. Not
being fully able to show that the slave was really shared by both, Cicero mostly
underlined how much the teaching itself could account for the fact that part of
the increased value should go to the teacher who had trained him.
Disciplina, patrocinium, nomen, which can be translated as ‘the teaching of
skills, patronage, and fame’, were, for Cicero, the rewards that an apprentice in
art or craftsmanship could expect to get.5 Although the topic of his discourse is
only artistic training, which is specific, because artists can be much sought
after and very rich, the text can be regarded as paradigmatic. To different
degrees, the craftsmen who had succeeded all seem to agree that apprentice-
ship had brought much to them: on their funeral stones, they recall their
professional skills, their master’s reputation, and their integration in his social
networks, thus imitating the way of life and social ideals of the elite. They are,
for us, the best examples of the middle classes’ rise in the empire.
To explore this social practice, I will focus mostly on sources from the
eastern part of the Roman Empire: Egyptian contracts of apprenticeship,
didaskalikai, which have already been much studied, some literary texts, and
some inscriptions—obtained with the help of the Packard Humanities Insti-
tute database, although some had already been collected by Louis Robert.6 In
this corpus, we find dedications from pupils to their teachers, and also from
teachers to their pupils, some signatures of mathètai who liked to recall the
name of their teacher, and, finally, epitaphs, which provide interesting details

3
Duncan-Jones (1974: 348–50; app. 10); Mrozek (1975: 45–8); Tran (2011: 122).
4
Cic., Q. Rosc. 28, ed. J. Axer, Leipzig: this edition of the text, published in 1975, is based on
the only manuscript of the text, rediscovered some time previously. The figures used by Cicero
are more coherent than in other modern editions: Panurgum tu, Saturi, proprium Fanni dicis
fuisse, at ego totum Rosci fuisse contendo. Quid erat enim Fanni? Corpus. Quid Rosci? Dis-
ciplina. . . . Ex qua parte erat Fanni, non erat HS 6,000, ex qua parte erat Rosci, amplius erat HS
600,000 (‘Satyrus, you say that Panurgus was the property of Fannius. And I contend that he
belonged entirely to Roscius. What belongs to Fannius? His body. What belongs to Roscius? His
teaching. . . . Fannius’ part did not make him worth 6,000 sesterces; Roscius’ part made him
worth more than 600,000 sesterces’). See also Tran (2013: 164–77) for other examples showing
how important the training was to increase the value of slaves (Dig. 6, 1, 27, 5; 6, 1, 28; 13, 7, 25).
5 6
Cic., Q. Rosc. 30. Robert (1936: 56–7).
 The Benefits of Apprenticeship in the Roman World 185

about training and success in the craft. Compared with the numerous inscrip-
tions referring to professionals of artes liberales, rhetoricians, sophists, phys-
icians, or grammarians, detailed tombstones of craftsmen are rather rare.7
There may be several reasons for that. In the first place, there is the relative
modesty of the craftsmen’s background: although their names often appear on
stones, little additional information is given; fortunately for us, there are some
exceptions.8 Secondly, when the training took place inside the family circle,
which seemed to have still been regular practice in some milieux, pupils did
not have an incentive to specify the name of their master.
In order to examine the benefits of apprenticeship, I shall begin with the
content of the training, the disciplina; then, I shall show how the apprentices,
as Cicero said, wanted to claim legitimacy from the name and fame (nomen) of
their masters and from the social networks (patrocinium) they built during the
time of the training. To conclude, I will examine some clear-cut cases of
upward mobility resulting from apprenticeship.

DISCIPLINA

Informal apprenticeship undoubtedly existed in Greece and Rome: Plato tells

us about it in his Protagoras.9 The details of this training appear in our sources

7
Some epitaphs of physicians or rhetores are discussed later in the chapter. It must be noted that
the Greeks had the same world, technè, to qualify the activity of an artist, a physician, or a craftsman.
Therefore, the differences between them are of degree, not of nature. Cf. Frasca (1994: 6). Even
though the Roman elites differentiated artes liberales and craftsmanship (e.g. Dig. 50, 13, 1), the
same term, ars, applies to these occupations, and excellence of professional know-how was claimed
similarly by intellectuals and craftsmen. For that matter, the middle classes did not have the same
point of view as the elite. In Petronius, Satyricon 46, artes liberales were seen by a rich centonarius,
Euchion, as a profession like any other: quod si resilierit, destinaui illum artificii docere, aut
tonstreinum aut praeconem aut certe causidicum . . . Litterae thesaurum est, et artificium nunquam
moritur (‘if he gives it up [sc. his juristic studies], I decided to have him learn a craft: either
hairdresser, or herald, or pleader . . . Education is a treasure, and the science of a craft never starves’).
See Mayer (2012: 22–60); Monteix (2011: 13); Tran (2011: 122): ‘en réalité l’apologie de la doctrina
et du travailleur doctus était commune à l’artisanat et à d’autres activités que les modernes classent
parmi les métiers intellectuels.’
8
For the ‘silence’ of workers in antiquity, see Joshel (1992: 8–15), who wished to ‘give voice to
the silent Roman’; see also Salvaterra (2006: 19; 22).
9
Pl., Prt. 328 a: οὐδέ γ’ἄν, οἶμαι, εἰ ζητοῖς τίς ἂν ἡμῖν διδάξειεν τοὺς τῶν χειροτεχνῶν ὑεῖς αὐτὴν
τὴν τέχνην ἣν δὴ παρὰ τοῦ πατρὸς μεμαθήκασιν, καθ’ὅσον οἱός τ’ἧν ὁ πατὴρ καὶ οἱ τοῦ πατρὸς φίλοι
ὄντες ὁμότεχνοι (‘it won’t be easy, I think, if you try to find out who could teach our craftsmen’s
sons the same craft as their father, which they learnt from their father himself, as much as the
father is able to do, or his friends, who share the same craft’). Feyel (2006: 457–63) believed that
the text attests apprenticeship inside the family circle. However, in reality, Plato is more subtle:
even if he underlines the secret that surrounds the apprenticeship (Prt. 327 b), he says here that
the friends of the craftsman can also teach the young men. For informal apprenticeship in Rome,
see Tran (2010: 198; 2013: 161–70), who thinks that this type of apprenticeship was the most
common in the Roman world.
186 Christel Freu

only rarely, as apprenticeship was practised inside the craftsmen’s home and
was fully integrated into family labour and parental education.10 Yet even in
classical or Hellenistic times, this informal training was not the only scenario,
and the mobility of apprentices looking for well-known masters all over the Greek
world is widely attested.11 During the late Roman Republic, where the gentes were
still a leading element in society, we know about apprenticeship in the circle of the
gens, with liberti who learned the craft chosen by the dominus. Yet the texts do not
say where the training took place and under whose guidance.12
As for imperial provinces, the Egyptian didaskalikai of the Roman period
show, for their part, that many craftsmen families wanted to place their
children outside the family, be it at friends’ homes, or in the households of
close or even distant relatives, as if moving to another home was already part
of the training. Plato certainly thought that the friends of fathers were like the
family. Nevertheless, they brought a distinct savoir faire and an authoritarian
distance. It is not surprising that Lucian of Samosata left his position as an
apprentice at his uncle’s, and ran back to his mother’s arms, after being beaten
on the very first day: we see here a clear difference between the two worlds.13
Disciplina, far from being secret, was thus shared by professional circles of
friends, which could contribute to improving techniques.14 Yet in the imperial
period, at least in Egypt, an interesting evolution may be observed: even
though, in the beginning of the imperial era, apprenticeship was still often
entrusted to friends, longer training entrusted to more remote craftsmen
seems to have become increasingly frequent.15
Undoubtedly, apprenticeship was provided not in a school but in the
master’s workshop or in his house. In most cases, the child was thus placed
in a subordinate position to an external authority and had to comply with
discipline imposed by a relative stranger, which was all new for him.16 We do

10
Nevertheless Greek sculptors were in the habit of including the name of their master or
father in their signatures. Examples can be found in the work of Marcadé (1953); e.g. Charmolas
and Menodotos, Rhodian, sons of Artemidoros and active c.100–80 BC (I.17–20); Cephisodotos II
and Timarchos of Athenes, sons of Praxiteles and grandsons of the sculptor Cephisodotos I,
active at the end of the fourth century BC (I.53–6 ). See also I.65, and I.110–11. For more complete
information, see now Muller-Dufeu (2002).
11
In his Periegesis Pausanias gives examples of apprentice sculptors travelling for their
training, from the archaic to the imperial period. Pausanias’ sources were either the sculptors’
signatures, or the local guides (exegetai). e.g. Periegesis 5, 17.1, 3-4, 20.2; 6, 1.3; 3.4–6, on
Cantharos from Sicyon, active in the fourth century BC; 6, 3.11, and 4.4 on Pythagoras from
Rhegion, active in the fifth century BC. For a complete survey of ancient sources mentioning
Greek sculptors, see Muller-Dufeu (2002).
12
Donderer (1996: 58), giving the example of the gens Cossutia, who specialized in marble
production and whose freedmen were famous temple architects.
13 14 15
Lucian, Somn. 4. Freu (2011: 30–1). Bergamasco (1995: 150–61).
16
See Cribiore (1996: 6) for the writing schools, and Frasca (1999) for a comparison between
apprenticeship in literary ludus and in taberna (sp. 151: ‘la bottega artigiana si sostituisce alla casa,
il magister si assume in toto le prerogative del pater, e con analoga autorità dispone di loro’).
Whatever the place used for the teaching was, the important thing was that it was away from home.
 The Benefits of Apprenticeship in the Roman World 187

not know much about the content of the teaching. Most testimonies alluding
to this issue are quite vague, merely mentioning that the master will teach all
he knows. The didaskalikai say: ὥστε μαθεῖν . . . τέχνην πᾶσ[αν ἀυτὸ]ν ὡ[ς καὶ
ἁυτὸς ἐπίσταται (‘so that he learns all the art, as [the teacher] himself knows
about it’).17 Indeed, contracts insisted mostly on the personal and individual
character of the professional technè taught by the master. Only the teaching
contract—the Lehrvertrag, which differs from the Lehrlingsvertrag and is
much rarer—shows the different steps: in the two or three years of the
training, the child must learn precise techniques, in a predefined order.18
It has already been pointed out by Heinz Schulz-Falkenthal that profes-
sional training did not only provide technical instruction; it also introduced
the young apprentice into social networks that would be useful for his career.19
First of all, it brought a nomen—that is, a reputation, even fame.

NOMEN

Nowadays, in the crafts and in the liberal arts, the name and fame of the
master still reflect upon his pupils. No matter what the personal quality of the
apprentice, being able to highlight the name of one’s master, especially a
renowned one, is an immediate benefit. When a craftsman’s skills were
transmitted inside the family, the reputation of the family was transmitted
from generation to generation, and spread locally or over larger areas; father
and son, uncle and nephew, grandfather and grandson, could work together
and the youngest could benefit from the name and reputation of his ancestors:
in Greece, temple archives show collaborations of this type between fathers
and sons.20 In imperial times, the obvious example of Lucian’s life reminds us
of this fact; in Roman Egypt, the papyri tell the same story.21
When the master was not a relative, but a renowned skilled worker, his
name was all that the pupil inherited from him, and thus was all the more
important. Artistic and intellectual groups took good advantage of this: they
inscribed themselves in the lineage in which fictive and symbolic affiliations

17
P. Wisc. 4 (Oxyrhynchos, AD 53). See also P. Tebt. 385 (Tebtunis, AD 117); P. Fouad I.37
(Oxyrhynchos, AD 48); P. Oxy. II.322 (AD 26); P. Oxy. XL.2971 (AD 66); PSI 1,132 (Talei, AD 61),
with lacunas.
18
We possess only three of these contracts, two about the shorthand apprenticeship and the
third about an apprenticeship of playing music. Cf. Bergamasco (1995: 100–1).
19
Schulz-Falkenthal (1972: 196).
20
Feyel (2006: 457–8). For painters, see Plin. NH 35.53–148, where apprenticeship appears to
have been a family practice for some classical Greek painters (NH 35.54; 35.60; 35.108–11).
21
Lucian was apprenticed to his uncle, whose father was already a sculptor. Lucian, Somn.
1–2; see Jones (1986: 9–10).
188 Christel Freu

enabled one to claim the transmission of professional art. This phenomenon is

known for artists and sophists.22 For the Greek sophists of imperial times, the
bonds between pupils and master were so tight that they were compared with
family ties: the ζηλωταί or ἑταῖροι were brothers, son of their masters; in late
antiquity, Athenian philosophers used to claim that their families went back to
Plutarch or Philostratus.23
For that reason, placing emphasis on the master’s name was crucial. Cicero
highlights precisely this point:
What hope and expectation Panurgus aroused, what interest and favour he
brought to the stage for the reason he was the pupil of Roscius! Those who
loved Roscius favoured his pupil, those who admired Roscius, approved his pupil,
those, in short, who heard Roscius’ name pronounced, felt that his pupil was well
and perfectly trained. Such is the crowd: many of its judgements are inspired by
fame, a few by truth.24
What is true for artistic training also holds for a craftsman’s apprenticeship.
Thus, a detailed funeral carmen found in Tarragona, written on an altar by a
discipulus, begins by underlining the importance of the master craftsman’s
name and renown: Iulius hic fuerat nomine summo artificioque Statutus (‘This
was Julius Statutus, very great by his name and art’).25
Another way for a mathètès to emphasize his master’s name was to sign his
product with his name set side by side with the name of the didaskalos or the
magister. We thus find in Rome beautiful statues signed by sculptors who liked
to draw their glory from their famous master: a statue representing Orestes
and Electra, dating from the Augustan or Tiberian era and belonging to the
Ludovisi collection, is signed by Menelaos, mathètès, ‘pupil’, of Stephanos;26
another statue representing an athlete and dating from the second half of the
first century BC is signed by Stephanos, mathètès of Pasiteles, who, according to
art specialists, was a famous sculptor from the middle of the first century,
known for imitating classical models.27 The signature mentions the name of
the master in the same position as a patronymic, thus showing an interesting
parallel between biological and intellectual affiliation. In the same way,
apprentices of workers in mosaics could mention proudly the name of their

22
For Greek painters, see Plin. NH 35.75–77; 25.110–14.
23
Goulet-Cazé (1982: 255–6) underlines the existence of idealized kinship ties between the
philosopher’s pupils and the master; see also Puech (2012).
24
Cic., Q. Rosc. 29: ‘Quam enim spem et expectationem, quod studium et quem fauorem secum
in scaenam attulit Panurgus quod Rosci fuit discipulus! Qui diligebant hunc, illi fauebant, qui
admirabantur hunc, illum probabant, qui denique huius nomen audierant, illum eruditum et
perfectum existimabant. Sic est uolgus; ex ueritate pauca, ex opinione multa aestimat.’
25
RIT 447 = AE 2000, 802.
26
IGUR 4.1575: Μενέλαος Στεφάνου μαθητὴς ἐπόει (‘Menelaos, pupil of Stephanos, made
[it]’).
27
IGUR 4.1584: Στέφανος Πασιτέλους μαθητὴς ἐπόει (‘Stephanos, pupil of Pasiteles, made
[it]’).
 The Benefits of Apprenticeship in the Roman World 189

master, as an example in Michael Donderer’s corpus of mosaic inscriptions

shows: a Gallic mosaic-worker in Lillebonne signed side-by-side with his
master from Puteoli, as his discipulus.28 Their names, within the tabulae
ansatae, were put clearly in parallel.

PATROCINIUM

The master did not only transmit his skills and his fame; like a father to his
son, he bequeathed also his social networks, friends, and customers. In
Cicero’s speech, the master left to his pupil his patrocinium: raised, nourished,
and educated during the time of his training, the apprentice belongs to his
master’s ‘house’, like his friends or his children.29
As is well known, the large houses of the elite became, in the late republic
and during the empire, hives of economic activity where slaves, freedmen,
salaried persons, and apprentices lived and worked.30 Working there was the
best way to get acquainted with customers and friends of the master. The
better off the master was, the more important was the number of the threptoi,
or persons living in.
Craftsmen, generally, did not possess the means of the Roman elite. Their
workshops seem, on average, to have been rather small, especially in small
towns or villages.31 At best, they had a couple of pupils. Nevertheless, they
claimed to model themselves on the masters of the big households, where
amicitia and philia ruled the relationships between free persons.32 In Egypt,
too, apprentices lived with their master from dawn to dusk and even slept at
their master’s home when their families were far away.33 Gravestones that

28
The mosaist’s signature at Lillebonne is in CIL 13.3225 (=AE 1978, 500): T(itus) Sen(nius) c
(iuis) Puteolanus fec(it) || et Amor, ciuis Kaletus, discipulus. (‘Titus Sennius, citizen of Puteoli, and
Amor, citizen of Cales, his pupil, did [it].’). See Donderer (1989:108–11 (no. A86)).
29
Cic., Q. Rosc. 30–1.
30
For social relationships within the Roman house, see Wallace-Hadrill (1994).
31
For Asian trades, see Drexhage (2007: 182–6), and, for more general views, see Kehoe
(2007: 559–66), who nevertheless recognizes the existence of big workshops in great cities.
Hawkins (2012) emphasizes that Roman craftsmen during the Roman Empire, when labour
costs increased, preferred to ‘accommodate an increasingly specialized workforce in subcon-
tracting networks rather than in integrated firms’. But his best examples come mostly from the
building industry, which is specific because of the irregular employment of workers on building
sites (as we see also in Egyptian papyri). We also have evidence for permanent workshops and
service industry on a greater scale in big cities or on large estates. For example, as Flohr (2011:
esp. 91–3; 2013), showed, Italian fulling workshops in big cities such as Rome or Ostia can be of
significant size. On this issue, see also the important and nuanced remarks of Wilson (2008).
32
e.g. Lucian, Περὶ τῶν ἐπὶ μισθῷ συνοντῶν, 1 (= On salaried posts), where philia appears as a
bond claimed hypocritically between the house’s master and his employees.
33
For references, see Bergamasco (1995: 127–8); see also Laes (2011: 192).
190 Christel Freu

commemorate teachers and pupils, whatever their crafts were, corroborate this
picture. We see apprentices, alone or as part of a group, paying for the
gravestone of their master.34 Pupils participated with family members in their
homage to the deceased. Conversely, masters paid tribute to their deceased
pupils.35 Most significant gravestones placed mathetai after the children: in
one inscription from Saittai in Lydia, a city that has revealed a great number
of craftsmen’s epitaphs, the mathètès of an unknown trade was recalled as a
relative outsider, after the children and the threptoi—the young slaves born in
the house or exposed children raised by the family—but he clearly belonged to
the home that took him in: ‘The 211th year, in the sixth day of Xandikos month,
Ioulios and Loukios honoured their mother Eirene and so did her daughter-
in-law Ammia, the threptoi Eirenaios and Peia, and Gaius the apprentice.’36
Similarly on a stone from Rome, of the Severan era, a man from Smyrna
who was perhaps a doctor or a rhetor set up a funerary monument for himself
and the members of his big household:
To the Manes. Caius Septimius Heracleitos made [this monument] for himself
and his companion Aufidia Capitolinè, for their freedmen and their descendants,
for his heirs and their descendants, for Satorninos Claudianos son of Abascantos
from Smyrna and his descendants, for Loucios Catilios Chryserôs, his pupil, and
Marcos Gabios Dios and their descendants. As for the rest, Heracleitos, rejoice
without envy, for you have to die one day, claimed by the Fates.37

34
A sophist who taught his art in Ephesos is thus honoured by a monument voted by the local
senate with a donation of his pupils: IEph. 1548, l. 4–5= SEG 13.506: κ[ατ’ ἐπίδοσιν] | τῶν
μαθ[ητῶν]. Puech (2002: 455) reads κ[αὶ στησάντων?] | τῶν μαθ[ητῶν] ([statue] erected by his
pupils).
35
IG 10.2.1, 879. This inscription, from second-century AD Thessaloniki, shows a certain
Dioscorides honouring the memory of his pupil Iulianos: Διοσκουρίδης | Ἰουλιανῷ τῷ | μαθητῇ |
ἐκκ τῶν ἐκίν- | ου μνείας χάριν. See also Laes (2011: 193–4) for two other examples from Cordoba
(CIL 11.2243) and Rome (CIL 6.10013).
36
SEG 31, 1006 (second century AD): ἔτους σιαʹ, | μη(νὸς) Ξανδικοῦ | &ʹ. Ἰούλιος καὶ Λούκιος |
Εἰρήνην τὴν μητέ- | ρα, Ἀμμια νύμφη, Εἰρη- | ναῖος, Πεία οἱ θρεπτοὶ | καὶ Γάϊος ὁ μαθητὴς |
ἐτείμησαν (I do not know how to interpret the link between the honoured woman and the
apprentice. Was she his mistress or his master’s wife?). For the threptoi in the family, see Nani
(1943–4). See also Frasca (1999: 151–3), who speaks of a ‘coinvolgimento affetivo implicito in
questo singolare tipo di rapporto docente-discente’. A literary but important testimony of the
middle of the nineteenth century, referring to a particular context of pre-industrial type of trade,
seems to illustrate perfectly what we read in our Roman sources. H. de Balzac describes the
familial integration of young apprentices in a Parisian shop: ‘Les maîtres adoptaient leurs
apprentis. Le linge d’un jeune homme était soigné, réparé, quelque fois renouvelé par la maîtresse
de maison. Un commis tombait-il malade, il devenait l’objet de soins vraiment maternels. En cas de
danger, le patron prodiguait son argent pour appeler les plus célèbres docteurs ; car il ne répondait
pas seulement des mœurs et du savoir de ces gens à leurs parents. Si l’un d’eux . . . éprouvait
quelque désastre, ces vieux négociants savaient apprécier l’intelligence qu’ils avaient développée, et
n’hésitaient pas à confier le bonheur de leur fille à celui auquel ils avaient pendant longtemps confié
la fortune’ (H. de Balzac, La Maison du chat-qui-pelote (Paris, 1842)).
37
IGUR 2.936: Θεοῖς Καταχτονίοις {Καταχθονίοις} Γ(άιος) Σεπτίμιος Ἡράκλειτος ἐποίη- | σεν
ἑαυτῷ καὶ Αὐφιδίᾳ Καπετωλίνῃ | συμβίῳ καὶ | ἀπελευθέροις καὶ τοῖς | μετ’ αὐτοὺς ἐσομένοις καὶ
κληρο- | νόμοις ἑαυτοῦ καὶ τοῖς ἐξ αὐτῶν | ἐσομένοις καὶ Σατορνίνῳ Ἀβα- | σκάντου Κλαυδιανῷ
 The Benefits of Apprenticeship in the Roman World 191

These cases are, of course, rare; moreover, the inscriptions underline that
some pupils were beloved and more integrated into the family. One pupil
stood out from the group of his companions, because of his skills or his social
origins, or because he had gained the special affection of his master: in a stone
from Thessalonike, dating from the second century AD, the master dedicator
said that the mathètès for whom he carved the inscription was one of his pupils
(ἐκκ τῶν (μαθητῶν)).38 On the altar from Tarragona, the beloved pupil says:
scripsi haec unus ego ex discipulis prior omnibus illis (‘I, the one and first of all
pupils, wrote this’).39 He was the one who profited most from his master’s
relations and friends.
For the master also bequeathed to his pupils his patrocinium—that is, in
this context, his trade and customers. This is how we should understand
their insistence in recalling how much their master was cherished and
popular. The altar of Tarragona emphasizes: non uno contentus erat, plur-
ibus gaudebat amicis . . . Reliquit suboles suae posteros stationis futuros, | per
quos ut statio Statutiq(ue) nomen habebit, | tres paene aetate pares artificio
ministros (‘He did not content himself with one friend, but he enjoyed
many . . . He left as offspring, heirs of his workshop, through whom the
workshop and the name of Statutus will remain, three servants almost
equal in age and skill’).

Σμυρναίῳ | καὶ τοῖς ἐξ αὐτοῦ ἐσομένοις καὶ | Λουκίῳ Κατιλίῳ Χρυσέρωτι μα- | θητῇ καὶ Μ(άρκῳ)
Γαβίῳ Δίῳ καὶ τοῖς | [ἐ]ξ ̣ αὐτῶν ἐσομένοις τὰ δὲ | [λοι]πά Ἡράκλειτε εὔφραινε | [σαυτ]ὸν ἀφθόνως
τὸ γάρ ποτε | [δεῖ]ν σε θανεῖν Μοίραις | μ ̣εμέληται.
38
IG 10.2.1, 879.
39
RIT 447 = AE 2000, 802. I quote the reference in integro, for its particular interest: Iulius hic
fuerat nomine summo artificioque Statutus/tractabatque uiris aurum, mulieribus atque puellis.
Plenus omni ope moribus, uita, disciplina beatus, non uno contentus erat, pluribus gaudebat
amicis. | H(a)ec illi semper uita fuit: mane et sexta lauari. | Reliquit suboles suae posteros stationis
futuros, | per quos ut statio Statutiq(ue) nomen habebit, | tres paene aetate pares artificio
ministros. | Scripsi haec unus ego ex discipulis prior omnibus illis, | Secundinius Felicissimus ego,
set nomine tantum. | Hoc quot potui, magister: tibi contraria munera fo[ui]? | Addo scriptura tuis
tumulis sensus, siue exter ubique serues utque tuos amicos meque cum illis; | ut quotienscumque
tibi annalia uota dicamus, | ut et uoce pia dicamus : Carnunti sit tibi terra leuis (‘This was Julius
Statutus, very great by his name and art. He worked gold for men, women and young girls. Full of
every kind of wealth, happy with his lifestyle, life, and profession. He did not content himself
with one friend, but he enjoyed many. This was always his life: to bathe in the morning and at six.
He left as offspring, heirs of his workshop, three servants almost equal in age and skill, through
whom the workshop and the name of Statutus will remain; I, Secundinius Felicissimus [“very
happy”], but only by name, the one and first of all apprentices, wrote this. Master, I did what
I could: and I promoted gifts converse to those given by you. I bestow to your grave, by these
words, ideas and feelings, in order that, everywhere, whatever stanger you may be, you could
keep your friends and I among them; in order that, whenever we pronounce birthday vows for
you and that we say with a pious voice: may the earth of Carnuntum rest lightly on you’). For
commentary, see Gomez Pallarès (2000: 417–28), and for the interpretation of the last line see AE
2000, 802: Carnuntum may be the signum of Statutus or the place of his death, which is the
interpretation I retain because of the lines 12–13 (siue exter ubique serues utque tuos amicos).
192 Christel Freu

A Bithynian epitaph from the imperial period illustrates how an apprentice,

thanks to his master, acquired many friends:
In fond memory of you, master, I engraved this stele and set it on your grave,
I Euphras, who love Vestalis, even though he lies dead, for when one has received
some good, he does not forget it; he taught me his art with the greatest devotion
possible; he gave me the means to live and eat, by making me a craftsman tailor dear
to all my friends; after raising and teaching me his art, in the hope that he would get
back from me as much gratefulness, he did not get it but died at the age of thirty-five.40
These lines evoke very clearly the reciprocity of favours, typical of the ex-
change between amici or philoi.41 This amicitia motif is a privileged theme of
funerary inscriptions of the plebs media, because it emphasizes the sociability
of the deceased as well as his ties of friendship with the dedicator. There was
surely a desire to imitate the sociability of the elite and to reinforce the dignity
of the craftsman by depicting him as surrounded by many friends. We must
dwell on this point for a while: it has been said that middle-class members,
among whom we have to rank the craftsmen who celebrate, with a ‘modest
pride’, their memory on stone, cultivated a ‘class wisdom’ through which they
could differentiate themselves from the infima plebs, on the one hand, and
from the civic elite, on the other.42 To be sure, they valued their work and skills
with a specific and proper emphasis, but this did not prevent them from
imitating the elite to some extent: the focus on amicitia must be seen in that
light.43 But does this mean that there really was a specific class wisdom?
I doubt it. To explain the claim of these epitaphs, it makes more sense to use
the model of ‘cultural revolution’—the upward mobility of popular classes and
the desire to imitate the elite’s culture now spread throughout the empire.44

40
TAM 4, 132 = Labarre and Le Dinahet (1996: no. 76: 6). Gravestone from the area around
Nicomedia, imperial period (Bithynia): μνήμης χάριν γλυκίας σῆς, ἐπιστάτα, | στήλην ἔγρ<α>ψα
τήνδε κὲ ἐθέμην τάφῳ | Εὐφρᾶς ποθῶν Βεστᾶλιν, κὲ ἂν κεῖτε νέκυς· | ὁ γὰρ καλῶς παθών τις λήθην
οὐκ ἔχει· | τέχνην διδάξας τῆς πάσης μετ’ [ε]ὐνοίας | ἔδωκεν ἀφορμὴν τῆς [ζ]ωῆς κὲ τῆς τροφῆς |
ῥάπτην τεχνείτην παρὰ πᾶσιν φίλοις φίλον· | ὅτα<ν> ἐξέθρεψε κὲ ἐδιδάξατο τέχνην, | ἐλπίζων παρ’
ἐμοῦ χάριτας [[—]] τὰς αὐτὰς ἀμοιβὰ[ς] λαβεῖν, | οὐκ ἔτυχεν, ἀλλ’ ἔ[θ]ανε πέντε κὲ τρειάντα ἐτῶν.
41
Was there a difference between Roman amicitia and Greek philia? It depended on the
cases, of course, but we may consider that the two terms were roughly equivalent: the definition
of philia by Greek philosophers helped to theorize Latin amicitia. And, in practice, amicitia and
philia were equally an essential social tie in imperial times, supporting exchanges of goods and
services in Latin West as well as in the Greek world. On this topic, see Verboven (2011).
42
Veyne (2005: 131–61), also insists on the amicitia motife. e.g. the excellent faber tignuarius
from Arles, Q. Candidus Benignus, was also remembered by his wife and daughter as a good
friend: ‘he knew how to entertain his friends’ (CIL 12.722: ‘nosset qui pascere amicos’).
43
Veyne (2005: 137–40) underlines this elite’s imitation (‘ces épitaphes s’inspirent de la
tradition indigène de l’éloge funèbre des grands personnages: à son heure dernière, la classe
moyenne imite l’aristocratie’), which seems to me somewhat contradictory to the assumption of
a specific ‘class wisdom’.
44
For discussion of this model, see Wallace-Hadrill (2008: 28–37, 435–54); Mayer (2012:
100–65) contests the model, by emphasizing the specificities of middle-class culture. I think that,
even if there are some cultural peculiarities in the craftsmen’s self-representation, forms of
 The Benefits of Apprenticeship in the Roman World 193

After becoming masters of their craft, thanks to an apprenticeship, skilled

‘craftsmen’ belonged to the same world as the Roman elite.45
However, their imitation of the elite did not prevent craftsmen from
appropriating in their own way other cultural models: for the Greek author
of the Bithynian poem, this kind of friendship made it easier to become richer
by practising the craft (‘he gave me the means to live and to eat, by making me
a tailor dear to all my friends’). The reciprocity of favours is here expressed in
terms of economic benefits. We know indeed how amicitia served as a
framework for economic exchanges in Roman society, allowing friends to
justify and to structure deposits and guarantees, money loans, legacies, and
donations.46 In the context of crafts, these philoi or amici would include, in my
opinion, both befriended craftsmen who might help the craftsman in his work
and also regular customers, those who brought prosperity to his shop by their
orders. In the Bithynian epitaph already quoted, the master expected his
pupil to share his new ‘friends’. Apprenticeship would thus be a means for
the adults to control the youth’s work, by demanding from them a return on
investment.47

APPRENTICESHIP AND S OCIAL ADVANCEMENT

As we have seen, the few existing testimonies show that some craftsmen,
rather than develop their own ethics, were eager to integrate into the prevail-
ing social order, appropriating the values and moral codes of the elite, and
placing public emphasis on their reputation, their life of ease, and their many
friends.
Clearly, depending on circumstances, on the reputation of the teacher, and
on the fortune the child owned previously, the benefits of apprenticeship could
be more or less important and fast. It is not, therefore, surprising that manual
crafts had attracted some members of the notability. While gravestones do not
say much about the social origins of the young apprentices, Egyptian papyri
remind us that, although craftsmen’s children were predominant in some
branches, professional training was also attractive for middle-class children,
and sons of veterans or of privileged town residents. One could object that
statutory origins do not always reflect levels of wealth and that these

imitation (in portrait sculpture or in values publicized in inscriptions) are evident. On imitation
processes, see especially Tarde (1979: 205–64); see also Elias (1973).
45
For the pride of craftsmen and their search for social recognition, see Morel (1992: 290–3);
Ferdière (2001: 8).
46
Verboven (2002; 2011: 415–18). See also Veyne (2005: 142–8).
47
Bradley (1991: 116–18); Dixon (1999: 222); Tran (2013: 184–5).
194 Christel Freu

honourable families who placed their offspring within a craftsman’s house had
become impoverished and expected benefits from the work of their children.
However, we do not have any proof for such a scenario, while, on the contrary,
we know that at least one of these middle-class families was well off.48
The fact that apprenticeship could attract children of a good family, not
acquainted with craftsmanship, implies that economic benefits could be
expected from that training. Some gravestones evoke the accomplishments
of men who, having gained their skills in apprenticeship, have succeeded in
reaching the top of their urban or rural community. A beautiful piece of
marble, probably a statue base, inscribed on three sides and found in a large
village of Pisidia, honouring a master ironsmith, very fond of his art, who
executed various difficult works in iron, recalls his training, in the imperial
period, in one of the most famous cities of the eastern Empire, Alexandria:
but he [the son] behaves well and has erected a very precious monument; because
of which his father, the craftsman that much loved his work and, thanks to
Hephaistos’s fire, worked hard to make various objects, a beloved man, is talked
about by the whole of this city, and that of Alexandria, where he learned perfectly
the great laborious art of working iron . . .49
This text shows that initiatory journeys were not reserved for the sons of
important people who became artists, sophists, or physicians; the middle
classes too, in this peaceful imperium Romanum, had the ambition to send
their offspring to the best teachers of technai. At the end of the training the
new technitès had the choice of remaining in the big city or returning to his
native home, where he could become one of the local masters. Craftsmen do
not seem to have preferred one option over the other: perhaps the choice
between moving for professional reasons and settling in their local town may
have been influenced by local traditions. Moreover, the fame of one’s business
could be local, regional, or more distant.
In his biographical dream, where a contest between Sculpture (Ἑρμο-
γλυφικὴ τέχνη) and Liberal Education (Παιδεία) takes place, the Syrian Lucian
of Samosate presents the benefits of apprenticeship of a technè: Paideia
considers the sculptor’s technè to be a synonym for penia—that is, poverty
and anonymity—and thinks that it does not promote a man.50 Technè, on the
contrary, estimates that she will bring a certain comfort and social reputation
to the sculptor:

48
Bergamasco (1995: 112–13). The question arises about the professional future of these
apprentices of good families. On this issue, see Ferdière (2001: 9) and Freu (2011: 30, 36–40).
49
SEG 31.1284: l. 2–7 (Konana, Pisidia; imperial period): ἀλλὰ καλῶς ποιεῖ μνήμην στήσας
πολύτειμον, | ἐξ ἧς τὸν γονέα, τὸν φιλότεχνον ὅλον | καὶ πυρὶ ἐργοπόνον πολυποίκιλον Ἡφαίστοιο |
τεχνείτην, ἐρατὸν πᾶσα πόλις λαλέει | ἥδε κἀλεξάνδρεια, [τῇ] ἐν τέχνην πολύεργον | ἐξέμαθεν
μεγάλην τήν γε σιδηρόδετον. For text and translation, see Merkelbach and Stauber (2001).
50
Lucian, Somn. 11: ὁ νῦν πένης ὁ τοῦ δεῖνος (‘you are a poor man now, son of no one’).
 The Benefits of Apprenticeship in the Roman World 195
If you want . . . to follow me and live with me, first you’ll be generously fed and
you will have strong shoulders and you will be a total stranger to envy. Moreover
you will never have to go abroad, after leaving your country and your relatives,
and it will not be for mere words, either, that everyone will praise you.51

Although the speech is only a school and literary exercise, the opposition
between Paideia and Technè is not without relevance.52 Whereas sculptors
and builders were sometimes itinerant craftsmen, exporting their abilities,
they might also be local craftsmen integrated in powerful associations and
part of local civic life.53 In Lucian’s speech, Sculpture offers to her followers the
comfort of notability. This is not so surprising: we find in fact, in the same
geographical context, a parallel in a Syrian epitaph dating from the beginnings
of the fourth century AD. It comes from the rich region of Apamene, which in
this period was thriving. The inscription is written upon a beautiful grave with
a dromos: the man, looking back on his career, underlines how much profes-
sional training in an unnamed trade made him prosperous, so that he could
buy some lands and stay in the village, near his local deity:
This is what Abedrapsas says, giving thanks (to the God): when I was young, my
ancestral god, god of Arkesilaos, having clearly appeared to me, bestowed upon
me many benefits. For at the age of twenty-five, I was committed to learn the
craft, and after a short period, I acquired this craft, and thanks to his foresight,
I bought myself a small estate, without anybody knowing. And I freed myself
from going down to town. And I was right and I was rightly led.54

He explains his enrichment by the benefits given to him by the local

deity. However, one should be more inclined to think that the mathèsis,
mentioned first, was the real reason for the success of his enterprise. It is
very interesting to see Syrians seeking in the craft an insurance of social

51
Lucian, Somn. 7: εἰ δ’ἐθέλεις . . . ἕπεσθαι δὲ καὶ συνοικεῖν ἐμοί, πρῶτα μὲν θρέψῃ γεννικῶς καὶ
τοὺς ὤμους ἕξεις καρτερούς, φθόνου δὲ παντὸς ἀλλότριος ἔσῃ. Καὶ οὔποτε ἄπει ἐπὶ τὴν ἀλλοδαπὴν,
τὴν πατρίδα καὶ τοὺς οίκείους καταλιπών, οὐδὲ ἐπὶ λόγοις . . . ἐπαινέσονται σε πάντες.
52
Jones (1986: 8–10).
53
For the integration of professional associations in public life of the Roman East, see the now
classic Van Nijf (1997), to which add the more recent article by Drexhage (2007: 168–82,
189–95). Robert (1960: 30–9) sketches Lucian’s first steps in the sculptor’s career and the craft
of marble sculptors in the Roman Empire. The craftsmen from Nicomedia or from Prokonnesos
were famous all over the empire and travelled to export their skills: some had been buried in Asia
Minor, Macedonia, Tripolitana, and Rome; others were involved in local guilds and stayed home (e.g.
IGBulg. 2.674, ὑ]πὲ ̣[ρ] τ̣ῆς̣ συνόδ ̣ου Νει- | κομηδέων λιθο -̣ | ξόων).
54
IGL Syr 4. 1410 (Frīkya in Apamene; AD 325): ταῦτα εὐχαριστῶν λέ<γι> | Ἀ ̣βεδράψας· ἐμοῦ
ἐφ’ ἡλικίας | ὄντος, ὁ πατρῷός μου θεὸς | Ἀρκεσιλάου, δήλως μοι φ[ε]νό- | μενος, ἐν πολλοῖς με
εὐέ|ργησεν. ὡς ἔτων γὰρ κὲ παρεδό- | θην εἰς μάθησιν τέχνη|ς, καὶ διὰ ὀλίγου χρόνου παρέλα- |
{λα}βον {παρέλαβον} τὴν αὐτὴν τέχνην, | καὶ ἔτι διὰ τῆς αὐτοῦ προνοίας | ἐπριάμην αὑτῷ χωρίον,
μηδένος γνόντος, | καὶ ἐλευθέρωσα αὑτὸν μὴ καταβένιν αὑτὸν εἰ<ς> | τὴν πόλιν· κὲ ἐ<γ>ὼ δίκεος
ἤμην, κὲ δικέως ὀδηγήθην.
196 Christel Freu

stability: Lucian offers the same ideal in his work.55 But, in this regard, things
were complex: other regions of the empire were lands of departure for skilled
migrants.56

CONCLUSIO N

Apprenticeship, when it was formally guaranteed by a contract, was thought to

be an essential element of qualification and social promotion inside the
hierarchy of a craft. In increasing the value of the skilled worker trained by
renowned masters, outside the family, it was not really that archaic and even
played a role in the functioning of the labour market. The best pupils acquired
chances to improve their condition and enter the local notability by gaining
friends and customers from their previous master. Even if forms of appren-
ticeship could vary a lot in history, the social and economic benefits of
technical formation seem to be a permanent historical fact.57
My arguments are based on scarce evidence, that is certain. Yet I think that,
as the workshops were similar everywhere, there were similar models of
sociability in the Roman world, especially as the empire carried on. The
small number of mentions of the kind of success that was due to the appren-
ticeship does not allow us, of course, to evaluate this phenomenon. Of all the
pupils of a master, how many were to climb the social ladder and improve
their social condition? It is difficult to tell, but it does seem that there existed
some ‘beloved pupils’. This becomes clear in cases where one of the appren-
tices was charged to write the epitaph or was mentioned among the members
of the family. Moreover, evidence for the hierarchy of shop employees and
workshops indicates clearly that former apprentices had different fates. In
Egypt, as elsewhere, papyri show the existence of at least three groups of
workers in a shop: the head, who was the most specialized of all and supervised

55
See also Lucian, Πατρίδος ἐγκώμιον.
56
e.g. Asia Minor, and especially the provinces of Pontus and Bithynia and Asia. See Pliny,
Ep. 10, 37–42 and 61–2; in letters 37 and 39, Pliny asks Trajan, as a legatus pro praetore of Pontus
and Bithynia, to send him skilful architects in order to help build the Nicomedian aqueduct, the
theatre of Nicea, and the balineum at Claudianopolis. The emperor’s answer (letter 40) is very
interesting: ‘Architecti tibi deesse non possunt. Nulla prouincia non et peritos et ingeniosos
homines habet; modo ne existimes breuius esse ab urbe mitti, cum ex Graecia etiam ad nos uenire
soliti sint’ (‘You cannot lack architects. There is no province where there are no such skilful and
inventive men; unless you think it quicker to have one sent from Rome, whereas they usually
come to us from Greece’). See also Robert (1960: 30–9); Drexhage (2007: 181); Ruffing (2008).
57
As was suggested by one of the anonymous reviewers, the Roman period was not an
isolated case for the search of good masters: in the early modern period, for instance, well-
reputed masters often charged premiums from apprentices who wished to learn from them.
 The Benefits of Apprenticeship in the Roman World 197

the labour; specialized employees, who were former apprentices and earned
different types of wages; and unskilled workforce.58 Interestingly, the crafts-
men who recalled their apprenticeship in their inscriptions called themselves
technitai, skilled workers and possessors of the craft: apprenticeship was
obviously very helpful to promote oneself.59
Thus, whatever the apprentice’s former condition was, and regardless of
whether he worked for himself or became a specialized employee, his earnings
increased.60 At the same time, apprenticeship also had important benefits for
the economy as a whole because of the transmission and even the improve-
ment of technical skills.61

ACKNOWLEDGEMENTS

This chapter has benefited from the useful remarks of Ilias Arnaoutoglou and
Patrice Hamon on Greek inscriptions, and of A. Cristofori and A. Ripoll for
the whole. The anonymous reviewers and the editors, Miko Flohr and Andrew
Wilson, also improved it with their readings, for which I am very grateful to
them. Errors remain the sole responsibility of the author.

REFERENCES

Bergamasco, M. (1995). ‘Le διδασκαλικαί nella ricerca attuale’, Aegyptus 75: 95–167.
Bradley, K. R. (1991). ‘Child Labour in the Roman World’, in K. R. Bradley (ed.),
Discovering the Roman Family: Studies in Roman Social History. Oxford, 103–24.
Cribiore, R. (1996). Writing, Teachers and Students in Graeco-Roman Egypt. Atlanta.
Dixon, S. (1999). ‘The Circulation of Children in Roman Society’, in M. Corbier (ed.),
Adoption et fosterage. Paris, 217–30.
Donderer, M. (1989). Die Mosaizisten der Antike und ihre wirtschaftliche und soziale
Stellung. Erlangen.
Donderer, M. (1996). Die Architekten der späten römischen Republik und der Kaiserzeit:
epigraphische Zeugnisse. Erlangen.
Drexhage, H.-W. (2007). Wirtschaftspolitik und Wirtschaft in der römischen Provinz
Asia in der Zeit von Augustus bis zum Regierungsantritt Diokletians. Berlin.

58
Freu (2011); Tran (2013).
59
It is the case in the stele from Bithynia (TAM 4, 132) and in the Pisidian text (SEG
31, 1284).
60
Morel (1992: 280–2), has a phrase that seems to me a good concluding word: ‘un spécialiste
vaut de l’or’.
61
Tran (2013: 175–85).
198 Christel Freu
Duncan-Jones, R. (1974). The Economy of the Roman Empire: Quantitative Studies.
Cambridge.
Elias, N. (1973). La Civilisation des mœurs. Paris.
Ferdière, A. (2001). ‘La “Distance critique”: Artisans et artisanat dans l’Antiquité
romaine et en particulier en Gaule’, Les Petits Cahiers d’Anatole 1: 2–31.
Feyel, C. (2006). Les Artisans dans les sanctuaires grecs aux époques classique et
hellénistique à travers la documentation financière en Grèce (BEFAR, 318). Paris.
Flohr, M. (2011). ‘Exploring the Limits of Skilled Craftsmanship: The fullonicae of
Roman Italy’, in N. Tran and N. Monteix (eds), Les Savoirs professionnels des
hommes de métier romains. Naples, 87–100.
Flohr, M. (2013). The World of the Fullo: Work, Economy, and Society in Roman Italy.
(Oxford Studies on Roman Economy). Oxford.
Frasca, R. (1994). Mestieri e professioni a Roma: Una storia dell’educazione. Florence.
Frasca, R. (1999). ‘Il profilo sociale e professioniale del maestro di scuola e del maestro
d’arte tra repubblica e alto impero’, in G. Firpo and G. Zecchini (eds), Magister:
Aspetti culturali e istituzionali. Turin, 129–58.
Freu, C. (2011). ‘Apprendre et exercer un métier dans l’Égypte romaine (Ier–VIe siècles
ap. J.-C.)’, in N. Tran and N. Monteix (eds), Les Savoirs professionnels des hommes
de métier romains. Naples, 27–40.
Gomez Pallarès, J. (2000). ‘Nueva lectura, con commentario, del epitafio métrico de
Julio Estatuto (R.I.T., 447)’, in G. Paci (ed.), Ἐπιγραφαί: Miscellanea epigrafica in
onore di Lidio Gasperini. Rome, 417–28.
Goulet-Cazé, M.-O. (1982). ‘L’Arrière-plan scolaire de la Vie de Plotin’, in L. Brisson
et al. (eds), Porphyre, ‘La Vie de Plotin’, I. Travaux préliminaires et index grec
complet. Paris, 231–327.
Hawkins, C. (2012). ‘Manufacturing’, in W. Scheidel (ed.), The Cambridge Companion
to the Roman Economy. Cambridge: 175–94.
Jones, C. P. (1986). Culture and Society in Lucian. London.
Joshel, S. R. (1992). Work, Identity, and Legal Status at Rome: A Study of the
Occupational Inscriptions. Norman, OK, and London.
Kehoe, D. P. (2007). ‘The Early Empire: Production’, in W. Scheidel, I. Morris, and
R. Saller (eds), The Cambridge Economic History of the Greco-Roman World.
Cambridge, 543–69.
Labarre, G. and Le Dinahet, M.-Th. (1996). ‘Les Métiers du textile en Asie Mineure de
l’époque hellénistique à l’époque impériale’, in Aspects de l’artisanat du textile dans
le monde méditerranéen (Égypte, Grèce, monde romain). Lyons, 49–115.
Lattimore, R. (1962). Themes in Greek and Latin Epitaphs. Urbana.
Marcadé, J. (1953). Recueil des signatures de sculpteurs grecs. Paris.
Marrou, H.-I. (1948). Histoire de l’éducation dans l’Antiquité, i. Le Monde grec. Paris.
Mayer, E. (2012). The Ancient Middle Classes: Urban Life and Aesthetics in the Roman
Empire. Cambridge, MA, and London.
Merkelbach, R., and Stauber, J. (2001). Steinepigramme aus dem griechischen Osten, III.
Leipzig.
Monteix, N. (2011). ‘De “l’artisanat” aux métiers’, in N. Tran and N. Monteix (eds), Les
Savoirs professionnels des hommes de métier romains. Naples, 7–26.
Morel, J.-P. (1992). ‘L’Artisan’, in A. Giardina (ed.), L’Homme romain. Paris, 267–302.
 The Benefits of Apprenticeship in the Roman World 199
Mrozek, S. (1975). Prix et rémunération dans l’Occident romain (31 av. n.è.–250 de
n.è.). Gdansk.
Muller-Dufeu, M. (2002). La Sculpture grecque: Sources littéraires et épigraphiques.
Paris.
Nani, T. G. (1943–4). ‘ΘΡEΠΤˇΙ’, Epigraphica V–VI: 45–84.
Puech, B. (2002). Orateurs et sophistes grecs dans les inscriptions d’époque impériale.
Paris.
Puech, B. (2012). ‘Transmission de pouvoir et transmission de valeurs: Les Dynasties
d’intellectuels en Orient du IIIe au Ve siècle’, in Chr. Badel et Chr. Settipani (eds),
Les Stratégies familiales dans l’Antiquité tardive (IIIe–VIe siècle). Paris, 301–20.
Robert, L. (1936). Collection Froehner: Inscriptions grecques I. Paris.
Robert, L. (1960). ‘Épitaphes et acclamations byzantines à Corinthe’, Hellenica 11–12:
21–39.
Ruffing, K. (2008). Die berufliche Spezialisierung in Handel und Handwerk: Untersuchun-
gen zu ihrer Entwicklung und zur ihren Bedingungen in der römischen Kaiserzeit im
östlichen Mittelmeerraum auf der Grundlage griechischer Inschriften und Papyri. 2 vols.
Rahden.
Salvaterra, C. (2006). ‘Labour and Identity in the Roman World: Italian Historiography
during the Last Two Decades’, in B. Waaldijk (ed.), Professions and Social Identity. New
European Historical Research on Work, Gender and Society. Pisa, 15–38.
Schulz-Falkenthal, H. (1972). ‘Zur Lehrlingsausbildung in der römischen Antike—
discipuli und discentes’, Klio 54: 193–212.
Tarde, G. de (1979). Les Lois de l’imitation: Étude sociologique. Paris and Geneva.
Tran, N. (2010). ‘L’Apprentissage et le statut de travail des artisans en Gaule romaine’, in
Aspects de l’artisanat en milieu urbain: Gaule et Occident romain, Revue archéologique
de l’Est, 28e supplément. Dijon, 195–200.
Tran, N. (2011). ‘Les Gens de métier romains: Savoirs professionnels et supériorités
plébéiennes’, in N. Tran and N. Monteix (eds), Les Savoirs professionnels des
hommes de métier romains. Naples, 119–33.
Tran, N. (2013). Dominus tabernae: Le Statut de travail des artisans et des commerçants de
l’Occident romain (BEFAR, 360). Rome.
Van Nijf, O. M. (1997). The Civic World of Professional Associations in the Roman
East. Amsterdam.
Verboven, K. (2002). The Economy of Friends: Economic Aspects of ‘Amicitia’ and
Patronage in the Late Republic. Brussels.
Verboven, K. (2011). ‘Friendship among the Romans’, in M. Peachin (ed.), The Oxford
Handbook of Social Relations in the Roman World. Oxford, 404–21.
Veyne, P. (2005). ‘La Ville de Rome et la “plèbe moyenne” ’, in P. Veyne, L’Empire
gréco-romain. Paris, 117–61.
Wallace-Hadrill, A. (1994). Houses and Society in Pompeii and Herculaneum. Princeton.
Wallace-Hadrill, A. (2008). Rome’s Cultural Revolution. Cambridge.
Wilson, A. I. (2008). ‘Large-Scale Manufacturing, Standardization, and Trade’, in
J. P. Oleson (ed.), The Oxford Handbook of Engineering and Technology in the
Classical World. Oxford, 393–418.
 9

Women, Trade, and Production in Urban

Centres of Roman Italy
Lena Larsson Lovén

This chapter discusses the involvement of women in production and trade in

urban centres of Roman Italy and the possibilities of identifying women’s
economic and professional roles in manufacturing and retail. For modern
scholars, working women belong to the most invisible groups in the Roman
world: literary texts do not offer detailed information about female workers
and working lives and neither does archaeology. While there is epigraphic and
iconographic evidence for female working lives, earlier studies have demon-
strated very clearly that there is an extensive gap between the documentation
of male work, on the one hand, and that of female work, on the other. The
marginalization of women is obvious in any category of evidence for work and
occupations in the Roman world. The scarcity of evidence has meant that
female working lives were long understudied. Largely neglected by modern
scholars, working women appear to us as a marginal group.1
The present chapter will focus primarily on how female occupations and
professional identities were represented in funerary epigraphy, and discusses
what this evidence can tell about female economic roles and gender structures
in Roman society. The majority of the texts that will be discussed in this
chapter come from the Corpus Inscriptionum Latinarum, and there is a certain
emphasis on the city of Rome, but evidence from some other Italian cities will
be discussed as well. The intention here is not to discuss an extensive list of
female jobs (as has been done by some in the past), but rather to analyse how
female professional identities have been documented and represented, and
how socio-economic roles of women can be interpreted in relation to over-
arching Roman gender hierarchies.

1
For an overview of previous research on work and gender issues, from the last decades of the
twentieth century and the early twenty-first century, see Salvaterra (2006); see Knapp (2011) for
a recent study of marginal or invisible groups in the Roman world.
 Women, Trade, and Production 201

APPROACHING THE WORKING L IVES OF WOMEN

Epigraphy has been used unevenly in studies on aspects of Roman economy

and more rarely with the specific purpose of discussing the economic roles of
women.2 Ground-breaking work was done by Susan Treggiari in the 1970s,
when she published two articles focusing on female occupations.3 Both are
based on epigraphic evidence from the city of Rome in the imperial period.
Treggiari’s work can be seen as part of the 1970s trend towards greater interest
in women’s history in general.4 She discussed an extensive number of female
jobs in the imperial household and in the domestic staff of wealthy families in
the city of Rome as well as among non-domestic craftspeople and traders.
These studies showed how women’s work was represented in epigraphy, but
they also made the gender imbalance in the documentation of male and female
work in the epigraphic record very obvious: not surprisingly, the inscriptions
commemorating men’s work comprised a much wider range of job titles than
those commemorating female work. Treggiari made several important con-
clusions about male and female jobs, and her work was an eye-opener to
many, but it did not engage in a wider discussion of the economic and social
parameters of women’s work. However, Treggiari’s work is still a point of
departure for further studies of the occupational roles of Roman women.
After the 1970s, the history of women was gradually and partly replaced by
gender studies; in the search of women or ‘the female’ in history, new
methodologies developed. These entailed new ways of reading ancient sources
and started from the notion that gaps and silences are as significant as what is
actually in the sources. For instance, Sandra Joshel, in her 1992 study on work,
identity, and legal status in the city of Rome, raised the issue of ‘listening to
silence’ right at the beginning of her argument.5 Suzanne Dixon made a
similar point in her paper on the epigraphy of textile production aptly called
‘How do you count them if they’re not there?’ Dixon argued that, in addition
to what is in the sources, it is also vital to look for what is not there.6 This
methodological approach starts from the issue of exclusion.7 The exclusion of
information, such as on women’s work, Dixon says, can be either deliberate or
casual, but in both cases it is likely to have some significance.8 Recent work by

2 3
Edmondson (2015: 671–3). Treggiari (1975, 1976).
4
Treggiari’s paper on jobs in the household of Livia was published within a year of Sarah
Pomeroy’s Goddesses, Whores, Wives and Slaves: Women in Classical Antiquity.
5
Joshel (1992: 3–23). ‘Listening to silence’ is again the argument in Amy Richlin’s recent
monograph on women in the Roman world (Arguments with Silence). Richlin (2014: esp. 1–35).
6 7
Dixon (2000–1). Dixon (2000–1: 10).
8
In her 2001 study on Reading Roman Women, Dixon presented a wider methodological
discussion of how to read ancient genres in general, not only epigraphic sources, and discussed
how male and especially female social identities were constructed through work. Dixon (2001:
16–25, 113–33).
202 Lena Larsson Lovén

Roth, Groen-Vallinga, and Holleran builds further upon these observations,

also emphasizing the gap between the picture suggested by the evidence and
historical reality.9 In short, recent scholarship has emphasized the need for a
deepened rereading of various genres and types of evidence, and for alertness
to gaps, silences, and absent information. In this way, it becomes possible to
reach not only a reliable, but also a more complex and nuanced, picture of
female working lives, and to give women their due place in debates about
craftsmen and traders in the Roman world.

JOB COMMEMORATION IN THE I MPERIAL

HOUSEHOLD

In her paper on jobs in the household of Livia (1975), Treggiari used funerary
inscriptions from the Monumentum Liviae, a columbarium on the Via Appia,
as the basis for a discussion of jobs in the imperial household.10 The men and
women buried here were slaves and freedmen and freedwomen of the early
imperial city household; more than 1,100 persons are estimated to have been
buried in this complex between the later part of the reign of Augustus and
shortly after the deification of Livia in AD 41. It is named after Livia, who
appeared most frequently as the owner of slaves in the inscriptions from this
sepulchral complex. Some of the slaves and freed persons were commemor-
ated with a job title, and this information reveals part of the structure of this
extensive household in terms of job specializations and work hierarchies. It
also gives an indication of the distribution of male and female jobs. Not all jobs
that could be expected to occur in this type of household are represented in the
inscriptions from the Monumentum Liviae, but the dissimilarities in the
frequency of male and female jobs are still clear.11
The epigraphic evidence from the Monumentum Liviae showed a variety of
professional specialists divided over several occupational sectors, including
administrators, an extensive domestic staff, and craftsmen. Among the crafts-
men of the imperial household was a shoemaker and a few occupations
involved in dealing with luxury goods such as some goldsmiths (aurifices), a
silversmith (argentarius), a margaritarius (a dealer or trader in pearls), and,

9
Roth (2007: 1–24); Groen-Vallinga (2013: 309); Holleran (2013: 328).
10
The Monumentum Liviae no longer exists but was excavated in the early eighteenth
century: Treggiari (1975: 48–9).
11
It is important to take into consideration that the inscriptions do not give a full picture but
may be reflecting only part of the staff and that they do not record the entire staff at one moment.
Cf. Treggiari (1975: 57). See also Joshel (1992: 100–5) for a discussion of slaves, freedmen/
freedwomen, and occupational titles in the Monumentum Liviae.
 Women, Trade, and Production 203

possibly, a gilder (inauratur) who belonged to one of Livia’s freedmen.12 One

inscription documents a lanipendus, who would be responsible for weighing
the wool for the spinners—who were normally women.13 The occurrence of a
lanipendus may point to textile manufacturing, but, as there is only one single
inscription, this does not in itself mean that textiles were produced in the
imperial household: such a conclusion would require the presence of other
occupational categories as well, such as spinners and weavers, who represent
the central stages in the production chain. Yet no inscriptions mentioning
these occupations are known from this columbarium.
In most job categories from the Monumentum Liviae, women are not
attested, but some female workers appear in a group of inscriptions linked
to textile work, as clothes-menders (sarcinatrices).14 Only a fifth of the women
from the Monumentum Liviae are commemorated with a job title. Treggiari
noted that, on the whole, ‘one striking factor in the job structure is the low
proportion of women’.15 Thus, the sarcinatrix appears as one of the few female
jobs documented by several texts from the household of Livia Some other
attestations of sarcinatrices are known from Rome, but they come from
outside the imperial household.16 A few examples come from other cities
elsewhere in central and northern Italy.17 Some of the sarcinatrices from
Monumentum Liviae had been manumitted, but the majority of the women
commemorated with this job title were slaves.18

WOMEN I N THE PRO DUCTIO N AND TRADE OF

LUXURY GOODS

Luxury products, such as jewellery, silk clothes, gems, pearls, gold, purple,
perfumes, and more, were available in many Roman cities, and a number of
occupational titles exist that reflect their manufacturing and retail. Some of

12
Shoemaker: CIL 6.3939; Treggiari (1975: 54); aurifices: CIL 6.3927, 3945, 3949; argentarius:
CIL 6.8727; margaritarius: CIL 5.3981; gilder: CIL 6.3928.
13
CIL 6.3976.
14
CIL 6.3988, 4029–31, 5357, 8903, 9038. There is also a possible male mender (sarcinator):
CIL 6.4028. In Rome, evidence for sarcinatrices is found outside the imperial household too
(Treggiari 1976: 85). The majority of the sarcinatrices are female slaves.
15
Treggiari (1975: 58). The total number of women in Monumentum Liviae will, however, be
higher, as many were commemorated after being freed and married, with no mention of a job.
See also Treggiari (1976: 94).
16
See Treggiari (1976: 85) for inscriptions of sarcinatrices from Rome, from the household of
Livia, other imperial women, and from the Statilii family. According to Eichenauer, the total
number of inscriptions of sarcinatrices is 26; cf. Eichenauer (1988: 94).
17
CIL 5.2542 (Este), 5.2881 (Padua), and 11.5437 (Assisi); Eichenauer (1988: 96).
18
Eichenaurer (1988: 94, n. 4).
204 Lena Larsson Lovén

these appear in the inscriptions from the Monumentum Liviae, generally

referring to male workers. In general, epigraphically attested dealers of luxury
products are men, but women have been documented in this branch of the
urban economy as well. Female perfume-sellers, unguentariae, have been
attested both in Rome and in Puteoli.19 In Rome, luxury products were sold
along the Via Sacra, as is attested by several inscriptions, including the
combination of occupational titles and the name of the street.20 An example
is an inscription listing a group of five persons who were jewellers (gemmarii).
The first of these is a woman named Babbia Asia, while the others are male.21
Other inscriptions from the neighbourhood of the Via Sacra attest traders in
purple, which was a symbol of status throughout antiquity.22 The most
frequently recurring job title in epigraphy linked to purple products is the
purpurarius, variously interpreted as either ‘producer or dealer of purple
dyestuff ’ or ‘producer or dealer of purple products’.23 Whatever its actual
meaning, the term has a wide distribution in urban centres in Italy from the
south to the northern regions. In addition to the evidence from Rome,
purpurarii have been found in urban centres such as Puteoli, Capua, Parma,
Chiusi, Sarsina, and Aquileia. Numerically, there is a male dominance.24
The most direct evidence that women were involved in purple trade and
production is an inscription from Rome that documents a group of freedwomen
who may all have been in the trade of purple products.25 In addition to this,
there are several inscriptions from Rome mentioning purpurarii.26 In some of
them, both men and women appear, but the occupational title is primarily
linked to men. These inscriptions can be used to highlight the difficulties in
reading the evidence concerning female involvement in Roman trade, produc-
tion, and the Roman labour market in general.27 One example commemorates a
couple from the Veturii family, well known for its involvement in the textile
trade. The couple is a man named D. Veturius Atticus, described as purpurarius
de Vico Iugario, and a woman Veturia Tryphea (Fig. 9.1).28 The inscription does

19
CIL 6.10006 refers to a woman in Rome commemorated by her husband after thirty years
of marriage; CIL 10.1965 refers to a woman from Puteoli who died at the age of 71; Treggiari
(1979a: 71).
20
For a more detailed discussion of the commercial activities in the Forum district, see Papi
(2002); for shopping in Rome in general, see Holleran (2012).
21
CIL 6.9435. All five persons were ex-slaves and two of them were manumitted by a woman,
see Treggiari (1979a: 67); Papi (2002: 58); Holleran (2013: 314).
22
For a detailed discussion of purple as a long-lasting symbol in antiquity, see Reinhold (1970).
23
For a discussion and various interpretations of purpurarius see von Petrikovits (1981: 100);
Gregori (1994: 742); Pelletier (1996: 133). See also Hughes on the discussion of purpurarii as
producers/dealers of purple products other than textiles: Hughes (2007: 89–91).
24
CIL 5.1044 (Aquileia); 10.1952 (Puteoli); 10.3973 (Capua); 11.1069a (Parma); 11.2136
(Chiusi); 11.6604 (Sarsina).
25 26
CIL 6.9846, purpurar(ii). Gardner (1986: 238–9).
27
Joshel (1992: esp. ch. 1); Dixon (2000–1).
28
NSA 1922, 144 = AE 1923, 59; from the Via Praenestina; cf. Dixon (2001: 118–19).
 Women, Trade, and Production 205

Fig. 9.1. Purple-dealers from the Vicus Iugarius (AE 1923, 59). Photo: Miko Flohr.

not specify whether the woman also worked in the purple trade, but it is possible
that both the man and the woman worked in a shop in the Vicus Iugarius
trading purple products. The same can be said about an inscription of three
freed people, one man—another purpurarius, L. Plutius Eros, who worked in
the Vicus Tuscus—and two women, named Plutia Auge and Veturia Attica.29
The man and one of the women are from the family of the Plutii, while the
other woman is from the Veturii family. One of the women, Plutia Auge,
commissioned the monument for all three of them, but the relationships
between these individuals and between the two families are not altogether
clear.30 In this case, too, the occupation of the women is not specified, but
both women may have been involved in the trade of purple products.31 The
inscriptions of purpurarii may serve as examples of how ‘reading the silence’ can
work. When an occupational group of mixed sex occurs in an inscription with a
job title in the masculine form, it is obvious that it refers to male work, but it
does not exclude that women were involved in the business as well. If, in
analysing the inscriptions with such job titles, women are more regularly
considered, this still does not constitute direct evidence for female work, but it
does offer a better understanding of women’s occupational possibilities and of
their roles in urban economies and the labour market.
More luxury textiles were cloth and clothes made of silk, which was
imported from the east. In the early imperial period, silk clothes became
highly fashionable in Rome, not just for women but also for men, which
probably caused an increasing demand for silk products.32 Most known

29 30 31
CIL 14.2433. Gardner (1986: 238). Eichenauer (1988: 98).
32
See Hildebrandt (2013) for a recent chapter on silk clothing in Rome.
206 Lena Larsson Lovén

Latin inscriptions referring to people involved in the silk trade come from the
city of Rome or urban centres in its vicinity.33 Again, most of these inscrip-
tions document men, sericarii, but there is evidence for at least one female silk-
worker, a woman named Thymele, who was commemorated as a sericaria.34
Further specialists in luxury production are attested by two funerary in-
scriptions from Rome. Dating to the fourth century AD, both commemorate
women who appear to have been involved in highly specialized jobs related to
the textile economy. One epitaph documents a girl who was an aurinetrix, a
spinner of gold thread; the other records an aurivestrix.35 These are very
specific job titles reflecting a high degree of specialization in the manufactur-
ing economy of the Roman metropolis. However, as both are attested only
once, it is hard to tell how common such jobs may have been, though it is
significant that they come from the city of Rome, where demand for luxury
products is likely to have been exceptionally high.36 Sellia Epyre, the aurives-
trix from the Via Sacra, was possibly a specialist in gold embroidery; her case
is another example of the trade of luxury products in the commercial district
in and around the Via Sacra (Fig. 9.2). For the aurinetrix, the spinner of
gold thread, there is no indication in the inscription of where her work
took place, but it is worth noting that the inscription is an epitaph to the
memory of a girl named Viccentia, dulcissima filia, who died at the age of only
9 years and 9 months.37 The girl was commemorated by her parents; while the
occupation of these people is not mentioned in the text, it is possible that the
girl worked in a family business where she was trained by her parents from an
early age.38
As demonstrated by the examples discussed so far, inscriptions document-
ing the trade of luxury goods are found in cities throughout Roman Italy, but
the lion’s share of the evidence is concentrated in the Roman metropolis and
its vicinity. The variation in job titles related to luxury trades reflects the

33
Cf. two epigraphic instances from Tivoli commemorating the same man, M. Nummius
Proculus, who was a sericarius; CIL 14.3711–12, and two from Gabii of Aulus Plutius Epaphro-
ditus, a negotiator sericarius; CIL 14.2793 (= ILS 5449), CIL 14.2812 (= ILS 7601). A Greek
inscription mentioning a silk specialist comes from Naples (IG 14.785). For examples from urban
centres outside Italy, cf. also Athens (IG 3.3513).
34
CIL 6.9892, Thymele was the slave of a woman named Marcella.
35
Aurinetrix: CIL 6.9213. Aurivestrix: CIL 6.9214. The occupational title aurivestrix has also
been interpreted as ‘dressmaker in gold’, see Lefkowitz and Fant (1992: 219, no. 321, n. 21); Gleba
(2008: 63); Holleran (2013: 315).
36
Edmondson (2015: 673).
37
The age of death of the girl and the occupational title linked to her name raise questions
about Roman child labour and at about what age a child might start specialist training for a job,
but these are issues beyond the scope of the present chapter. Cf. Bradley (1991); Groen-Vallinga
(2013: 305–6).
38
On the informal training of craftsmen, see Freu, Chapter 8, this volume.
 Women, Trade, and Production 207

Fig. 9.2. Sellia Epyre, aurivestrix from the Via Sacra, Rome (CIL 6.9214). Photo: Roger
B. Ulrich.

variety of products that were traded and consumed mainly in the capital—the
primary market for a conspicuous consumption—and the evidence clearly
shows that women played a role in this process as well.

WOME N IN TH E P RODUCTI ON AND

TRADE OF EVERYDAY CONSUMER GOODS:
THE CASE OF TEXTILES

If one turns to the evidence for the production and trade of more mundane
goods than pearls, perfumes, and silk clothes, the production of everyday
textiles may be used as an example. Textile production formed a vital part of
the Roman economy, with a large number of people involved as producers,
traders, and consumers. Textiles were professionally manufactured and
traded, not just in Roman Italy but all over the empire, but the evidence
available nowadays is scattered and limited, and has until recently been
studied only to a very limited extent.39 Nevertheless, the economic importance
of textile production should not be underestimated. Through information in

39
See Roth (2007: 53–87) for a discussion of evidence for textile production and especially
female slave labour.
208 Lena Larsson Lovén

some literary sources we can get a glimpse of the development and the scale
of textile production already in the Republican period. Livy, for instance,
reports that towards the end of the Second Punic War, in 204 BC, Tiberius
Claudius Nero, who was praetor in Sardinia, had a cargo of 1,200 togas and
12,000 tunics—a huge amount of garments—sent to supply the Roman legions
serving in North Africa under Scipio the Elder.40 Less than four decades later,
in 169 BC, an even larger amount of garments, 6,000 togas and 30,000 tunics,
was contracted by the urban praetor C. Sulpicius to be sent from Rome for
army supply in Macedonia.41 Cato the Elder advised wealthy estate-owners to
buy cheap clothing in Rome for their slaves, as this would be more economical
than having clothes for slaves made on the estate.42 Cato does not specify the
scale of the production, but his writings coincide well, chronologically, with
the episodes mentioned by Livy and reflect the existence—by that time—of a
market for ready-made clothes and possibly other textile products in Rome
and other urban centres.
The scale of the textile economy as well as the involvement of female labour
has been debated and variously interpreted over the years. In the episodes
mentioned above, nothing is said about who produced the textiles contracted
for the Roman army or the cheap clothing for slaves. Literary sources give us the
names of various textile products, but, again, the best information comes from
funerary epigraphy, where a range of occupations related to various stages of
textile production and trade. The occupational inscriptions document both men
and women in domestic service as well as non-domestic workers. The greatest
concentration of textile-related occupations in the city of Rome comes from
inscriptions of the families of the Statilii and the Veturii. In the sample of the
Statilii family, there is, for example, evidence of a group of spinners, quasillariae,
who were all women, and a couple of inscriptions of weavers: two male textores
and one female textrix.43 If we consider that spinning and weaving were two of
the central stages in the chain of textile production, the epigraphic evidence of
these work moments is very modest. Some more wool-weighers, lanipendi,
occur, and interestingly there are also some women in this supervisory position
as lanipendae.44 As already mentioned, a lanipendus or lanipenda was primarily
responsible for weighing the daily amount of wool to be given to the spinners.
Yet this person could probably also have been in charge of the whole sequence
of textile manufacturing in households where the entire process was performed.

40
Livy 29.36. Two years later, in 202 BC, the same Tiberius Claudius Nero was elected as
Roman consul.
41 42
Livy 44.16. Cato Maior, De Agr. 1.135.
43
Textile-workers from the Monumentum Statiliorum: quasillariae, CIL 6.6339–46; textores,
CIL 6.6360–1, and a female weaver, textrix, CIL 6.6363.
44
Lanipendii: CIL 6.3976–7, 6300, 8870, 9495, 37755. Lanipendae: CIL 6.9496–8, 34273,
37721, AE 1969–70, 49. From outside the city of Rome come CIL 9.3157 (Corfinio) and 4350
(Amiternum).
 Women, Trade, and Production 209

It is unusual to find a woman in a position as manager or supervisor, and the

position as lanipenda seems to have been restricted to households outside both
the wealthiest families and the imperial circle.45 More clothes-menders, sarcina-
trices, appear from wealthy Roman households, and, as in the case of the
Monumentum Liviae, female workers dominate this occupation.
Occupations related to the stages in production of a finished textile product
were the vestifici and vestarii. Finished cloth could be made into new garments
by male and female tailors—vestifici and vestificae. These people are found
among the household staff of private employers.46 Vestiarii were non-domestic
traders, and evidence of male vestiarii is found in many urban centres of Roman
Italy, though the majority comes from Rome.47 A possible female vestiaria
appears in an inscription together with a group of men, all liberti who seem to
have specialized in the business of delicate cloth/clothes: they are described as
vestiarii tenuarii.48
The examples of women’s work discussed so far have been defined in
epigraphy by a number of job titles, with inscriptions linking women to a
job title either on an occupational basis or as part of a group. Most of the
epigraphically attested female job titles have parallel titles for men, and, in
most occupations, references to men outnumber those to women. One excep-
tion appears to be the clothes-menders, the sarcinatrices, where the majority
consists of women. The only exclusively female group appears to be the
spinners, the quasillariae, with no obvious equivalent male occupational title.
Outside the textile economy, the same pattern, with more men than women,
can be found in most other areas of production as well.49 Shoe-making may be
added as yet another example of an artisan’s job pursued by several men and
by some women. Most inscriptions referring to sutores (cobblers) refer to men.
From Milan comes an inscription mentioning a sutor caligarius, a specialist in
making boots, and from Rome comes the well-known stele of the cobbler
Gaius Julius Helius, with an inscription listing his occupation and a portrait
showing his professional activities.50 However, a funerary monument from

45
Treggiari (1976: 83).
46
Vestificae: CIL 6.5206, 9744, 9980. According to Treggiari, there are four vestifici in
domestic service: CIL 6.7476, 8544, 9979, and 37724; Treggiari (1976: 85).
47
A sample of inscriptions of vestiarii from Roman Italy shows the diffusion from southern
urban centres to the north: CIL 4.3130 (Pompeii); 9.1712 (Benevento); 10.3959 (Capua); 1.1216;
6.4044, 4476, 7378, 7379, 9962–6, and 9970–6 (Rome), 11.6839 (Bologna), 11.869, 6962a
(Modena), 5.3460 (Verona), 5.774 (Aquileia).
48
CIL 6.33920. In addition to the women, and men, involved in the production of textiles in
wealthy aristocratic households, there were also those who took care of the wardrobe of the
master or the mistress, the vestiplica (vestipica). For vestipliciae, see CIL 6.9901, 33393, 33395,
37825. One example of a vestiplicia from outside Rome is found in CIL 9.3318, from Castelvec-
chio Subrego. For the male equivalent, a vestip(l)icius, see CIL 6.7301, 8558–60, 9981.
49
Treggiari (1975, 1975, 1979a); Eichenauer (1988); Groen-Vallinga (2013).
50
Milan: CIL 5.5919. Rome: CIL 6.33914. Cf. Zimmer (1982: 137–8).
210 Lena Larsson Lovén

Ostia commemorating a woman named Septimia Stratonice has a partially

preserved inscription and an image representing a seated woman, dressed in a
simple belted tunic and holding a shoe last in her right hand (Fig. 9.3).51 The
preserved part of the inscription does not provide the job title (sutrix), but the
image gives a clear but rare indication of a woman in her professional role as
shoe-maker.52

MALE AND F EMALE J OB TITLES IN THE SAME

LINE OF BUSINESS

A majority of those men and women who were commemorated with a job title
were slaves and freed persons—people to whom work constituted a central
part of their social identity. Sharing work for people in these groups was
often a reason for commissioning a shared burial in the form of an epitaph
or a memorial. In funerary epigraphy and iconography we meet various

Fig. 9.3. The shoemaker Septimia Stratonice from Ostia (CIL 14.4698).

51
CIL 14.4698.
52
Kampen (1981: 64–9, fig. 47). In the iconography of work the same tendency with fewer
women commemorated at work is obvious. Holleran (2013: 218–19).
 Women, Trade, and Production 211

constellations of people, such as legally married couples, contubernales—slave

couples that were not legally married—family groups, slaves, and conliberti,
who refer to their work in words, in iconography, or through both text and
image. Artisans’ epitaphs including freedmen or freedwomen often indicate a
working relation between them.53 However, only rarely does it occur that male
and female workers are attested with the same job title in one inscription. Yet a
few examples do exist, among them two from northern Italy.54 One example is
an inscription from Milan that documents a man and a woman involved in the
trade of linen textiles and explicitly mentions the occupation of both: linarius
and linaria.55 While there is more evidence of linarii, this particular inscrip-
tion represents the only known instance of the job title in the feminine form.56
The inscription was commissioned by the man, the linarius Gaius Cassius
Sopater, and it commemorates not only the man himself but three freed-
women of his as well, all bearing the name Cassia. One of them, the linaria
Cassia Domestica, was also his wife. From Turin is an inscription of nail-
makers, a female clavaria (nail-maker), Cornelia Venusta, and a clavarius, a
freeborn, P. Aebutius.57
Further examples of couples in the same line of business are from Rome.
Two texts document producers of gold leaf, each explicitly mentioning a man
and a woman as brattiarius and brattiaria.58 Both men and one of the women
were clearly former slaves, since freed, while the other woman, Fulvia Melema,
is of unclear status.59 A comparable form of closely related male and female
work documented in inscriptions concerns men and women doing different
jobs but in the same sector of production. An example is an inscription from
Rome of a slave couple where both the man and the woman were involved in
textile production. The woman, Musa, was a spinner (quasillaria), while the
man, Cratinus, was a lanipendus.60
The examples discussed show women in a number of occupational roles
among others as producers or traders of jewellery and perfume, dealers in
purple, silk, or linens, as spinner, weaver, wool-weigher, clothes-mender,
seamstress, tailor, shoe-maker, nail-maker, and more. Women also worked
as salespersons of various goods, as is demonstrated by both epigraphy and
iconography.61

53
Joshel (1992: 138).
54 55
Huttunen (1974: 48–9); Treggiari (1976: 98); Dixon (2001: 118). CIL 5.5923.
56
One inscription from Spain attests a possible linaria (lintearia) from Tarragona, Spain: CIL
2.4318a.
57 58
CIL 5.7320; Holleran (2013: 315). CIL 6.9211 and 6939; Treggiari (1979a: 66–7).
59
In CIL 6.9211 the man Gaius Fulcinius Hermeros was clearly a freedman, while there is no
clear indication of manumitted status for the woman, Fulvia Melema; Holleran (2013: 315).
60
CIL 6.9495. For more examples of other combinations of male and female occupations in
the same inscription, although not always related jobs, see Treggiari (1976: 98).
61
See Kampen (1982, 1985) for saleswomen in iconography; Holleran (2013: 321–5).
212 Lena Larsson Lovén

W O M EN’ S WORK: READINGS BEYOND

MARGINALITY

Discussing female work through epigraphic evidence where women have been
commemorated by an occupational title will give an overview of jobs that were
open to and pursued by Roman women. In several of the examples discussed,
the link between the women and their work is clear and a combined female and
occupational identity have been commemorated through a job title. However,
many inscriptions with job titles concern a couple or a whole group of people,
often both men and women. In such cases, the name of the job is normally in the
masculine plural and often thought to refer to an occupation pursued by the
men in the group. Readings of such inscriptions with the grammatical exclusion
of women is an example of how women and female work are marginalized and
more difficult to identify. Questions of women’s occupational roles in such
epigraphic contexts have already been touched upon, concerning the evidence
of purpurarii from Rome. Yet another example—also from Rome—is an
inscription commemorating a man named Gaius Cafurnius Antiochus and his
wife, Veturia Deuteria (Fig. 9.4).62 It includes the name of an occupation in
connection to the man’s name, lanarius, but there is no job title linked to the
woman’s name. Does this mean that the woman did not have a job? Did she
perhaps work in the same trade as the husband? Or did she do something
completely different? The exact meaning of lanarius is not altogether clear, but
both ‘producer of wool’ and ‘dealer in wool’ have been suggested.63 Whatever its
meaning, the job title has an obvious link with lana, wool, and there are more
examples of lanarii from Rome and from other urban centres in Italy such as
Gubbio and Pesaro.64 As an occupational title in epigraphy, lanarius exists only
in the masculine. Should this be interpreted as implying that no women were
involved in the work done by lanarii?
This particular instance, of C. Cafurnius Antiochus and Veturia Deuteria,
is, furthermore, one of the few where an inscription with a job title is
combined with an image including a possible occupational symbol—a sheep.
The sheep is a unique symbol in the iconography of Roman textile production
and underlines the close connection of this job to wool. Above the sheep is also
depicted a pair of joined hands, one of them a woman’s arm with a bracelet. In
Roman iconography a man and a woman clasping hands—the gesture of
dextrarum iunctio—usually symbolizes a married couple.65 It was of particular

62
CIL 6.9489; Zimmer (1982: no. 34), of late Republican date; Larsson Lovén (forthcoming:
no. 1.1).
63
Von Petrikovits (1981: 100); Frayn (1984: 151–2). More recently Suzanne Dixon has suggested
it could be a male catch-all comparable to the female lanifica; see Dixon (2001: 119).
64
Rome: CIL 6.94990–4, 31898, 33869; AE 1971, 49. Gubbio: CIL 11.5835. Pesaro: CIL
11.6367; this last inscription includes a group of men, among them two lanarii and two vestiarii.
65
Larsson Lovén (2010), with further references.
 Women, Trade, and Production 213

Fig. 9.4. Funerary plaque of the lanarius C. Cafurnius Antonchus and his wife Veturia
Deuteria, Rome (CIL 6.9489). Photo: Lena Larsson-Lovén.

importance for people who had not always had the right of a legal marriage,
such as ex-slaves, to demonstrate their acquired rights and new status. In this
case, both the man, the lanarius, and the woman have clearly attested their
status as former slaves, but they come from different households. The woman
was an ex-slave from the Veturii family. As Holleran has recently argued,
female slaves could learn a trade that could make them more useful to the
household.66 After manumission, such training could become their profes-
sional basis and social identity. This may have been the case for a woman like
Veturia Deutera. For a female ex-slave of the Veturii—who, as already men-
tioned, were known for their engagement in textiles and the dyeing trade—it
does not seem unlikely that she had been engaged in some form of textile
production. If one adds the evidence for marriages between people belonging
to different families working in the same branch of the urban economy, it
seems plausible that Veturia Deuteria was involved in a family-run textile
business, first as a slave and later, in her marriage with the lanarius, perhaps
together with her husband—but her work was not specified in the commem-
orative inscription.67
Epitaphs commemorating slaves and freed people were more inclined to
document work than were those commemorating freeborn. This is particu-
larly true for urban slaves; not all slaves and freed people were equally likely to
be commemorated. A work position could influence the way people were
remembered, and those with a higher position in the work hierarchy occur

66 67
Holleran (2013: 315). For a discussion on this item, see Dixon (2001: 119–21).
214 Lena Larsson Lovén

more frequently in the epigraphic record than those with a job of a lower
status. Thus, a supervisor such as a lanipenda/-us stood a better chance of
being commemorated than someone who held a low-status position, such as a
weaver or a spinner. This situation is clearly reflected in the inscriptions,
where only a handful of weavers and spinners are documented, a very meagre
body of evidence that does not accurately reflect this type of work, which must
have involved an extensive number of workers.
Gender structure is another vital agency directing the modes of commem-
oration and of how female work was documented in the sources. The public
identities of Roman women were usually not defined primarily by their work.
However, the reluctance to document women at work, as we can see both in
writing and in images, reflects gender roles more than a social reality. The
above discussion of textile work can serve as an example of how gender
structures affected the commemoration of men’s and women’s professional
roles. Cloth production was considered to be the work of women; in an ideal
world it took place in a domestic setting. In spite of the traditional view of a
close link between textile production and women’s work, male jobs appear
much more common in the epigraphic evidence of occupations related to
textile production and trade than does the work of women. Instead of infor-
mation of women’s professional and economic roles, the stress in ancient
sources is more often on their traditional roles as wives and mothers than on
their work.68
The absence of explicit information on female work makes the identifica-
tion of women’s professional and economical roles more complex. To make it
possible to identify the place and role of women in the Roman labour market
beyond simply mapping jobs pursued by women, a reading of the sources
needs to be seen in relation to overarching gender structures. This could be
done by ‘listening to silence’ or reading gaps and absences, as discussed at the
beginning of this chapter, in order to understand what may have been
regularly excluded from the sources. An example of the exclusion of women’s
work could be what Susan Treggiari has suggested to be a parallel between
the absence of women’s work in modern evidence and in ancient sources.
Treggiari used the signs of shops and businesses that mention a male name in
combination with ‘and son(s)’ or ‘and co.’ to reinterpret ancient sources where
women have been excluded.69 Does the absence of references to women in
modern business signs mean that no women worked in there? From a present-
day Western perspective most people would probably not be surprised to find
women among the personnel in positions as secretaries, switchboard oper-
ators, or cleaning staff, and in other positions as well, in a business called
‘NN & son(s)’ or ‘NN & co.’, even without female names appearing in the signs

68 69
Saller (2007); Groen-Vallinga (2013: 297–9). Treggiari (1979a: 78–9).
 Women, Trade, and Production 215

on public display. This forces us to think again about the information of

female labour in Roman sources: in contrast to the impression the ancient
sources may give, Roman women must have worked extensively in production
and trade. They could work as professional partners of their husbands
or in family businesses, and were allowed to make profits from their
business. However, their contributions were rarely expressed in job titles on
inscriptions.
Applying modern theoretical models to ancient sources may give a better
understanding of female socio-economic roles, within the family and on the
labour market, and by adding information from other types of evidence the
picture may be more complex.70 Ulrike Roth has discussed in-depth methods
and the uses of material culture to identify the work of female slaves in
agriculture—the most silent and invisible women of all—to identify their
work and to discuss their economic significance.71 Iconography is another
source for the working lives of women that can add information, as has already
been shown by Natalie Kampen in her pioneering study on images from Ostia
of women at work.72 At Pompeii, the paintings from the facade of what is
known as the felt shop of the vestiarius M. Vecilius Verecundus provide a key
iconographic example of a woman at work.73 While we know the name and
the job title of the man—they are written underneath his portrait on the
facade of the workshop—there is no evidence specifying the name or the
occupation of the woman who, in one of the paintings on the same facade,
stands behind a desk to sell the products of the workshop (Fig. 9.5). The status

Fig. 9.5. Pompeii, facade to the left of shop IX 7, 7: woman selling products in shop.
Reproduced from Notizie degli Scavi di Antichità (1912: 179).

70 71
See Groen-Vallinga (2013) for women’s work. Roth (2007).
72
Kampen (1981); see also Kampen (1982, 1985) for studies of saleswomen in iconography.
For an overall study of work in Roman iconography, see Zimmer (1982).
73
Pompeii, IX 7, 7. CIL 4.3130. See also Angelone (1986).
216 Lena Larsson Lovén

of the woman is ambiguous, but she has been interpreted as the wife of
Verecundus because of her prominent place in the iconographic representa-
tion of the workshop.74
Another example of a man and woman working in the same business is a
funerary relief from Rome representing a scene in a butcher’s shop (Fig. 9.6).
The right half depicts a man at work: dressed in a tunic, he is chopping meat
behind a working table. Around him, a variety of meat products and work
equipment is displayed. To the left is a woman seated in a high back chair with
her feet resting on a stool. She is dressed in a foot-long garment, possibly a
tunic, over which she wears a mantle. Her hair is carefully arranged in a style
that came into fashion during the reign of Hadrian. She is holding a set of wax
tablets that probably symbolize a book of accounts.75 By dress and pose the
woman can be identified as a person of a higher status than the man, and she
may actually have been the owner of the shop, the butcher being one of her
workers rather than her husband. Women could take over the management of
a business after the death of a husband; they could also own property and
administer it themselves. They contributed to the commercial culture of
Roman urban centres, but only rarely were they commemorated in their
professional and economic roles; the woman in the butcher’s scene is a rare
example.76 Women’s work opportunities were undoubtedly more limited than
those for men in Roman society, especially for freeborn women, but still some
were commemorated by an occupational or a professional identity and some
were obviously financially successful in both smaller and larger businesses in

Fig. 9.6. Funerary relief from Rome depicting a butcher at work and a woman doing
accounting. Photo: Dresden, Staatliche Kunstsammlungen.

74
Clarke (2003: 109); Flohr (2013: 282–4, figs 117–18); Holleran (2013: 316–17).
75 76
Zimmer (1982: 94–5, n. 2). Dixon (2004: 65).
 Women, Trade, and Production 217

various urban centres of Roman Italy. Women from Pompeii such as Euma-
chia, Naevoleia Tyche, and Julia Felix are often mentioned as examples of
women’s successful involvement in business. From Ostia comes the example
of a woman named Junia Libertas, who put up a long inscription telling that
she left gardens, houses, and shops to her freedman/women. It is not clear
what kind of business Junia Libertas was involved in, but she was obviously a
wealthy woman who could afford a public legacy to be remembered by.77
Economically successful women or those from wealthy families could be made
more visible in urban centres than women of lower social status. In the
imperial period, it was common practice to honour these women with statues,
as happened with Eumachia, who got a statue dedicated to her in the role of
priestess by the fullones of Pompeii. Alternatively, women could pay for
statues of themselves or of family members, or could donate publica munifi-
centia to their city. Emily Hemelrijk has shown in a number of studies how
women in cities of Roman Italy and the Latin West demonstrated wealth and
social status by paying for public buildings, which were given their (family)
name, by erecting lavish funerary monuments, and by becoming visible in the
urban landscape by public statuary.78
Yet another example of a woman in business is a liberta named Trosia
Hilaria from Aquileia in the north-east. She identified herself as lanifica circ(u)
latrixs (= circulatrix).79 The woman had a funerary stele made to herself and to
her freed male and female ex-slaves. There is no image on the stele to guide us
further about the meaning of the unique job title, but it is quite clear that the
expression lanifica in this case was used not as the customary praise of an
industrious housewife, but rather as a job title. Trosia Hilaria, the lanifica
circulatrix, may have been a spinner or weaver who worked on a professional
basis as a pedlar in different households in the region of Aquileia, but it is also
possible that she ran a small-scale business where some of her slaves worked.80
Trosia Hilaria, and her slaves may have been collecting wool from local
households, processing it, and returned it for a fee in a sort of putting-out to
cottage industry. The woman had been a slave herself, but at the time of
commissioning the funerary stele she was a slave-owner as well, of both male
and female slaves. We have no information of how many slaves she had, but at
least some of them had been freed and were buried together with the former
owner. The fact that she was a slave-owner indicates that Troisa Hilaria had
some financial means, but on a more modest scale than, for instance, Naevo-
leia Tyche in Pompeii and Junia Libertas in Ostia. Trosia Hilaria was not
without professional success or economic means, but commissioning a joint

77 78
AE 1940, 94; Dixon (1992). Hemelrijk (2012, 2013) with further references.
79
Museo Archeologico di Aquileia, inv. no. 49941 (= AE 2003, 115). The stele is 1.10 m high
and 0.59 m wide. The style of the writing points towards a late Republican date.
80
See Holleran (2013: 322) for a discussion of circulatrix.
218 Lena Larsson Lovén

funerary monument and grave with (some of) her manumitted slaves points
more in the direction of shared work and a social identity with slaves and
other manumitted persons rather than with the higher echelons in society.

DISCUSSION

Based on epigraphy, the general picture of women’s occupational roles is

that they were frequent in domestic service jobs, in taking care of small
children, in cloth production with partly highly specialized jobs, and in the
trade of some goods such as luxury products, while they are completely
absent from many other sections.81 A prima facie interpretation of the
evidence could lead to the conclusion that this was the occupational realm
of women and that most of the labour market was closed to them. Some
past scholars have also fallen into this trap. For instance, A. H. M. Jones
argued in his influential paper on the cloth industry under the Roman
Empire that women’s role in the production of textiles appeared to be of
no particular significance: their foremost task was spinning, which was
done in an ‘entirely unorganised’ way and in their ‘spare time’.82 Weaving,
on the other hand, was, according to Jones, mostly done by men and
regularly organized on a large scale.
In all ancient media, we are presented with a stereotyped picture of women
and their professional roles, which were more often excluded than included;
instead, it was their traditional family roles such as wives and mothers that
were regularly stressed. This situation reflects long-standing gender ideologies:
for women of various social classes it was not an occupational life, but rather
life as a traditional housewife, that remained the ideal. This leads to women
being represented as persons who were supposed not to work outside the
home but to be dutiful and industrious within the household. In reality,
however, both men and women were part of a labour market, but Roman
ideologies of work, of masculinity and femininity, had such an impact on the
documentation of work that women workers and their economic roles have
been strongly marginalized in ancient sources. Thus, the documentation of
their professional roles is a result of social conventions and Roman gender
ideologies, which encouraged female identities in their family roles rather than
their professional roles.
The study of Roman women has developed in many directions since
scholars began to address the issue in the 1970s: new theoretical models
have been applied to ancient sources, which have led to new readings that
have helped us better to understand women’s socio-economic roles. Reading

81 82
Treggiari (1976); see also Holleran (2013). Jones (1960).
 Women, Trade, and Production 219

the evidence by trying to see not only what is there but also what is not there
opens possibilities to see women as professionally active in more sectors than
those that have been documented in epigraphy by a female job title. It is a way
of looking beyond marginality and to see a more diversified and less stereo-
typed picture of women’s professional and economic roles in the labour
market in Roman society as well as seeing both male and female craftsmen
and traders as vital socio-economic agents in urban Roman communities.

REFERENCES

Angelone R. (1986). L’officina coactiliaria di M. Vecilio Verecundo a Pompei. Naples.

Bradley K. R. (1991). Discovering the Roman Family: Studies in Roman Social History.
Oxford.
Clarke, J. (2003). Art in the Lives of Ordinary Romans. Berkeley and Los Angeles.
Dixon, S. (1992). ‘A Woman of Substance: Junia Libertas of Ostia’, Helios 19: 162–74.
Dixon, S. (2000–1). ‘How do you count them if they’re not there? New Perspectives on
Roman Cloth Production’, Opuscula Romana 25/26: 7–17.
Dixon, S. (2001). Reading Roman Women: Sources, Genres, and Real Life. London.
Dixon, S. (2004). ‘Exemplary Housewife or Luxurious Slut: Cultural Representations of
Women in the Roman Economy’, in F. McHardy and E. Marshall (eds), Women’s
Influence on Classical Civilization. London and New York, 56–74.
Edmondson, J. (2015). ‘Economic Life in the Roman Empire’, in C. Bruun and
J. Edmondson (eds), The Oxford Handbook of Roman Epigraphy. New York and
Oxford, 671–95.
Eichenauer, M. (1988). Untersuchungen zur Arbeitswelt der Frau in der römsichen
Antike. Frankfurt am Main.
Flohr, M. (2013). The World of the Fullo: Work, Economy, and Society in Roman Italy
(Oxford Studies on the Roman Economy). Oxford.
Frayn, J. (1984). Sheep-Rearing and the Wool Trade in Italy during the Roman Period.
Liverpool.
Gardner, J. (1986). Women in Roman Law and Society. London and Sydney.
Gleba, M. (2008). ‘Auratae vestes: Gold Textiles in the Ancient Mediterranean’, in
C. Alfaro and L. Karaki (eds), Purpurare Vestes II. Vestidos, Textiles y Tintes:
Estudios sober la produccion de bienes de consumo en la antiguidad. Valencia, 63–80.
Gregori, G. L. (1994). ‘Purpurarii’, in Epigrafia della produzione e della distribuzione:
Actes de la VIIe rencontre franco–italienne sur l’épigraphie du monde romaine.
Rome, 740–3.
Groen-Vallinga, M. (2013). ‘Female Participation in the Roman Urban Labour Mar-
ket’, in Hemelrijk and Wolf (2013), 295–312.
Hemelrijk, E. (2012). ‘Public Roles for Women in Cities of the Latin West’, in S. Dillon
and S. L. James (eds), A Companion to Women of the Ancient World. London, 478–90.
Hemelrijk, E. (2013). ‘Female Munificence in the Cities of the Latin West’, in Hemel-
rijk and Wolf (2013), 65–84.
220 Lena Larsson Lovén
Hemelrijk, E., and Wolf, G. (eds) (2013). Women in the Roman City of the Latin West
(Mnemosyne Supplement, 360). Leiden and Boston.
Hildebrandt, B. (2013). ‘Seidenkleidung in der römischen Kaiserzeit’, in M. Tellenbach,
R. Schulz, and A. Wieczorek (eds), Die Macht der Toga: Dresscode im römischen
Weltreich. Mannheim, 58–61.
Holleran, C. (2012). Shopping in Ancient Rome: The Retail Trade in the Late Republic
and the Principate. Oxford.
Holleran, C. (2013). ‘Women and Retail in Roman Italy’, in Hemelrijk and Wolf
(2013), 313–30.
Huttunen, P. (1974). The Social Strata in the Imperial City of Rome: A Quanti-
tative Study of the Social Representations in the Epitaphs Published in the CIL
VI. Oulu.
Jones, A. H. M. (1960). ‘The Cloth Industry under the Roman Empire’, Economic
History Review 13/2: 183–92.
Joshel, S. R. (1992). Work, Identity, and Legal Status at Rome: A Study of the
Occupational Inscriptions. Norman, OK, and London.
Kampen, N. (1981). Image and Status: Roman Working Women in Ostia. Berlin.
Kampen, N. (1982). ‘Social Status and Gender in Roman Art: The Case of the
Saleswoman’, in N. Broude and M. D. Garrard (eds), Feminism and Art History:
Questioning the Litany. New York, 63–71.
Kampen, N. (1985). ‘Römische Strassenhändlerinnen: Geschlechte und Sozialstatus’,
Antike Welt 16/4: 23–42.
Knapp, B. (2011). Invisible Romans, Cambridge, MA.
Larsson Lovén, L. (2010). ‘Marriage and Marital Ideals on Roman Funerary Monu-
ments’, in L. Larsson Lovén and A. Strömberg (eds), Ancient Marriage in Myth and
Reality. Newcastle upon Tyne, 204–20.
Larsson Lovén, L. (forthcoming). The Imagery of Textile Making. Oxford.
Lefkowitz M. R., and Fant, M. B. (1992). Women’s Lives in Greece and Rome: A Source
Book in Translation. London.
Le Gall, J. (1970). ‘Métiers de femmes au Corpus Inscriptionum Latinarum’, REL 47bis:
123–30 (= Mélanges M. Durry).
Papi, E. (2002). ‘La turba inpia: Artigiani e commercianti del Foro Romano e dintorni
(I. sec. a.C.–64 d.C.)’, JRA 15/1: 43–62.
Pelletier, A. (1996). ‘Les Métiers du textile en Gaule d’après les inscriptions’, in Aspects
de l’artisanat du textile dans le monde méditerranéen (Egypte, Grèce, monde romain).
Paris, 133–6.
Petrikovits, H. von (1981). ‘Die Spezialisierung des römischen Handwerks’, in
H. Jankuhn, W. Janssen, R. Schmidt-Wiegand, and H. Tiefenback (eds), Das
Handwerk in vor- und frühgeschichtlicher Zeit. Göttingen, 63–131.
Pomeroy, S. (1976). Goddesses, Whores, Wives, and Slaves: Women in Classical
Antiquity. New York.
Reinhold, M. (1970). History of Purple as a Status Symbol in Antiquity. Brussels.
Richlin, A. (2014). Arguments with Silence: Writing the History of Roman Women.
Ann Arbor.
Roth, U. (2007). Thinking Tools: Agricultural Slavery between Evidence and Models.
London.
 Women, Trade, and Production 221
Saller, R. P. (2007). ‘Household and Gender’, in W. Scheidel, I. Morris, and R. P. Saller
(eds), The Cambridge Economic History of the Greco-Roman World. Cambridge,
87–112.
Salvaterra, C. (2006). ‘Labour and Identity in the Roman World: Italian Historiog-
raphy during the Last Two Decades’, in B. Walldijk (ed.), Professions and Social
Identity: New European Historical Research on Work, Gender and Society. Pisa,
15–38.
Treggiari S. (1975). ‘Jobs in the Household of Livia’, PBSR 30: 48–72.
Treggiari S. (1976). ‘Jobs for Women’, American Journal of Ancient History 1: 76–104.
Treggiari S. (1979a). ‘Lower Class Women in the Roman Economy’, Florilegium 1:
65–83.
Treggiari S. (1979b). ‘Questions on Women Domestics in the Roman West’, in
M. Capozza (ed.), Schiavitù, manomissione e classi dipendenti nel mondo antico:
Atti del Colloquio internazionale (Bressanone, 25–27 novembre 1976) (Pubblicazioni
dell’Istituto di storia antica, 13). Rome, 186–201.
Zimmer, G. (1982). Römische Berufsdarstellungen (Archäologische Forschungen, 12).
Berlin.
 10

Freedmen and Agency in Roman Business

Wim Broekaert

When analysing the division of a society’s workforce according to status,

economic historians of the medieval or modern periods usually present tables,
figures, and graphs, for statistical evidence neatly illustrates which status group
conducted a specific trade and offers an explanation for the dominance or
absence of particular status classes in the professional world. Yet, when trying
to analyse the relationship between status and work division for the Roman
world, this kind of evidence is generally lacking. The reasons are all too well
known. First, archival records in which a person’s profession and status were
explicitly cited are virtually absent, with the obvious exception of Egypt.
Secondly, serial analysis of occupational inscriptions has turned out to be a
less than ideal alternative. Limitations include the inevitable loss of thousands
of texts and the so-called epigraphic habit, which forces us to reconsider the
actual value and representativeness of the surviving inscriptions.1 We need to
take into consideration that some social groups were far more likely to be
commemorated in an inscription than others, and that many reasons may
determine whether or not to include a profession and status marker. Regional
differences in the epigraphic habit also blur the overall picture of the Roman
workforce. Particularly striking is, for example, the fact that the epigraphy of
the Spanish and African provinces so far has yielded only a handful of
inscriptions mentioning traders, even though one would expect several traces
of the activity of merchants and shippers, as both regions are rich in epigraph-
ic remains and played a crucial role in Rome’s food supply. Yet it seems that
merchants originating from these provinces mentioned their trade only when
erecting an inscription in an area where citing one’s profession was part of the
local epigraphic habit. Most telling are the inscriptions of the oil merchant
L. Marius Phoebus. His name, without any additional information, features
among the initiators of a monument located in Cordoba, presumably his

1
MacMullen (1982); Mouritsen (2001, 2005).
 Freedmen and Agency in Roman Business 223

native town, but only his epitaph in Rome mentions his profession.2 In this
particular case, one might argue that the nature of the inscriptions (dedicatory
versus funerary) determined whether or not to include a profession. This is
not an isolated instance, for it is probably no coincidence either that African
shippers and merchants decorated their offices on the Piazzale delle Corpor-
azioni in Ostia with mosaics identifying their profession and cities of origin,
but that so far none of these cities has yielded a contemporary inscription of a
negotiator or navicularius.3
Still, despite these difficulties, the importance of status easily features
among the most contested topics of Roman trade and, in a wider sense, the
Roman economy. No ancient historian will deny that a major part of com-
merce was conducted by slaves and freedmen, but the consequences of having
trade run by legally dependent and semi-dependent businessmen have been
hotly debated. This chapter therefore aims to discuss the importance of status
and dependency in the organization of Roman trading.
The outline of the text is as follows. First, I will briefly sketch out the role
attributed to status and dependency in earlier discussions of Roman trade and
the apparent evolution in assessing both features. The second section will
discuss the relationship between principals and agents and rely on several case
studies to illustrate how agency by freed businessmen was organized. The third
section focuses on the ratio of economic dependence and independence
among freedmen. A final section will offer some concluding remarks.

STATUS, DEPENDENCY, AND AGENCY I N

HISTORICAL RESEARCH

Since the 1970s, economic historians have tried various ways to incorporate the
notion of status in the analysis of the Roman economy. For some, the status of
people engaged in trade determined the very nature of the economy and proved
to be a crucial part in assessing the economic performance of the Roman business
world. For others, status mattered only in describing the economic relationship
between patrons and freedmen and the analysis of aristocratic portfolios. This
section offers a quick overview of the dominant views.4
Every discussion of status and dependency in the Roman trading world
obviously starts with Finley’s The Ancient Economy. Finley and the adherents
of the substantivist view argued that status is a crucial trait of Roman business,
which therefore clearly separates the Roman economy from contemporary
economies. They claim that the way of organizing a business enterprise and

2
Cordoba: AE 2000, 734. Rome: CIL 6.1935. See also Remesal Rodríguez (2000).
3 4
CIL 14.4549.10–14, 17–19, 21–2, 34–6, 38–9. Schleich (1983); Andreau (1995).
224 Wim Broekaert

spending profits was heavily influenced by the status of the people involved:
economic behaviour is completely determined and even predictable by the
place one occupies along the ‘spectrum of statuses’.5 The substantivist model
thus centres around a proposed bipolar structure of the Roman economy. At
one end of the economic spectrum, we find people of freeborn status and
social prominence concentrating on agriculture and recoiling from the risks of
financing trade and industry, so that these last two sectors of the economy
remained deprived of major capital investments. The other end consists of
people of low and freed status who did not have sufficient means to run their
own estate and were thus forced into a profession in trade and industry. Being
freedmen and foreigners, they could rely on only a small starting capital and
limited means of investment. Small-scale trade and commercial failure was the
evident consequence.
This model thus relies on two interrelated premisses. First, the economic
world of the aristocracy, specializing in agriculture, is virtually separated from
that of the lower social strata, which engaged in trade and industry. Agency by
slaves and freedmen or silent partnerships were only a marginal phenomenon
in the economic decision-making of the Roman aristocracy.6 This economic
dichotomy can easily be recognized in Finley’s description of the business
community in Gaul, clearly separating the freed and foreign merchants from
the local aristocracy.7 Castrén offered a similar description of the Pompeian
society, in which he stressed the differences between freedmen dominating
urban production and commerce, and the landed aristocracy.8 Secondly, as
soon as a businessman becomes wealthy and powerful, he immediately turns
away from commerce to live as a landed aristocrat. Money made in trade and
industry never returned to the business world to start a new enterprise, but
was used to invest in land and to enhance the merchant’s social prominence by
ostentatious display of wealth and benefactions.
With a landed elite shunning investments in commerce, the Roman business
world was apparently doomed to failure, inefficiency, and small-scale transac-
tions. Yet, despite his categorical conclusion, Finley duly recognized that his
model relied mainly on literary evidence and that scrutinizing the epigraphic
evidence might adjust, even alter, the picture of the Roman trading world. Only
recently has this call been answered by a few researchers who have ventured on a
systematic analysis of Greek and Roman occupational inscriptions.9
The substantivist model met severe criticism, most notably in the work of
John D’Arms, who claimed that Finley’s sociological model could indeed work
for a few members of the imperial upper class, but did not accommodate the

5 6 7
Finley (1973: 68, 87). Finley (1965: 37; 1973: 57–9). Finley (1973: 59–60).
8
Castrén (1975: 121).
9
Ruffing (2008); Broekaert (2011, 2012a, b). Joshel (1992) offers an interesting sociological
insight into the occupational world of Rome, but is disappointing for economic historians.
 Freedmen and Agency in Roman Business 225

economic portfolios of the majority of the senators, knights, and members of

the local urban aristocracies.10 He convincingly argued that the imperial and
municipal aristocracy did invest in trade and industry, but only indirectly, by
relying intensively on their slaves and freedmen. Nevertheless, his model
provided for major capital injections in business by the Roman aristocracy.
Hence, with the presumed bipolarity of the economy adjusted, the importance
of status in the organization of Roman trade evidently dwindled and status lost
its meaning as a variable in the assessment of the performance of the Roman
economy. Nonetheless status did not disappear completely. In D’Arms’s
model, members of all status groups apparently engaged in commerce, but
assumed different roles: high-status groups only invested in trade, while
dependent, lower-status groups were responsible for the actual day-to-day
business deals and voyages. The Finleyan gap between high and low status was
bridged, but it did not disappear.
In 2007, a new analysis of the interrelation of dependency and agency was
put forward in the Cambridge Economic History of the Greco-Roman World, in
which the rich contributions of the New Institutional Economics were applied
to the ancient business world.11 This modern analytical framework describes
how law and institutions influence and shape a society’s economy and try to
find solutions to inherent constraints. One of the topics most relevant for this
chapter concerns agency. It is well known that agency always suffers from a
major inhibitory problem—namely, the risk of assigning an untrustworthy
agent. One always has to reckon with the problem of information asymmetry,
for the agent usually has far more information and immediate control of the
economic activities than his principal, so the latter may be unable to closely
supervise and assess his agent’s decisions. Hence, the agent may be tempted to
neglect his principal’s interests or to capture part of the investments and
profits. To prevent this, the principal must either offer sufficient incentives
to his agent, or be able to have recourse to additional control mechanisms.
This is where the real value of working with dependent managers becomes
apparent, for legal and social dependency can be crucial assets in monitoring
the actions of an agent. Freedmen, for instance, could be obliged by their
patron to perform well-defined operae or services. Ulpian asserted that a
freedman is under a natural obligation to perform services for his patron.12
If a patron stipulated that managing his affairs belonged to his freedman’s
operae, then neglect, theft, and any other kind of fraud would be interpreted as
a refusal to perform legally compulsory services and would hence be liable to
punishment. Patrons were well aware of the potential benefits this operae-
system had to offer, for one of the categories comprised operae fabriles or
professional services, with very few restrictions as to the nature or setting of

10 11 12
D’Arms (1977, 1981). Frier and Kehoe (2007: 122–34). Dig. 12.6.26.12.
226 Wim Broekaert

the services.13 Additionally, the deliberately vague duty of obsequium or

compliance, which every freedman was obliged to follow and which included
gratitude, obedience, and respect to one’s patron, may also have eased agency.
Paul, for instance, states that ‘a freedman is ungrateful when he does not show
proper respect for his patron, or refuses to administer his property, or undertake
the guardianship of his children’.14 Thus, freedmen who operated as managers
and deceived or robbed their patron clearly violated the obsequium. In theory,
this legal obligation was thus supposed to stimulate freedmen to behave as
trustworthy and honest business agents.
This short overview has stressed the changes in assessing the role of status
and dependency in the Roman business world. From an apparent structural
weakness hampering the growth of the economy at large in the substantivist
model, status and dependency have become important features easing agency
and stimulating large-scale business ventures.
So far, however, the analysis of Roman agency has predominantly relied on
legal sources, and for obvious reasons: business management by slaves is
meticulously analysed in the various actiones adiecticiae qualitatis, and agency
by freedmen is regularly hinted at when the patron’s rights are discussed.15
This chapter, on the other hand, intends to adduce mainly inscriptional
evidence to clarify the interaction between dependency and agency by freed-
men and assess whether inscriptional evidence fits the model of freedmen as
managers par excellence. The discussion will focus on three main topics.
First, all three models accept that aristocratic patrons did not engage in
trade personally. Apparently, in this respect status continues to matter for the
organization of the Roman business world. But was it actually true that high-
status investors never took an active part in the decision-making and merely
provided financial back-up and infrastructure to their freedmen in exchange
for a share in the net return?
Secondly, when agency by freedmen provided a solution to the dangers of
information asymmetry, how did principals actually employ their agents? Is
inscriptional evidence able to offer some insights in the actual structure of
agency and the tasks assigned to freedmen?
Thirdly, if agency by freedmen proved to be a successful tool in organizing
business, why would patrons ever forgo the opportunity of using their former
slaves as agents? Do inscriptions allow us to detect economically independent
freedmen?

13
Dig. 38.1.16–19 and 38.1.26 pr.; Waldstein (1986).
14
Dig. 37.14.19. Cf. also Dig. 1.12.1.10, 1.16.9.3 and Joshel (1992: 33–4).
15
Kirschenbaum (1987); Wacke (1994); Földi (1996); Stelzenberger (2008); Gamauf (2009).
Aubert (1994) also includes inscriptional evidence, but focuses mainly on production and shop
management.
 Freedmen and Agency in Roman Business 227

TRADE AND THE ROMAN A RI STOCRACY:

TRACING THE I NVISIBLE

The Roman aristocracy apparently attached high value to two attitudes

towards the sources of wealth on which the family fortune was based: highlight-
ing the virtues of the socially respectable gentlemanly absentee landownership,
and creating as much distance as possible between personal engagement and
investments in commerce. Yet surely some members of the aristocracy had a
keen eye for business and may have preferred closely to monitor their agents’
activities or even to organize part of the enterprise themselves. The lack of
archives and the loss of correspondence between investors and managers prevent
a straightforward substantiation of this assertion, yet inscriptions mentioning
merchants may provide some telling hints.
The most famous example is that of Sex. Fadius Secundus Musa.16 In AD 149,
Musa was honoured as a rich and generous magistrate in his native city of
Narbonne.17 His name also occurs on twenty-five tituli picti on Spanish oil
amphorae, all discovered on Monte Testaccio in Rome.18 These tituli identify
the merchant distributing the vessel’s contents, so there can be no doubt that
Musa, a municipal aristocrat, was personally engaged in commerce. Yet, one
could argue that perhaps Musa was shipping oil only before joining the local
upper class and that his rise in the Gallic society started only when he had retired
from trade, so no deprecation would hinder his political career. Luckily, however,
the tituli picti are carefully dated. The monumental inscription from AD 149
remarked that, by then, Musa had attained every single function in the local
cursus honorum. Moreover, the family’s social mobility did not end there, for the
text also stresses that his grandson Iucundus had joined the ordo senatorius.
Apparently, one of Musa’s children married a member of the senatorial aristoc-
racy. Surely one would expect that the Gallic Fadii by then had started imitating
the lifestyle of the rich and famous and shunned away from trade. Yet, the tituli
mentioning Musa’s name prove otherwise. His engagement in the oil business
ranges from AD 146 to 161, so we can safely conclude that, both during his
political career as well as after having reached the highest functions and married
one of his children to a member of the imperial aristocracy, Musa continued
shipping Spanish oil to Rome. Tituli also attest to the presence in the oil trade
during the same period of other members of the Sexti Fadii, whose freedman
status is indicated by their Greek cognomina and who most likely belonged to
Musa’s familia.19 We thus encounter a socially prominent patron, a wealthy
member of the Gallic aristocracy who personally engaged in trade and launched

16 17
Héron de Villefosse (1915). CIL 12.4393.
18
CIL 15.3863–73a–p.
19
CIL 15.3855–61a–d (Anicetus); Rodríguez Almeida (1972: no. 16 [Antiochus]) and CIL
15.3862 (Paon).
228 Wim Broekaert

several of his freedmen in the same business. Musa was responsible for the
decision-making and organization of the oil trade and had at least the oppor-
tunity of closely monitoring his freedmen. He may even have cooperated with
them in a commercial partnership, for seven tituli document a societas of Fadii.20
Sadly, these tituli never cite cognomina, so we cannot know for sure whether
Musa had concluded a partnership with his freedmen or the latter were cooper-
ating without their patron’s personal involvement. The rich dossier of the Sexti
Fadii indeed neatly illustrates how prominent patrons could move across the
boundaries of status and work in the same economic environment as their
freedmen.
For the next example illustrating the simultaneous presence of aristocratic
merchants and their freedmen in trade, we turn to the Ostian grain market.
Given the importance of Rome’s port in organizing the imperial supply system,
Ostia must have housed a large association of grain merchants. One of its most
prominent members seems to have been P. Aufidius Fortis, who was doing
business during the second quarter of the second century AD. He originated
from the African city of Hippo Regius, where he was elected a duovir. Yet he
appears to have left his native city and settled permanently in Ostia to organize
his trade. There he again quickly rose to the ranks of the local elite, for an
honorary inscription initiated by the mercatores frumentarii mentions that he
attained the functions of decurio, duovir, and quaestor aerarii Ostiensium.21
Wealth and prestige made Fortis one of the most influential Ostian businessmen,
with extensive connections in the local professional associations. The fabri
granted him the honorary title of praefectus, the grain-measurers and divers
elected him a patron, and his own association of grain merchants honoured him
as quinquennalis for life. In AD 146, Fortis eventually became patron of the city, a
distinction he celebrated by donating two silver statues and organizing games for
three days.22 Fortis was not the only Aufidius who engaged in the grain trade.
Around the middle of the second century, the Ostian grain merchants dedicated
a statue to Q. Calpurnius Modestus, a procurator ad annonam. The inscription
on the pedestal mentions several of the association’s magistrates, among them
P. Aufidius Faustianus, a quinquennalis, and P. Aufidius Epictetus, a quaestor.23
These grain merchants, whose cognomina suggest a freed status, obviously
belonged to the same family as Aufidius Fortis. Another Ostian inscription
clarifies the exact relationship, for both Faustianus and Epictetus, together with
their coliberti Euphrosynus and Ianuarius, feature among the initiators of a
statue, dedicated to their former master Fortis.24 The same four freedmen also
financed the erection of yet another statue, this time in honour of P. Aufidius

20
CIL 15.3874; Rodríguez Almeida (1991: no. 3) and Blázquez Martínez and Remesal
Rodríguez (2003: nos 61–6).
21 22 23
CIL 14.303 and 14.4621. AE 1946, 204. CIL 14.161.
24
CIL 14.4621.
 Freedmen and Agency in Roman Business 229

Fortis iunior, the homonymous son of their patron.25 We can thus follow the
continuation of the family’s interests in the grain trade. Are we to assume that
Fortis was still monitoring his freedmen’s actions? It seems at least very likely that
Fortis’s personal engagement did not end with his rise in Ostian society. When
his colleagues in the grain trade honoured Fortis as quinquennalis perpetuus, he
was already a member of the city council and had been elected quaestor aerarii
four times. Moreover, the fact that both Epictetus and Faustianus managed to
rise in the ranks of the mercatores and hold magistracies in the association when
their former master still held the highest magistracy can hardly be a coincidence.
Being freedmen of Aufidius Fortis apparently guaranteed them a favourable
starting position, protection, and success in the grain trade. This again suggests
that Fortis, as a lifetime member and magistrate of the association actively
supported his agents. We can imagine he shared his experience and commercial
network and provided financial back-up.
These two case studies suggest that at least some local aristocrats with
commercial interests did far more than merely financing their freed agents’
business. Some, like Fadius Secundus, may have continued trading, even after
having introduced several freedmen into the same business, and perhaps
cooperated with them in partnerships. Others, like Aufidius Fortis, used
their privileged position in professional associations to support not only
their agents’ business, but also their rise within these associations. These
aristocratic businessmen can hardly be styled absentee investors, but seem to
have actively monitored and sustained their agents’ transactions. How excep-
tional or commonplace this kind of cooperation really was is still difficult to
tell, for more detailed prosopographical analysis of local aristocracies, their
freedmen, and sources of wealth is called for. With the current state of
knowledge, however, this kind of personal engagement in business by local
elites apparently did occur, but still seems to be rather uncommon.

MODELLING ROMAN AGENCY:

A F RE E DM A N IN E V ER Y P O R T

When discussing the nature of agency in the Roman business world, the focus
often turns to agency by slaves, and for good reasons. Legal sources succinctly
analyse the actiones adiecticiae qualitatis, determining the liability of each
partner and documenting the presence of slaves in shops, land transport,
and maritime trade. Literary texts, too, frequently hint at slaves active in
business.26

25 26
CIL 14.4622. Lucian, Nav. 13; Plut. De lib. 7; Libanius, Ep. 177.1.
230 Wim Broekaert

The information on freed agents, however, is not systematically discussed

by the Roman jurists, for, theoretically, there was no need for a distinctive
body of laws structuring the cooperation between a patron and his freedman.
From a legal point of view, both were citizens with the right to enter into
a standard labour contract, which legalized the agreement, so no additional
clauses concerning liability were necessary. The consequence is that only mere
glimpses of agency relations between patrons and former slaves survive. Still,
legal and literary sources suggest that management by freed agents was very
similar to the way in which slaves were running their masters’ business. Paul,
for instance, depicts the difficulties of agency in commercial enterprises
beyond the sea and makes no distinction between slaves and freedmen.27
Papinianus describes how a master had a slave working as a manager in
banking, and the same business was carried on by his freedman after he had
been given his freedom.28 Cicero notes that freedmen were used to administer
business in distant provinces, in the same way as slaves did.29 These sources
suggest a continuity between management by slaves and freedmen. Not
surprisingly, for freedmen often performed the same professions already
entrusted to them when they were still slaves. Manumission hardly ever
entailed a rupture with the patron’s family and previous labour experience,
because freedmen often continued to live in the patron’s house and kept close
relations.30 The location of workshops, shops, and tabernae directly linked to
large residential houses presumably belonging to the urban elite can in fact
reflect this extended cooperation.31 Continuing relationships indeed made
perfect sense, for a prolonged cooperation offered benefits to both the patron
and the freedman. The former had invested time and money in teaching his
slaves a trade; he had had the opportunity to assess their character, aptitude
for business, and honesty, most likely based on previous experience when his
slaves were still managing their peculium; he would now be able to rely on a
well-known and trusted pool of skilled labour, instead of having to train new
and potentially unreliable slaves; and he could expect favourable terms of
cooperation. The latter continued to benefit from their patron’s investments,
business connections, and the infrastructure necessary to conduct their trade;
they were able to exploit previous experience and networks; and were not
forced to seek either another profession and new training or initial capital to
continue their trade independently.
If freedmen indeed continued to cooperate with their former master, and
agency by slaves or freedmen was very similar from an organizational point of
view, one may wonder why manumission was at all necessary. The answer is

27 28 29
Dig. 40.9.10. Dig. 14.3.19.1. See also Dig. 26.7.58.pr. Cic. Fam. 13.33.
30
Plin. Ep. 2.17.9; Dig. 7.8.2.1 and 7.8.6; Fabre (1981: 131–40); Kirschenbaum (1987: 135–8);
López Barja de Quiroga (1991: 163–4).
31
Wallace-Hadrill (1994).
 Freedmen and Agency in Roman Business 231

provided by the legal background to agency. As long as the agent remained

under the potestas of his master, agency was regulated by the actiones adiecti-
ciae qualitatis, which made the master liable for the business transactions of his
slave agent. Business partners of his agent could therefore bring the master to
court to claim damages. Once, however, the agent had been manumitted, he
became a person sui iuris. As a freedman, he could represent his patron in
court, join commercial partnerships with his patron, and cover all legal aspects
of agency. He was now completely liable for his business deals. A patron
investing in trade by relying on a freedman was thus never responsible for
any of the damages a partner of his agent incurred. No wonder, then, that
Ulpian mentions that slaves were manumitted in order to become agents or
that Gaius notes manumission before the age of 30 was allowed in cases where a
patron wished his former slave to become an agent.32 This regulation was
indeed very advantageous to patrons, but not to the agent (for he continued to
be liable) or the agent’s business partners (for they were never able to recover
the damage from the principal). By the end of the second century AD Roman
law recognized this discrepancy in responsibility and decided to protect agent
and partners with the introduction of the actio quasi institoria. Now the
principles of agency for agents in potestate also applied to free persons (freed
or freeborn), and the principal was liable for all transactions conducted by his
representatives.33 However, long before the introduction of this actio, several
other agency regulations, which made representation by free persons (includ-
ing freedmen) easier, were provided for by Roman law, such as mandatum,
negotiorum gestio, and procuratio.34 Their applicability for agency by freedmen
would repay detailed analysis, but cannot concern us here, for none of them left
any trace in the epigraphy of the Roman business world. Suffice it to say that
Roman law duly recognized that indirect agency by free persons was an
essential part of the Roman trading world and required separate legislation.
In short, many slaves working in the service of their master must eventually
have been manumitted to become, as Fronto calls them, liberti fideles ac
laboriosi.35
When trying to model management by freedmen, we need to account for a
large variety of agency relationships. On one side of the model, we have a patron
who is monitoring every single phase of the enterprise, who insists on making the
important decisions and assigns only small tasks to his freedmen with little or no
independence. On the other, we have a patron who is merely financing or
enabling business ventures, who completely relies on his freedmen for the
practical organization, and expects only a rent or share in the profits. To capture
these differences, we need to reckon with three main variables.

32
Dig. 40.2.13; Gai. Inst. 1.19.
33
Dig. 3.5.30, 14.3.19, 19.1.13.25. Kirschenbaum (1987: 142–3); Aubert (1994: 91–5).
34 35
Aubert (1994: 105–12); Verboven (2002: 227–51). Fronto, Ep. 2.7.2.
232 Wim Broekaert

First, the social status of the patron. The higher he ranks among the local
aristocracies and the closer his involvement in politics or other magistracies,
the less likely he will be personally engaged in business. The previous section,
however, made clear that there are exceptions to this rule. The patron’s wealth
and place in society also determine the magnitude of his household and thus
the amount of freedmen he will be able to rely on. The number of potential
agents is again crucial to the scale of business and the possibility of construct-
ing networks of agents.
Secondly, the geographical range of the enterprise. A freedman working in
or close to the city where his patron lived could easily be monitored by the
patron himself, his family, or friends. On the other hand, when a freedman
was doing business in distant places, control was far more difficult. The patron
could then either put his trust in the agent’s honesty and aptitude for business
but run the risk of being cheated, or try to create ‘check points’ by placing
other freedmen in ports along the route his freedman was frequenting, thus
creating a network of agents. As a few case studies will indicate, this strategy
proved to be particularly efficient in long-distance trading.
Thirdly, the nature of the merchandise. When an agent was managing high-
value merchandise, a patron may have wished to try and protect his investment
by monitoring his freedman’s actions by either interfering personally or relying
on his network of freedmen. When, on the other hand, small or low-value
cargoes were marketed, the loss of which would cause only minor damage to
the patron’s portfolio, strict control may not have been worth the effort.
With these variables in mind, we can now construct a typology of agency by
freedmen in Roman trade and try to fit in the dispersed data epigraphy has to
offer. Six basic levels of agency can be distinguished.

Freedmen sell the merchandise they receive from their patron

This category unites all freedmen marketing agricultural surpluses from their
patron’s estates or other merchandise produced or supplied by the patron.
They either run a shop, often situated close to the patron’s workshop, travel
around to distribute merchandise, or even settle in distant markets. This
organizational scheme must have structured a major part of agency relations,
yet is rather difficult to trace in epigraphy. Nonetheless, the following case
studies may reflect this particular business organization.
We first focus on the wine trade. During the first quarter of the second century,
a certain Caedicia M.f. Victrix was producing Dressel 2–4 wine amphorae in the
vicinity of Sinuessa, a city located slightly to the north of Naples.36 It seems likely
she was a descendant of L. Caedicius, a duovir in the same city mentioned in an

36
Manacorda (1985: 143–4) and Tchernia (1996: 209).
 Freedmen and Agency in Roman Business 233

epitaph of the late first century BC, and thus belonged to the local aristocracy of
Sinuessa, and possessed estates in the vicinity of the city.37 The amphorae
produced in her workshop may have been used to transport wine grown on
the family’s estates. During the same period, the freedman A. Caedicius Succes-
sus occurs in two inscriptions from Ostia, where he was a magistrate of both the
seviri augustales and the navicularii maris Hadriatici.38 This association of
shippers seems to have specialized in the distribution of wine, for two colleagues
of Successus, L. Scribonius Ianuarius and Cn. Sentius Felix, were also engaged in
the wine trade.39 Consequently, the gens Caedicia may very well have been
producing wine on their estates, making amphorae in workshops, and relying
on freedmen such as Successus to ship the wine to the capital and provincial
markets. Moreover, Festus notes that several Caediciae tabernae were sited along
the Via Appia.40 Perhaps the Caedicii were also involved in organizing local
distribution and had other freedmen and slaves running various inns.
A second case-study centres on modes of business organization in the
second-century Spanish oil trade. Several tituli picti on Dr. 20 oil amphorae
identify merchants belonging to the gens Aemilia. It seems, however, that two
different branches of the Spanish Aemilii had interests in the oil trade, the
MM. Aemilii and the LL. Aemilii. When the status of the merchants belonging
to these families is compared, a remarkable difference can be traced. The
MM. Aemilii all have Latin cognomina, none of which clearly points to a freed
status.41 The LL. Aemilii, on the other hand, are represented by a freedman,
Onesimus, and another merchant, whose cognomen is now illegible.42 An
explanation for this striking difference may be offered by Baetica’s monumen-
tal epigraphy. MM. Aemilii frequently occur in the epigraphy of cities bor-
dering the Guadalquivir, where most of the Spanish oil was produced, or in
Cadiz, from where the oil was exported.43 Yet not a single inscription explicitly
linked the MM. Aemilii to the local aristocracies. So far, it seems this family
was part of the wealthier levels of the plebs media, but never advanced to the
ranks of the elite. The LL. Aemilii, on the other hand, belonged in several
Baetican cities to the ruling classes and were seated in the city council of Carmona,
Montemayor, Arjona, and Villanueva del Rio.44 They no doubt possessed
several oil-producing estates in the vicinity of these cities and were marketing

37 38
AE 1986, 153. AE 1959, 149 and AE 1987, 191.
39 40
CIL 6.9682 and 14.409. De Salvo (1992: 433). Festus, Verb. Sign. 39L.
41
Rodríguez Almeida (1972: no. 3: Cutianus); CIL 15.4084a–b: Pastor; Rodríguez Almeida
(1979: no. 20: Rusticus).
42
Rodríguez Almeida (1994: no. 5: Onesimus). CIL 15.3695a–b (Alt[—]).
43
AE 1974, 380 (El Rubio); AE 1982, 559 (Jerez de la Frontera); CIL 2.1350 (Ronda); 2.1487
(Ecija); 2.1752–3 (Cadiz); 2.5539 (Peñaflor); 2.7.81 (Arjona); 2.7.812a (Fuente Obejuna); 2.7. 949
(Cabeza del Buey); ILS 6920 (Montellano).
44
CIL 2.1378; 2.1535; 2.2106 and AE 1972, 265 respectively. See also CIL 2.2150, a funerary
inscription of C. Pomponius Marullus, a duovir in Bujalance, dedicated by his relatives
L. Aemilius Avitus and C. Pomponius Lupus.
234 Wim Broekaert

surpluses, for which they relied on the family’s freedmen. The MM. Aemilii
apparently did not belong to this upper class and organized the distribution of
oil themselves. This way each family’s background may account for the
differences in organizing trade.
A final case study introduces the freedman M. Volcius Herma, an ivory
merchant who was running a shop in Capena, a city located slightly to the
north of Rome.45 His tombstone was decorated with a relief, depicting Venus,
Amor, and a large elephant, symbolizing his trade. When the gens Volcia is
traced in monumental epigraphy, it appears that, apart from a few inscriptions in
Italy, the Volcii were living mainly in Africa Proconsularis.46 We can imagine
that Herma’s family was somehow involved in the export of ivory to Italy and
that this freedman had settled in Capena to organize local distribution.

Freedmen produce or supply merchandise, which will

be sold by their patron

In this category, we find freedmen working in their patron’s workshop, while

the patron sells the merchandise in a shop, often adjacent to the workplace;
freedmen scouting the newly arrived ships in the port, in search of goods that
their patron can later resell; and freedmen shipping merchandise to their
patron’s stores.
These types of division of labour, though undoubtedly omnipresent in the
Roman business world, again leave few traces in the epigraphic record, for
occupational inscriptions usually do not distinguish between the level of pro-
duction or supply and that of resale. Inscriptions of a patron and his freedmen
jointly running a shop cite only one occupation, irrespective of the actual
contribution of each partner. In Rome, for instance, the freedwoman Cameria
Iarine dedicated a tomb to her fellow-freedman and husband Onesimus, their
patron Thraso, and their patron’s patron Alexander.47 All of them had been
working in the same shop as tailors of fine clothing on the vicus Tuscus (vestiarii
tenuarii de vico Tusco). Are we to assume that the senior artisan Alexander still
engaged in tailoring or did he perhaps concentrate on selling the clothing the
other freedmen were producing? A hint at the latter division of labour between
patron and freedman can be found in the epitaph of the engraver M. Canuleius
Zosimus.48 The tomb was dedicated by his patron, who explicitly added that
Zosimus had done nothing contrary to his wishes and never committed any
fraud, even though he had always had much gold and silver in his possession. The
phrasing suggests that the patron was supplying Zosimus with raw material and
making a profit from selling the finished products. Nevertheless, we cannot

45 46
CIL 11.3948. ILAfr. 479, 59; AE 1955, 126; CIL 8.1262, 1321.
47 48
CIL 6.37826. CIL 6.9222.
 Freedmen and Agency in Roman Business 235

completely rule out the possibility that Zosimus was marketing his produce in
the shop and giving his patron a share in the profits.
Merchants are also likely to have used their freedmen to import foreign
merchandise to their home market. This practice may be reflected in a legal
note by Paul, in which he discusses the case of a patron sending his freedman
into Asia for the purpose of buying purple.49 The text is silent about the
patron’s occupation or the destination of the merchandise, but nonetheless
confirms that this kind of task was entrusted to freedmen. An epigraphic trace
of this business organization might be found in the votive altar that
P. Arisenius Marius dedicated to the goddess Nehalennia, after having suc-
cessfully crossed the North Sea.50 Marius certainly was engaged in business,
for he thanks the goddess for having protected his merchandise (merces bene
conservatas) during the voyage. It is remarkable, however, that he fails to
mention his own occupation, while explicitly stating that he was a freedman of
P. Arisenius V[–]hus, a negotiator Britannicianus. The gentilicium Arisenius
seems to be unique but apparently points to a Germanic origin, for the name
may be derived from the cognomen Arusenus, attested in a third-century
inscription from Bonn.51 Apparently Marius considered himself to be not a
true merchant, but merely a shipper working for a merchant—namely, his
patron. It is possible that Marius’ patron was specializing in commerce with
Britain but was running his business from the German mainland and asked his
freedman to undertake the actual voyages to distant markets.

Freedmen and patron work closely together

in every single business stage

Here, freedmen remain closely tied to the patron and other coliberti, for they
constantly cooperate in a specific trade. They either worked together in a
single workshop, ran several workshops simultaneously, or jointly organized
business voyages. Each of these organizational schemes can be discovered in
occupational inscriptions.
First, the cooperation of a patron and one or several of his freedmen working
in the same shop must have been the key organization of many small-scale urban
enterprises, although, as mentioned before, no further division of labour can be
inferred from the inscriptions. For instance, D. Veturius Diogenes and his former
slaves Nicepor and Flora were all working together as purpurarii in a shop on the
Esquiline in Rome.52
Secondly, patrons deciding to organize business on a larger scale had their
freedmen running separate workshops with a modest degree of independence.

49 50 51
Dig. 34.2.4. AE 1983, 721. CIL 13.8066.
52
CIL 6.37820.
236 Wim Broekaert

Constant monitoring was impossible, as the patron was still engaged in

business himself, but regular visits and checking profits and expenses may
have been sufficient. The family business of A. Umbricius Scaurus, who
together with his slaves and freedmen dominated the Pompeian fish-sauce
industry, is a case in point: tituli picti on Pompeian amphorae indicate that
Scaurus and at least two of his freedmen, Agathopus and Abascanthus, were
each running an officina.53
Thirdly, the business organization in which patron and freedman travel
together to distant markets is reflected in the epitaph of Q. Dellius Optatus, a
merchant from Ravenna, who was doing business in southern Italy together
with his freedman Priscus.54 Optatus died in Venosa and was buried there by
Priscus. This kind of close cooperation is even reflected in Roman law, for,
according to Callistratus, the operae a patron was entitled to request from his
freedman included travelling together and transacting business.55

Freedmen are each separately responsible for all business

stages with their patron merely providing
infrastructure and finance

Freedmen belonging to this category organized their trade without constant

cooperation and supervision by the patron. They were each responsible for
purchasing or producing merchandise and marketing the goods. This, however,
does not exclude mutual assistance and cooperation by freedmen from the same
family. They obviously were best acquainted with their coliberti and knew about
their character and trustworthiness, so joining forces in the business world may
have been an attractive and efficient strategy. Nevertheless, the freedmen had to
answer separately to their patron.
In starting up a trade, they received some kind of support from their patron.
He may have provided capital investment or the necessary equipment and
infrastructure, such as a ship, a workshop, or a shop. These he could either
lease for a fixed sum or supply in exchange for part of the profits.
Contrary to business organizations in which the patron personally cooper-
ated with his freedmen and hence engaged in the same trade, here the patron
could either decide to focus on a single trade or diversify his investments.
The first option implies specialization and has several advantages to offer. For
instance, specific information and experience can be shared with all family
members, thereby consolidating their position in the business community.
Information on the best routes, trustworthy and cheap suppliers, stable markets,
or reliable customers could quickly be dispersed within the family network. In

53
CIL 4.2574a, 7110, 5689. Curtis (1984) and Étienne and Mayet (1991).
54 55
CIL 9.469. Dig. 38.1.38.1.
 Freedmen and Agency in Roman Business 237

this way, members theoretically would be able to optimize and accelerate

commercial decision-making. We might trace this kind of agency organization
in two small Roman inscriptions, mentioning two cloak-dealers, the coliberti
L. Arlenus Demetrius and L. Arlenus Artemidorus.56 They originated from
Cilicia and Paphlagonia respectively and had both been sold to L. Arlenus
Philogenus. The latter’s epitaph, dedicated by his freedmen, fails to mention a
profession, so we do not know whether Philogenus had been active in the textile
trade or not. Nevertheless, his two freedmen both specialized in the same trade,
apparently prompted to this profession by their patron. More telling, however, is
the information on freedmen’s involvement in the oil trade. Several merchant
families, such as the DD. Aticii, the LL. Memmii, and the LL. Segolatii, appear to
have relied almost completely on freedmen to ship Baetican oil to Rome.57
Within a time span of a mere ten years, these families stimulated a number of
freedmen to engage in the same trade, so it seems plausible that at least some of
them were coliberti backed by a patron with a major interest in the oil trade. So
far, none of their patrons has been identified as a businessman actively partici-
pating in commerce, so this business organization fits the typology quite well.
The second option allowed patrons to diversify their business interests. The
major advantages of this agency organization include the ability of allocating each
freedman to the profession he was best at and the possibility of spreading risks over
several trades. This strategy was in practice reserved for larger households with
major assets to manage and many freedmen to rely on. Although we can safely
assume that this kind of agency was part of aristocratic economic policies in
Roman business, evidence is scarce owing to the absence of accounts describing
aristocratic portfolios and the various investments made by patrons through their
freedmen. Literary sources, however, sometimes offer an insight in the manage-
ment of large fortunes by freedmen. Cicero’s letters, for instance, illustrate how
single tasks were entrusted to different freedmen. Hilarus, for instance, was staying
in Macedonia to collect debts made by Antonius, while another freedman, Eros,
recovered the rents of Cicero’s urban property. Many of his estate bailiffs presum-
ably were freedmen too.58 Even though this example mainly concerns financial
and agricultural management, one can easily imagine how the administration of
commercial investments was organized along similar lines. The famous passage in
Plutarch’s biography of Cato the Elder, describing how Cato was represented by
his freedman Quintio in a commercial partnership, may be a case in point, as one
can assume that Cato had more freedmen working in the many investments cited
by Plutarch.59 Epigraphy, on the other hand, rarely allows us to identify several

56
CIL 6.9675, 12331.
57
DD. Aticii: CIL 15.3735–40, 4057; Rodríguez Almeida (1972: no. 10; 1979: no. 29; 1983:
no. 2) and Ehmig (2007: no. 128). LL. Memmii: CIL 15, 3962–70; Rodríguez Almeida (1972:
no. 24–5, 66). LL. Segolatii: CIL 15.3993–9.
58 59
Shatzman (1975: 421–2). Plut. Cat. Mai. 21.6.
238 Wim Broekaert

freed agents working for one and the same patron in different trades. Moreover,
inscriptions usually reflect more small-scale and specialized enterprises and can
hence contribute little to understanding this specific kind of business organization.

Different business stages are entrusted to different freedmen,

with the patron providing merchandise
or cooperating in one or more stages

Patrons with an active interest in business may have wished to reduce risk and
uncertainty and enhance efficiency by exerting control over different stages of
an enterprise. Instead of using their freedmen to monitor a single stage, such
as production or distribution of merchandise, they assigned different tasks to
different freedmen, thus creating a network of agents, with each manager
being responsible for a single part of the business enterprise. Ideally, every
single stage (production or purchase of merchandise; transport; distribution)
would be supervised by either the patron or one of his freedmen. The benefit
of this business organization is evident: agents can now concentrate on a single
business stage, gather specific experience, and become specialized managers.
In this construction, freedmen often settled for longer periods of time or even
permanently in towns that were of particular importance for the enterprise.
One can imagine that some were living close to the spot where merchandise
was produced or could easily be purchased, while others settled in cities where
demand was particularly high in order to organize local distribution. This way,
agents would be able to gather relevant information for a long time, gain a
better insight into the organization of local markets, supplies, and demands,
and create long-standing business relationships. Especially on shuttle routes
between two ports or regions, these networks of freedmen were able to
maximize knowledge about markets, goods, routes, buyers, and sellers.
This occupational specialization and the creation of networks offered
several advantages and must have resulted in a more efficient enterprise.
First, permanent agents in distant cities had the opportunity to take over the
merchandise imported by their patron or fellow-freedmen immediately
and have a return cargo ready. Family members responsible for transport
would thus be less dependent on current market conditions and lose no time
waiting in the port searching for customers and suppliers. The period during
which capital outlays were tied up could hence be reduced. This particular
advantage of working with fixed agents has frequently been analysed as a
major time-saving technique and prerequisite for an increase of scale, in
particular in the study of eighteenth-century transatlantic trade.60 Secondly,

60
Walton (1967); North (1968); Price and Clemens (1987). See also French (1987) for a
comparable study of the Atlantic shipping industry.
 Freedmen and Agency in Roman Business 239

permanent agents were able to stock merchandise and wait for favourable
market conditions to achieve larger profits. Thirdly, they were firmly rooted in
local business communities and became acquainted with native producers,
entrepreneurs, and bankers. These connections may again ease transactions
for the whole family, as finding suppliers, customers, and investors was
organized by local agents. Itinerant businessmen, on the other hand, would
find it far more difficult to create this kind of network. Fourthly, these intra-
family networks may have stimulated the dissemination of reliable business
information. Members were more likely to receive trustworthy information
from other members than from businessmen they were not connected to.
Fifthly, dense networks of freedmen may discourage free-riding. A merchant
doing business with colleagues he had never met before and would perhaps
not encounter again in the near future was far more vulnerable to cheating
than an agent transacting with coliberti who were all linked to the same family.
Frequent communication and the fact that each was responsible for a different
stage made fraud easier to detect.
The commercial benefits of this agency model are thus self-evident. Epigraphy
now allows us to detect traces of this kind of business organization in the Roman
trading world.
First, we return to the Gallic merchant family of the Sexti Fadii. We have
already mentioned in the previous section that several tituli picti indicated
that Fadius Secundus Musa and at least three of his freedmen were shipping
Baetican olive oil to Rome. These tituli, however, offer very little information
on the actual organization of the oil trade or the identity of trading partners.
Yet the distribution of inscriptions mentioning other Sexti Fadii can be
revealing, for members of the gens Fadia only rarely adopt the praenomen
Sextus, and it is therefore very likely that the few Sexti Fadii who do feature
in inscriptions were somehow related.61 When we map the presence of Sexti
Fadii in monumental epigraphy, the geographical distribution clearly suggests
that these businessmen were able to rely on a large network of family members,
and freedmen in particular, who had settled in specific ports to assist in
managing the export and distribution of olive oil. Apart from two isolated
occurrences in Venosa and Verona, Sexti Fadii appear only in the epigraphy of
cities closely linked to the family’s origin or their involvement in the oil trade.62
A first cluster of inscriptions can be found in Narbonne, the family’s home
city.63 Secondly, the Fadii also occur in Baetica, for in Astigi the epitaph of Sex.
Fadius Lamyrus, obviously a freedman, has been discovered.64 This town was a
major fiscal control centre for the oil trade, where oil, originally stored in skins,
was poured into amphorae. Many tituli picti document this stage, as they
invariably start with the phrase ‘checked in Astigi’.65 The monument is dated

61 62
Di Stefano Manzella (1989). Venosa: CIL 9.422. Verona: CIL 5.3607.
63 64 65
CIL 12.4393, 4486, 4802, 4804. CIL 2.1495. Aguilera Martín (2000).
240 Wim Broekaert

to the first decades of the third century, but, as Lamyrus apparently had died at
the age of 50, he may very well have been involved in the oil trade organized by
the Fadii during the 160s and beyond. Perhaps he was responsible for the
purchase and packaging of the oil in Astigi and the onward transport to one of
Baetica’s ports, where his family members would collect the merchandise. The
third cluster is located in Ostia and Rome, the final destination of the oil
transport.66 The Sexti Fadii mentioned in these inscriptions and their relatives
may have assisted the family business by accepting, storing, and distributing
the merchandise and purchasing return cargoes. This geographical distribution
of inscriptions thus reflects the business interests of the Fadii and at least
suggests how the family was relying on freedmen, settled in distant ports, to
organize trade.
The second case study focuses on another merchant family we have already
encountered, the Publii Aufidii. We previously described how, around the
middle of the second century, P. Aufidius Fortis, a member of the aristocracy
of both Hippo Regius and Ostia, was selling grain in Ostia and introduced at
least two of his freedmen to the same trade. Yet it seems that the Aufidii were
using their family members to monitor an additional stage of the grain trade.
In AD 173, an honorary inscription was dedicated at Ostia to the grain
merchant M. Iunius Faustus by the African and Sardinian shipowners.67
Among the initiators was a certain P. Aufidius [—]us, who no doubt belonged
to the same family as Fortis. It seems at least plausible that this shipper was
one of Fortis’ freedmen, who was entrusted with a different task from that of
the grain merchants in Ostia. He may have been responsible for shipping the
merchandise, which was then handed over to Fortis and his freedmen and
distributed in the port. The Aufidii may even have controlled yet another
stage. As Fortis was a member of the city council in Hippo Regius, he must
have possessed some estates in the vicinity of this town, which most likely were
used to grow grain. One can thus imagine that part of the grain that the Aufidii
were shipping to Ostia had been produced on their own lands. This way, the
Aufidii apparently managed to monitor every single stage in the grain trade,
from the production in Africa and the shipment to Italy to the final distribu-
tion in Ostia. Fortis and his family hence created a perfect vertically integrated
business, which had many benefits to offer.68 Their involvement in production
provided them with information on expected yields, so shippers would know
how much additional grain they had to purchase to fill the hold and the agents
in Ostia could perhaps make an educated guess on the future availability of
grain. The Aufidii possessed their own ships, so there was no need to rely on
the services of strangers to have the grain shipped across the Mediterranean.
Once arrived in Ostia, the merchandise could immediately be handed over to

66
Ostia: CIL 14.995–6, 4563.5. Rome: AE 1989, 97; CIL 6.17651, 26537.
67 68
CIL 14.4142. Broekaert (2012b).
 Freedmen and Agency in Roman Business 241

Fortis and his freedmen. Like the Fadii, the Aufidii appear to have installed
managers in the area of production, the shipping industry, and the port of
destination.
These two case studies illustrate the possibilities of employing freedmen as
business agents and the potential advantages of simultaneously monitoring
different stages. One may argue that the increase in business efficiency these
two merchant families successfully introduced may very well have engendered
the families’ wealth and display of benefactions. Nevertheless, this kind of
business organization, depending on the widespread use of freedmen as
managers, obviously demanded a major initial investment: only families who
were able to purchase and manumit sufficient quantities of slaves to manage
every single stage could hope to initiate this agency mechanism successfully. It
comes as no surprise then that the businessmen who controlled this system,
P. Aufidius Fortis and Sex. Fadius Secundus Musa, each belonged to the local
aristocracy of their native town.

Different business stages are entrusted to different freedmen,

without the patron being personally engaged in business

This commercial organization obviously also requires a large household, for

the more freedmen a patron could engage in trade, the more business stages
could be monitored. The advantages of this strategy combine those of the two
previous typologies. First, the patron can allocate each agent to the profession
he excels in and use this division of labour to corner as many business stages as
possible. Secondly, he has the additional benefit of eliminating personal
engagement and completely relying on his network of freedmen. He interferes
only to provide finance or infrastructure.
As already noted, this kind of complex business organization, in which
several freedmen from the same patron operated simultaneously in a specific
trade and in which the patron remained anonymous, is difficult to trace in
epigraphy. We can hope only for inscriptions mentioning freedmen who most
likely belonged to the same family and followed different trades that may have
coincided with the various stages of a business enterprise. This obviously
implies that, even though a few inscriptions hint at this agency organization,
they will hardly ever provide clear-cut evidence.
The first case study will analyse the local business interests of the gens
Caedicia in the city of Rome. During the last decades of the republic or the first
of the empire, two former slaves of this family, M. Caedicius Eros and
M. Caedicius Iucundus, were working as goldsmiths in the Via Sacra.69

69
AE 1971, 41 and CIL 6.9207.
242 Wim Broekaert

Another freedman, M. Caedicius Faustus, operated as a merchant in the same

street.70 Nothing is known of their patron(s), but we can imagine that the
MM. Caedicii were running one or several workshops in the Via Sacra, maybe
belonging to their patron(s), with Eros and Iucundus apparently producing
gold jewellery and Faustus selling the finished objects.
The second example shows the application of a similar strategy in the long-
distance trade. The famous dossier of the Puteolean aristocratic gens Annia
illustrates how agency by freedmen, each responsible for a single business
stage, can even be traced in the organization of Roman trade with Arabia and
beyond.71 The central node in this commercial network is P. Annius Plocamus,
who, during the reign of Claudius, had contracted to farm the customs dues in
the Red Sea. During the same period, he must have used his freedmen to invest in
the import of eastern luxuries, for Pliny notes that one of them had accidentally
discovered Sri Lanka while sailing to Arabia.72 Moreover, two graffiti of AD 6,
discovered in the Egyptian Wadi Menih near Coptus, mention a certain Lysas,
slave of Annius Plocamus.73 The site was located close to the ancient caravan
route connecting Coptus and Berenice on the Red Sea coast and used to transport
Indian and Arabian merchandise to the Nile. The early date of the graffiti may
suggest that Plocamus’ network of slaves and freedmen was already active in the
first decade of the present era. The final chain in the network—the distribution of
the goods in the Mediterranean and Italy in particular—may be represented by
the presence of another freedman of Annius Plocamus in Puteoli, P. Annius
Eros.74 Each of these freedmen and slaves monitored a single aspect of the import
of eastern merchandise, but, together, the family’s agents cornered every stage in
this trade. Plocamus must have financed their transactions, but may also have
used his connections in the eastern tax system to assist and protect his agents.
This kind of long-distance trade had to rely on a dense family network of agents
and could thus be conducted only by a large and wealthy household. Moreover,
wealth was also a prerequisite to invest in the eastern trade, for huge sums were
involved in the import of luxuries. This way, prosperous families like the Annii
could combine both their fortune and their freedmen to build and finance a
commercial network stretching from the Mediterranean to the Indian Ocean.
Furthermore, onomastic research on Roman citizens in Egypt during the second
century AD has shown that the network of the Annii had survived the first century
and that the family probably continued to engage in trade with Arabia and India.

These six models of agency aptly illustrate the wide application of the system and
the variety of assignments entrusted to freedmen. The success of this strategy is
self-evident. Freedmen enabled their patrons to organize business far more
efficiently than working with hired personnel could ever have done. Patron

70 71
CIL 6.9662. Camodeca (1979); Rathbone (2003: 221–2).
72 73 74
Plin. NH 6.84. SEG 13, 614. CIL 10.2389.
 Freedmen and Agency in Roman Business 243

and freedmen had long been acquainted with each other, so they were never
trading with complete strangers whose trustworthiness could not be taken for
granted. The ready availability of possible agents thus saved the patron time
seeking and evaluating representatives. Large households that could rely on the
services of several freedmen were able either to diversify business interests as Cato
did; to intensify their engagement in a single trade, such as the Spanish merchant
families did in the oil trade; or to expand the geographical range of their invest-
ments by creating a widespread network of local agents, as the Puteolean Annii did.

THE FREQUENCY OF AGENCY: IN SEARCH

OF THE S ELF-MADE BUSINESSMAN

If agency by freedmen was such a convenient tool to organize trade, was there
any reason for patrons not to continue business relations with their former
slaves? Or, slightly rephrased, assuming that the patron was still alive, do we
even need to search for independent freedmen in Roman business at all?
This question is closely related to another, well-known debate.75 How
frequently did the imperial and municipal aristocracy rely on slaves and
freedmen to invest in business? Are we to assume that investments in trade
and industry commonly occurred in the aristocratic portfolios, or did only a
minority diversify its assets? Again, for obvious reasons, we cannot hope for
decisive statistical evidence listing the various properties and investments of
aristocratic families.76 Nonetheless, this subject has major implications for the
specific role of dependent and semi-dependent status groups—slaves and
freedmen—in business. The higher the proportion of aristocrats investing in
business with their freedmen operating as middlemen, the more likely it is that
the freed businessmen we encounter in inscriptions were working as agents. In
that case, legal independency must very often have coincided with economic
dependency. This is obviously not to say that behind every freedman in
business lurked a patron reaping the benefits of his managers’ successes, but
the occurrence of economically dependent freedmen or managers should have
increased according to the rate of aristocrats financing business enterprises.
Although historians nowadays accept that Roman elites were involved in
financing business enterprises, so far, however, no consensus on the ratio of

75
See now, also Mouritsen (2011) whose argument goes in a similar direction as the one
presented here.
76
See, however, Shatzman (1975) and Mratschek-Halfmann (1993). These prosopographical
studies, however, focus on literary sources, which offer information on senatorial and equestrian
wealth only. We still lack a detailed analysis of the fortunes of municipal aristocracies. Moreover,
valuable insights could also have been gained by confronting the literary evidence with archaeological
data, to identify the involvement of aristocratic families in the production of amphorae, tiles, etc. So
far, this kind of detailed analysis is available only for Baetica and Campania. For Baetica, see Haley
(2003); for Campania, see Los (2000).
244 Wim Broekaert

agriculture and business in aristocratic portfolios has been achieved.77 Some,

focusing on the rich evidence of Pompeii, claim that the local aristocracy
commonly engaged in commerce and that the category of economically
independent freedmen was only marginal.78 In this view, agency was ‘the
normal form of employment of freedmen in sea-trade: the freedmen were
agents and employees of the patron’.79 Others, however stress the difficulties
in linking freedmen in business to local aristocracies. Garnsey, for instance,
claimed that one could only ‘conclude that a sizeable number of freedmen
attained a position of independence or relative independence and wealth’.80
Yet the criteria suggested to identify these independent freedmen are vague
and hardly applicable to the surviving evidence. Garnsey suggested wealth and
positions of responsibility as the two main candidates.81
The first criterion seems a rather odd choice, for which businessman would
have the best chances for success: the completely independent freedman,
forced to start up, organize, and finance his enterprise all by himself, or the
freedman-agent, who was able to rely on investments and contacts provided
by his patron? Why, indeed, would dependency and wealth be mutually
exclusive? Agents working on commission or for a fixed wage were probably
better protected against business failure than independent entrepreneurs.
Moreover, economic dependency does not automatically imply that the patron
was deciding on the limits of enrichment. A patron who had entrusted some
money to his freedman to invest in seaborne trade created the possibility for
his agent to take aboard a small sample of merchandise and to trade on his
own account. The patron was actually financing the ‘infrastructure’, thus
allowing the agent to make a small profit on the side. This kind of incentive
must have been very common in maritime trade and was even institutional-
ized in the organization of the early modern Indian trade. The agents of the
India Companies who were directing the ships’ voyages, the so-called super-
cargoes, were officially allowed to take aboard their own merchandise, often
amounting to no less than 5 per cent of the cargo.82 Furthermore, Garnsey
added that possible indicators of this wealth of independent freedmen in-
cluded membership of a professional association or the erection of inscriptions.
If this were true, then the clear majority of freedmen in business should be styled
independent, for inscriptions are the only traces they have left for us. The case
studies previously discussed are sufficient proof that neither membership of a
collegium nor the ability to finance an inscription can be used as circumstantial
evidence to identify independent merchants. The freedmen of the gens Aufidia,
who had all joined a shippers’ or trade association and erected inscriptions but

77 78
López Barja de Quiroga (1991: 165–6). Andreau (1973); Mouritsen (2001).
79
Treggiari (1969: 103).
80
Finley (1973: 64, 78). Garnsey (1981: quotation from p. 369).
81 82
Garnsey (1981: 368). Morse (1921); Lee-Whitman (1982).
 Freedmen and Agency in Roman Business 245

clearly operated as agents, may be a case in point. The rise of two freed Aufidii in
the association of grain merchants can even be attributed to the lobbying of their
patron, who had been elected quinquennalis perpetuus of the same collegium. Yet
it is only when analysing these occupational inscriptions as different parts of a
larger business organization that a family network of dependent agents appears.
Take, for instance, the large and beautifully decorated inscription of the oil
merchant D. Caecilius Onesimus, who was trading in Rome during the Antonine
age and attained the office of viator, one of the most respected functions held by
the apparitores.83 According to Garnsey’s criteria, he may have been a successful
independent merchant, as both his career and the costs of erecting the stone
indicate at least a moderate level of wealth. However, when tituli picti of the same
period identify several oil traders all belonging to the same family and all former
slaves, one may conclude that it seems far more likely that Onesimus was but one
of the many agents in a family network of oil merchants.84
The second criterion, positions of responsibility, is equally problematic.
Garnsey has in mind ‘the ownership or management of or partnership in a
business, preferably a sizeable one or one that can be seen to be profitable’.85
Owning production facilities or commercial infrastructure (shops and ships)
may indeed point to independent freedmen, but does not necessarily exclude
cooperation with the patron and his family. The latter may still rely on the
marketing and transport services offered by his freedmen, as he was undoubt-
edly entitled to preferential treatment or profitable business deals. Partnership
is an equally dubious indicator. For instance, several members of the previ-
ously cited merchant family of the DD. Caecilii entered into a societas with
their own freedmen.86 In Capua, an epitaph dedicated to P. Octavius A.l.
Philomusus had been erected by Philargurus, his libertus et socius.87 Partner-
ships with freedmen were also mentioned by Cicero.88 The final part in
Garnsey’s definition, sizeable and profitable businesses, again discords with
the analysis of wealth as a possible indicator: the size and success of a business
would be increased by a patron’s financial backing and could, therefore, rather
be used as markers for dependency.
Apparently, identifying independent freedmen proves to be as difficult as
recognizing traces of dependent freedmen. Maybe Garnsey himself recognized

83
AE 1980, 98.
84
DD. Caecilii of the Antonine age, who were certainly freedmen: CIL 15.3751–3 (Calliphytus);
15.3754–5 (Chrysogonus); 15.3756, 3758–61 (a societas of Daphnus and Euelpistus); Rodríguez
Almeida (1994: no. 7: Nicephorus); CIL 15.3784 (Papia).
85
Garnsey (1981: 368).
86
CIL 15.3788–90 (Caeciliorum et lib(ertorum)). It is true that the reading lib(erorum) is
equally possible, but this seems less likely: tituli recording a cooperation between a father and his
children always use the words filius and filia. Cf., e.g. Blázquez Martínez and Remesal Rodríguez
(2003: no. 89). For the structure and rationale of commercial partnerships, see Broekaert (2011).
87 88
ILLRP 938. Cic. Parad. 46.
246 Wim Broekaert

the problems in applying his model to inscriptional evidence, for his discussion
lacks a concrete test of his two criteria. Nevertheless, D’Arms did try to apply
Garnsey’s model to Ostian epigraphy and the association of augustales in par-
ticular, the freedman elite parallel to the municipal aristocracy and mainly
consisting of wealthy freedmen who operated in trade and industry.89 He argued
that their level of wealth, membership in professional associations, the frequency
of marriages outside the own familia, and the fact that none of their patrons was
explicitly mentioned indicated that these freedmen ‘clearly operated independ-
ently’.90 But did they really? For the majority of the augustales we can neither
prove that they were doing business in their own interests nor that they operated
as agents. Some case studies, however, may be indicative. For instance, we have
already noted that the Gallic Fadii specialized in the Spanish oil trade and had
managers in Spain and Ostia. When Sex. Fadius Rufus, who clearly belonged to
this network, became a member of the Ostian augustales, this promotion should
be seen in connection to the wealth he had accumulated as local manager of his
family’s business interests.91 A recently discovered inscription from Pesaro can
corroborate this kind of network analysis. The fragmentary text mentions a
T. Ancharius T.l. G—], sevir augustalis and negotiator s[—], who erected the
stone for his late wife, Gavellia L.l. Ma[—].92 A wealthy freedman who married
someone from a different family and did not mention his patron seems, according
to D’Arms’s model, a plausible candidate for an independent businessman. Yet
when the occurrence of the TT. Ancharii is mapped, doubts arise. A first inscrip-
tion was also discovered in Pesaro and indicates that the freedman probably
belonged to the family of T. Ancharius T.f. Priscus and his son T. Ancharius T.f.
Priscianus, two members of the municipal elite and wealthy benefactors.93
A certain Ancharius Abascanthianus, who erected an epitaph for his late wife in
the same city, may also be linked to this family. It is interesting to note that,
outside Pesaro, epigraphy yields only a few texts mentioning TT. Ancharii: apart
from single attestations in Dacia and Africa Proconsularis, several Ancharii were
living in Salona, a commercial hub located more or less opposite Pesaro on the
other shore of the Adriatic.94 One of them, T. Ancharius Anthus, was a member of
the local seviri augustales. It does not seem too far-fetched to assume that these
two wealthy freedmen were local managers in a commercial network stretching
from Italy to the Balkans and beyond, especially because Salona was one of the
major ports connecting Italy to the Danube regions.95
Very similar problems are encountered when applying D’Arms’s criteria to
freedmen who held the office of sevir. This word may be a mere abbreviation

89 90
Duthoy (1974). D’Arms (1981: 144–6; quotation from p. 144).
91 92 93
CIL 14.4563.5. AE 2005, 484. CIL 11.6357.
94
Dacia: AE 1914, 116. Africa Proconsularis: ILTun 580a. Salona: CIL 3.2092, 2169, 2291,
2595, 9131.
95
For the importance of trade between Italy and Salona, see Tassaux (2004) and Glicksman
(2005).
 Freedmen and Agency in Roman Business 247

for sevir augustalis, but can also denote membership of another prestigious
association not connected to the imperial cult.96 Whether or not the seviri
belonged to the same social milieu as the Ostian liberti is thus difficult to find
out. Nevertheless, the fact that they managed to attain this office clearly
identifies them as belonging to an urban elite, which at least justifies a
comparison with the augustales. Two case studies may suffice. In Bologna,
an epitaph was erected for two coliberti, M. Papuleius M.l. Pudens and
M. Papuleius M.l. Primus, who were both members of the association of seviri
and iron merchants.97 They appear to have belonged to the same family as
M. Papuleius Latro, who held the quaestorship in the same city, for the rare
gentilicium is otherwise unknown.98 Whether these freedmen operated inde-
pendently or as agents cannot be discerned, but at least their business organ-
ization fits one of the typologies discussed earlier, in which an aristocratic
patron stimulates his freedmen to work in the same trade and thus increases
the scale of his business interests. A very similar analysis can be made for an
epitaph from Cordoba, dedicated to the sevir L. Vibius Polyanthus by his wife
Fabia Helpis and dated to the final quarter of the second century.99 Polyan-
thus’ name also occurs in a titulus pictus on a Spanish oil amphora, discovered
in Rome and precisely dated to AD 147.100 It seems this oil merchant meets the
criteria laid down by D’Arms rather well. However, several tituli from
the middle of the second century indicate that Polyanthus was but one of
the freedmen of the LL. Vibii engaged in the oil trade, so again we may assume
that he was operating as one of the family’s agents.101
We can hence safely conclude that, even for the freedman elites of the
augustales and seviri, the criteria for identifying independent entrepreneurs
hardly seem applicable to the inscriptional evidence. D’Arms’s statement, that
the majority of these elites were independent freedmen ‘in a position . . . where
success depended upon his own capacities, contacts and initiative’, still
remains to be proven.102
Perhaps we should then rely on circumstantial evidence, such as the con-
tinuity or discontinuity of powerful families who controlled the means of
production and distribution.103 It has been argued that, if freedmen were
indeed mainly employed as agents, one would expect the domination of a

96 97 98
For more details, see Duthoy (1978). AE 1922, 82. CIL 11.697.
99 100
CIL 2.7.329. CIL 15.4045.
101
Other LL. Vibii in the second-century oil trade: Blázquez Martínez and Remesal
Rodríguez (2003: nos 128, 162: Vibius [—]); CIL 15.3949; 3951–5, 3957–9 (a societas of
Restitutus and Viator). The LL. Vibii seem to have relied on the services of their freedmen
since the beginning of the first century, for an early Dr. 20 amphora discovered in Castra
Praetoria already records the name of an oil merchant L. Vibius Hermes (CIL 15.3668).
102
D’Arms (1981: 146). On the same page, the author calls the parallels between these
freedmen and Trimalchio ‘obvious and close’, but one should note that at the time Trimalchio
started to invest in business his former master was already deceased and had left him a fortune!
103
Meiggs (1960: 209); D’Arms (1981: 140–1); López Barja de Quiroga (1991: 168–9).
248 Wim Broekaert

handful of families in the urban economy to continue for several generations.

If, on the other hand, freedmen were allowed to operate more independently,
regular changes in the composition of the governing classes and sons of
wealthy freedmen entering the local aristocracy would be expected. We may,
however, wonder whether the dichotomy between ‘closed’ and ‘open’ elites
really matters in discussing freedmen’s (in)dependence. First, we have already
noted that, from an economic point of view, agency by freedmen is expected to
have benefited both parties. In this scenario, dependent freedmen would only
consolidate their patrons’ economic dominance, while independent freedmen
would hardly ever be able to capitalize on the best business opportunities,
would fail to accumulate wealth, and thus would never have gained access to
the governing class. In both cases, the outcome would be a rather closed urban
elite, as indeed applies to the majority of the Roman cities. Secondly, when
alterations in the composition of the local elites do appear, they should not
necessarily be attributed to independent freedmen and their families pervad-
ing the urban social structure. The openness of the municipal elites to sons of
freedmen, a phenomenon well attested in major commercial centres such as
Ostia and Puteoli, never distinguishes between descendants of dependent or
independent freedmen.104 Yet, because many merchant families from all over
the Mediterranean, such as the Gallic Fadii or the Spanish Caecilii, settled their
freed agents in Italian port cities and because from an economic point of view
these agents were most likely to accumulate personal wealth and enable social
mobility for their children, the chances are that newcomers in the municipal
elites of commercial hubs were rather descendants of dependent than of
independent freedmen.
One may thus conclude that, in cities such as Ostia and Puteoli, wealth
could buy you a ticket to the council, irrespective of the money’s origin or the
economic background of the families who had earned it. Yet, given the
cosmopolitan character of the population and the presence of freed managers
belonging to provincial business families, regular changes in the governing
classes may be attributed to agency rather than to the economic success of
independent freedmen.
This outcome, based primarily on dispersed inscriptional evidence, corres-
ponds quite well to the conclusion of similar research by Mouritsen, restricted
to the city of Pompeii, and to the institutional analysis of agency recently put
forward by Frier and Kehoe in the Cambridge Economic History if the Greco-
Roman World:
This method of managing businesses through friends or social dependants had
significant implications for the organization of the Roman economy. For one, it
tended to reinforce the strict social hierarchy that helped to preserve the economic

104
Gordon (1931).
 Freedmen and Agency in Roman Business 249
and social privileges of the landowning elite: there was little capacity for developing a
class of artisans let alone entrepreneurs who were fully independent of elite patron-
age or control. Successful freedmen who gained wealth as artisans or business
managers were ultimately dependent on a master or patron for an initial investment
in skills and capital, and they often remained socially bound at least to some degree
to their patron.105

CO NCLUSION

Will we ever be able to guesstimate the ratio of dependent and independent

freedmen and assess the true relevance of status and dependency? Probably
not. What we can do, however, is to try and determine the economic rationale
of both options. From this point of view, continuing economic relationships
had far more benefits to offer to both the patron and his freedmen than
immediately ending their cooperation upon manumission. Why indeed
would someone cede control and access to a ready labour force or allow
potential competitors onto the economic scene? From an entrepreneurial
point of view, manumission makes sense only when the patron would have
the opportunity to continue the cooperation with his former slaves. This clearly
does not imply that all or even the majority of freedmen in business were
dependent. Some, like the illustrious Trimalchio, would have been freed by
testamentary manumission and would never have known a patron. Others may
not have had the opportunity to continue the relationship with their patron,
whatever the reason may be. The patron may have terminated his investments
and no longer needed agents, or he may have found more qualified agents. We
can only conclude that, for patrons eager to invest in business, relying on
freedmen was the first and most rational option. Assuming they opted for the
easiest and quickest way to find agents—and there is no reason why they would
have failed to do so—agency by freedmen offered the best solution, or a decent
one at least. Relying on freedmen to manage business interests provided an
answer to the problem of information asymmetry and the difficulties of finding
honest agents, but never excluded the possibility that, in the end, it might
become clear that the patron had mistakenly entrusted them with his money.
Trimalchio himself relates how one of his guests, C. Iulius Proculus, used to
possess more than one million sesterces, but had lost his fortune to his cursed
freedmen, who had squeezed him dry. Nevertheless, in the same account
Trimalchio also makes clear that his own business interests were still managed
by freedmen.106 The advantages of having familiar agents ready at hand

105
Mouritsen (2001); Frier and Kehoe (2007; quotation from p. 134).
106
Petron. Sat. 43, 76.
250 Wim Broekaert

apparently outweighed the risks, which at any rate would still be much smaller
than working with complete strangers. The outcome must hence be that,
logically, as long as patrons were still alive and continued to invest in business,
in these families dependent freedmen would be expected to be far more
numerous than independent ones. This business organization, however, did
not necessarily result in ruthless economic exploitation of freedmen. The
majority of the freedmen-agents may actually have gained from the agency
relationship and worked in their own and their patron’s interests. If the
prospect of manumission was the incentive for a slave to work in his master’s
interests, then the potential benefits of a prolonged cooperation may have
urged freedmen to operate as agents. Perhaps the many success stories of
prosperous freedmen and the stock character of the wealthy but uncivilized
libertus as a nouveau riche in Roman satire mirror this continued relationship.

REFERENCES

Aguilera Martín, A. (2000). ‘Los tituli picti delta del convento astigitano en el primer
tercio del s.III d.C.’, in Actas del Congreso Internacional Ex Baetica Amphorae.
Conservas, aceite y vino de la Bética en el Imperio Romano (Écija y Sevilla, 17 al 20 de
Diciembre de 1998). Écija, 1231–40.
Andreau, J. (1973). ‘Remarques sur la société pompéienne (à propos des tablettes de
L. Caecilius Jucundus)’, Dialoghi di Archeologia 7: 213–54.
Andreau, J. (1995). ‘Présentation: Vingt ans après L’Économie antique de Moses
I. Finley’, Annales ESC 50: 947–60.
Aubert, J.-J. (1994). Business Managers in Ancient Rome: A Social and Economic Study
of Institores, 200 B.C.–A.D. 250. Leiden.
Blázquez Martínez, J. M., and Remesal Rodríguez, J. (2003). Estudios sobre el Monte
Testaccio (Roma) III. Barcelona.
Blázquez Martínez, J. M., and Remesal Rodríguez, J. (2007). Estudios sobre el Monte
Testaccio (Roma) IV. Barcelona.
Bonsangue, M. L. (2001). ‘Les Relations commerciales entre Pouzzoles et l’Égypte au
IIe siècle ap.J.-C.’, CCG 12: 199–212.
Broekaert, W. (2011). ‘Partners in Business: Roman Merchants and the Potential
Advantages of being a collegiatus’, AncSoc 41: 221–56.
Broekaert, W. (2012a). ‘Joining Forces: Commercial Partnerships or societates in the
Early Roman Empire’, Historia 61/2: 221–53.
Broekaert, W. (2012b). ‘Vertical Integration in the Roman Economy: A Response to
Morris Silver’, AncSoc 42: 109–25.
Camodeca, G. (1979). ‘La gens Annia puteolana in età giulio-claudia; Potere politico e
interessi commerciali’, Puteoli 3: 17–34.
Castrén, P. (1975). Ordo Populusque Pompeianus: Polity and Society in Roman Pompeii.
Rome.
Curtis, R. I. (1984). ‘A Personalized Floor Mosaic from Pompeii’, AJA 88/4: 557–66.
 Freedmen and Agency in Roman Business 251
D’Arms, J. H. (1977). ‘M. I. Rostovtzeff and M. I. Finley: The Status of Traders in the
Roman World’, in J. H. D’Arms and J. W. Eadie (eds), Ancient and Modern: Essays
in Honour of Gerald F. Else. Ann Arbor, 159–79.
D’Arms, J. H. (1981). Commerce and Social Standing in Ancient Rome. Cambridge.
De Salvo, L. (1992). Economia privata e pubblici servizi nell’ impero romano: I corpora
naviculariorum. Messina.
di Stefano Manzella, I. (1989). ‘Nuova iscrizione sepolcrale con doppia datazione
consolare’, ZPE 76: 262–6.
Duthoy, R. (1974). ‘La Fonction sociale de l’augustalité’, Epigraphica 36: 134–54.
Duthoy, R. (1978). ‘Les Augustales’, Aufstieg und Niedergang der römischen Welt II
16/2: 1254–1309.
Ehmig, U. (2007). ‘Tituli picti auf Amphoren in Köln’, KJ 40: 215–322.
Étienne, R., and Mayet, F. (1991). ‘Le Garum à la mode de Scaurus’, in J. M. J. M. Blázquez
and S. Montero (eds), Alimenta: Estudios en homenaje al dr. Michel Ponsich. Madrid,
187–94.
Fabre, G. (1981). Libertus: Recherches sur les rapports patron-affranchi à la fin de la
République Romaine. Rome.
Finley, M. I. (1965). ‘Technical Innovation and Economic Progress in the Ancient
World’, Economic History Review 18: 29–45.
Finley, M. I. (1973). The Ancient Economy. Berkeley and Los Angeles.
Földi, A. (1996). ‘Remarks on the Legal Structure of Enterprises in Roman Law’, RIDA
43: 179–211.
French, C. J. (1987). ‘Productivity in the Atlantic Shipping Industry: A Quantitative
Study’, Journal of Interdisciplinary History 17: 613–38.
Frier, B. W., and Kehoe, D. P. (2007). ‘Law and Economic Institutions’, in W. Scheidel,
I. Morris, and R. Saller (eds), The Cambridge Economic History of the Greco-Roman
world. Cambridge, 113–43.
Gamauf, R. (2009). ‘Slaves Doing Business: The Role of Roman Law in the Economy of
a Roman Household’, European Review of History 16: 331–346.
Garnsey, P. (1981). ‘Independent Freedmen and the Economy of Roman Italy under
the Principate’, Klio 63: 359–71.
Glicksman, K. (2005). ‘Internal and External Trade in the Roman Province of Dalmatia’,
Opuscula Archaeologica 29: 189–230.
Gordon, M. L. (1931). ‘The Freedman’s Son in Municipal Life’, JRS 21: 65–77.
Haley, E. W. (2003). Baetica Felix: People and Prosperity in Southern Spain from
Caesar to Septimius Severus. Austin, TX.
Héron de Villefosse, A. (1915). ‘Deux armateurs narbonnais: Sex. Fadius Secundus et
P. Olitius Apollonius’, BSAF 74: 153–80.
Joshel, S. R. (1992). Work, Identity and Legal Status at Rome: A Study of the Occupa-
tional Inscriptions. Oklahoma.
Kirschenbaum, A. (1987). Sons, Slaves and Freedmen in Roman Commerce. Jerusalem
and Washington.
Lee-Whitman, L. (1982). ‘The Silk Trade: Chinese Silks and the British East India
Company’, Winterthur Portfolio 17: 21–41.
López Barja de Quiroga, P. (1991). ‘La dependencia económica de los libertos en el
Alto Imperio Romano’, Gerión 9: 163–74.
252 Wim Broekaert
Los, A. (2000). ‘Les Affaires “industrielles” des élites des villes campaniennes sous les
Julio-Claudiens et les Flaviens’, MEFRA 112: 243–77.
MacMullen, R. (1982). ‘The Epigraphic Habit in the Roman Empire’, AJPh 103: 233–46.
Manacorda, D. (1985). ‘Schiavo “manager” e anfore romane: A proposito dei rapporti
tra archeologia e storia del diritto’, Opus 4: 141–151.
Meiggs, R. (1960). Roman Ostia. Oxford.
Morse, H. B. (1921). ‘The Supercargo in the China Trade about the Year 1700’, English
Historical Review 36: 199–209.
Mouritsen, H. (2001). ‘Roman Freedmen and the Urban Economy: Pompeii in the First
Century AD’, in F. Senatore (ed.), Pompei tra Sorrento e Sarno: Atti del terzo e quarto
ciclo di conferenze di geologia, storia e archeologia, Pompei, gennaio 1999–maggio 2000.
Naples, 1–27.
Mouritsen, H. (2005). ‘Freedmen and Decurions: Epitaphs and Social History in
Imperial Italy’, JRS 95: 38–63.
Mouritsen, H. (2011). The Freedman in the Roman World. Oxford.
Mratschek-Halfmann, S. (1993). Divites et Praepotentes: Reichtum und Soziale Stellung
in der Literatur der Prinzipatszeit. Stuttgart.
North, D. C. (1968). ‘Sources of Productivity Change in Ocean Shipping, 1600–1850’,
Journal of Political Economy 76: 953–70.
Price, J., and Clemens, P. G. E. (1987). ‘A Revolution of Scale in Overseas Trade:
British Firms in the Chesapeake Trade, 1675–1775’, Journal of Economic History 47:
1–43.
Rathbone, D. (2003). ‘The Financing of Maritime Commerce in the Roman Empire, I–II AD’,
in E. Lo Cascio (ed.), Credito e moneta nel mondo romano: Atti degli incontri capresi di
storia dell'economia antica (Capri 12–14 ottobre 2000). Bari, 197–229.
Remesal Rodríguez, J. (2000). ‘L. Marius Phoebus, mercator olei hispani ex provincia
Baetica: Consideraciones en torno a los términos mercator, negotiator y diffusor
olearius ex Baetica’, in G. Paci (ed.) Epigraphai. Miscellanea Epigrafica in onore di
Lidio Gasperini. Rome, 781–97.
Rodríguez Almeida, E. (1972). ‘Novedades de epigrafia anforaria del Monte Testaccio’,
in P. Baldacci, J. Kapitän, and N. Lamboglia (eds), Recherches sur les amphores
romaines. Rome, 107–242.
Rodríguez Almeida, E. (1979). ‘Monte Testaccio: I mercatores dell’olio della Betica’,
MEFRA 91: 873–975.
Rodríguez Almeida, E. (1983). ‘Altri mercatores dell’olio betico’, Dialoghi di Archeologia
1: 79–86.
Rodríguez Almeida, E. (1991). ‘Ánforas olearias béticas: Cuestiones varias’, in
J. M. Blázquez and S. Montero (eds), Alimenta: Estudios en homenaje al dr. Michel
Ponsich. Madrid, 243–61.
Rodríguez Almeida, E. (1994). ‘Scavi sul Monte Testaccio, novità dai tituli picti’, in
Epigrafia della produzione e della distribuzione: Actes de la VIe Rencontre franco-
italienne sur l’épigraphie du monde romain. Rome, 111–31.
Ruffing, K. (2008). Die berufliche Spezialisierung in Handel und Handwerk: Untersu-
chungen zu ihrer Entwicklung und zu ihren Bedingungen in der römischen Kaiserzeit
im östlichen Mittelmeerraum auf der Grundlage der griechischen Inschriften und
Papyri. St Katharinen.
 Freedmen and Agency in Roman Business 253
Schleich, T. (1983). ‘Überlegungen zum Problem senatorischer Handelsaktivitäten,
T. I: Senatorische Wirtschaftsmentalität in moderner und antiker Deutung’, MBAH
2/2: 65–90.
Shatzman, I. (1975). Senatorial Wealth and Roman Politics. Brussels.
Stelzenberger, B. (2008). Kapitalmanagement und Kapitaltransfer im Westen des
römischen Reiches. St Katharinen.
Tassaux, F. (2004). ‘Les Importations de l’Adriatique et de l’Italie du Nord vers les
provinces danubiennes de César aux Sévères’, in Dall’Adriatico al Danubio, L’Illirico
nell’età greca e romana, Convegno internazionale (Cividale 2003). Cividale, 167–205.
Tchernia, A. (1996). ‘Maesianus Celsus et Caedicia Victrix sur des amphores de
Campanie’, in Les Élites municipales de l'Italie péninsulaire des gracques a Néron:
Actes de la table ronde de Clermont-Ferrand (28–30 novembre 1991). Naples and
Rome, 207–11.
Treggiari, S. (1969). Roman Freedmen during the Late Republic. Oxford.
Verboven, K. (2002). The Economy of Friends: Economic Aspects of amicitia and
Patronage in the Late Republic. Brussels.
Wacke, A. (1994). ‘Die adjektizischen Klagen im Überblick, I: Von der Reeder- und der
Betriebsleiterklage zur direkten Stellvertretung’, ZRG 111: 280–362.
Waldstein, W. (1986). Operae libertorum: Untersuchungen zur Dienstpflicht freigelassener
Sklaven. Stuttgart.
Wallace-Hadrill, A. (1994). Houses and Society in Pompeii and Herculaneum. Princeton.
Walton, G. M. (1967). ‘Sources of Productivity Change in American Colonial Shipping,
1675–1775’, Economic History Review 20: 67–78.
 11

The Social Organization of Commerce and

Crafts in Ancient Arles: Heterogeneity,
Hierarchy, and Patronage
Nicolas Tran

In an article published in 1985 and entitled ‘Economic Modernity and Status of

Businessmen’, Jean Andreau discussed the social and legal position of economic
agents in the Roman world. In the historiographical context of the revival of the
Bücher–Meyer controversy, he commenced by exposing the opposition between
the two schools of thought that one traditionally refers to as ‘primitivist’ and
‘modernist’. For Moses Finley and his followers, the world of ‘work’ was popu-
lated with small people.1 The majority were slaves or freedmen, despised by the
elite. Their social inferiority was symptomatic of the marginal place of crafts and
commerce in the ancient economy. Conversely, scholars inspired by Mikhail
Rostovtzeff, particularly John D’Arms, defended the idea that Roman commerce
fostered the emergence of a category of businessmen of a significant calibre,
symbolized by the rich, independent freedman.2
Yet Andreau did not limit himself to noting and describing the controversy:
he pointed out the contradictions inherent to both positions, thus showing the
urgent need to escape from their opposition:
In this regard, the conclusions of Finley and D’Arms, while they contradict each other,
are also contradictory in themselves. Finley, who wants to insist on the social humility
of financiers, but also on the limitations imposed on the economy by the existence of
ties of dependence, describes freedmen as very dependent on their patron, but he
tends to minimize the financial interests of that same patron, without noting that this
means that decision-making processes become incomprehensible, and that one can
no longer see who benefits from the profits. Conversely, D’Arms, who wants to reduce

1
Finley (1985: 60–1).
2
Rostovtzeff (1957: pp. xiv, 142–79); D’Arms (1981: 146–8). On the notion of the independ-
ent freedman, see Garnsey (1981). Cf. Garnsey (1998) with an updated bibliography.
 Commerce and Crafts in Ancient Arles 255
the importance of juridical and social differences, but hesitates to show that the elites
were involved in commerce and financing, does not stop emphasizing the autonomy
of freedmen and the role of patrons in financing them. However, if, in the end, all
freedmen were independent (legally and financially), what role would be left for the
patrons to play? One has to escape from the opposition of these two models . . . 3
If we phrase this differently, the strong roots of craftsmen and traders in the
servile population suppose the involvement of masters and higher-ranking
patrons in artisanal and commercial activities, while independence is at odds
with the status of freedman. This argues against the emergence of a class of
business people made up of freedman parvenus. If the two leading models are so
unsatisfactory, how are we to escape from their opposition? First of all, the
question of the wealth or humble status of people involved in crafts and trade
can be understood only through detailed prosopographical analysis. In reality,
the world of work was diverse: like the urban plebs and the servile population in
general, it was finely stratified. Slaves and freedmen did not make up the totality
of the world of crafts and trade, nor did they form a homogeneous whole. Still,
they all had in common the fact that they had a master or a patron. One thus has
to explain the vertical relations that tied artisans and traders to their master or
patron and their impact on the economy and on social hierarchies.4
It is impossible to address all these issues in this short chapter, but the problem
posed by the shortcomings of both primitivist and modernist ideas about
craftsmen and traders will form its background. The primary focus will be on
the professional associations of Arles, in the second and third centuries AD. To
enable us better to understand this port society, the seviri Augustales corporati
will also be included in the analysis. The corpora fostered the social and civic
integration of their members, and preserved them from the marginality to which
their individual economic and legal condition could have consigned them.

HIERA R CHICAL COMMUNITIES: T HE MELTING

POT OF A PLEBEIAN ELITE

The epigraphic record of Arles attests a fair number of professional associ-

ations, which is a characteristic feature of large places of commerce in the
western part of the Roman Empire.5 The texts of these inscriptions are given at

3
Andreau (1985: 389).
4
This theme is developed in H. Mouritsen’s chapter on the place of Roman freedmen in
economic life (Mouritsen 2011: 206–43). This book was not yet available in France when
I planned this chapter (during summer 2011) and, as a result, I integrated it later into my
research. Its conclusions, especially its criticism of the model of the rich and independent
freedman (pp. 228–34), and mine are very similar.
5
Waltzing (1895–1900) is still essential.
256 Nicolas Tran

the end of this chapter. In the colony of Arles, we know of corpora of seafarers
(nos 1–8), carpenters (9–14), utricularii (15–18), ship-builders (19–21), cen-
tonarii (19), and lenuncularii (22–3). To these should be added an association
of lapidarii Almanticenses (24), whose provenance is unknown, and the ship-
pers of the Durance (7, 16, 25), of whom we know two examples and one
patron in relation to Arles. The partiarii, who appear on an altar dedicated to
the Manes of a slave (no. 26; Fig. 11.1), are enigmatic. Because the monument
was decorated with the representation of a ship, and because partiarius derives
from pars, which alludes to the act of dividing and thus to distributing, Marc
Heijmans has thought of partiarii as dockers.6 Yet this is only one hypothesis
among others. It is possible that the partiarii shared their burial grounds and
thus formed a purely funerary association, instead of a professional one.7 In
any case the list of collegia from Arles may grow longer in the future through
new discoveries: the two inscriptions of the corpus lenunculariorum were taken

Fig. 11.1. Funerary altar of Hermias (inscription 26). Musée de l’Arles antique. Photo:
Nicolas Tran.

6 7
Heijmans (2003). Christol (2010b: 413–14); Laubry (2012: 111–12).
 Commerce and Crafts in Ancient Arles 257

out of the Rhône only in 2007.8 This corpus is known through an altar
dedicated to its Genius and through the pedestal of a statue of Neptune,
which was donated by a certain Publius Petronius Asclepiades. This person
does not highlight any collegial title, so it makes sense to consider him an
external benefactor of the corpus. No lenuncularius has thus been attested—nor
do we know of an individual centonarius. In fact, only the patronage of Gaius
Paquius Pardalas of the corpus centonariorum attests the existence of this
community. On the other hand, the names of fifteen corporati have been
inscribed on epitaphs, which make up a small prosopographical series.9
One of the typical features of the evidence concerns the prominent represen-
tation of the dignitaries of the corpora. On his epitaph, Marcus Iunius Messianus
mentions that the utricularii had made him magister four times (no. 17). Among
the six known fabri tignuarii, Lucius Aventius (or Aventinius) Avitianus, Quintus
Candidius Benignus, and Caius Publicius Bellicus report the same title (nos 9–10,
14). For their part, an anonymous from Saint-Gilles, Caius Iulius Pomp[—] and
Marcus Frontonius Euporus were curatores of their corpus: the first of that of the
utricularii (no. 18), the second of the fabri navales (no. 20), the third of the
navicularii marini (no. 7). Such magistracies brought with them certain costs
ob honorem, and, in general, their award reflected the social position of the
benefactors, singling out the individuals and families that were seen as the most
respectable.10 Further, among the simple corporati, Lucius Iulius Secundus was at
once utricularius at Arles and shipper of the Durance (no. 16). This combination
can be explained in different ways. Secundus could have exercised both occupa-
tions successively. It is also possible that he practised only one of them but
maintained business relations with other corporati, who, as such, accepted him
as a member of their association. Finally, Secundus may have been shipper and
utricularius at the same time. The second profession is very enigmatic. Given the
places where the inscriptions mentioning them have been discovered, one has to
accept, at the very minimum, that the activity of utricularii was related to the
transport of merchandise and to transhipment points between river transport and
land transport. Some historians have gone so far as to define utricularii as
muleteers; others prefer to consider them, still, as river transporters.11 I have
not solved this enigma, but I would just emphasize that the occupations of
shippers and utricularii were two different things, and beyond that leave it at
the level of hypothesis. So, Lucius Iulius Secundus could have been involved in the
transport of merchandise—wine, for example, if one thinks about the relief of
Cabrières d’Aygues—relying on several modes of transport.12 He should, in any

8
Christol and Fruyt (2009).
9
ILGN 116, of which the reading is very uncertain, has been excluded from the study.
10
On the relation between collegial honores and social rank, see Tran (2006: 164–74).
11
On this complex issue, see Verdin (2005), who gives earlier bibliography.
12
For the relief of Cabrières d’Aygues, see Espérandieu (1907–66: ix. 6699).
258 Nicolas Tran

case, be seen not as a simple muleteer or a modest boatman, but as a freight

specialist, capable of organizing commercial relations on a regional scale.13 Only
then is the relative wealth suggested by his epitaph comprehensible: Secundus left
200 denarii to the utricularii to sacrifice at his tomb each year. The sum of this
foundation is limited, but it nonetheless suggests that Secundus did not belong to
the most humble workers of the port.
The corporati known from epigraphy were no marginal, poor people,
neither at Arles nor elsewhere. On the contrary, the position that many of
them occupied in their association seems to reflect their success in their
respective economic activities. Of course, this does not mean that there were
only a few poor workers in the port of Arles. Epigraphy simply fails to shed
light on these milieux, as it concentrates on the upper strata of society, and on
the upper strata of the plebs itself. This, at least, is the norm that seems to apply
to the epigraphy of professional associations, as Michel Christol has empha-
sized for Gallia Narbonensis.14 Above all, no association can be considered
purely on its own terms: besides the internal hierarchy of each community,
there were big differences between organizations.

A H I E R A R C H Y O F CORPORA AND OF
ECONOMIC ACTIVITIES

The harbour society of Arles was no homogeneous ensemble, and the occupations
and associations that emerged from it were unequal in dignity and prestige.15 As
one might expect, the navicularii occupied the top of the pyramid. All epigraphic
indications point in the direction of regarding maritime shipowners as the
exception rather than the norm. The inevitable consequence of this observation
is that it is impossible to base an understanding of the occupational structure of
Arles on the navicularii.
As far as the social composition of the corpora of Arles is concerned, the corpus
of the navicularii marini is the only known association that counted seviri
Augustales among its members. In this quality, Lucius Secundius Eleuther and
Marcus Frontonius Euporus advanced to a level of prestige close to that of the
colonial elite (nos 4 and 7). The latter was elevated to the same dignity by the
cities of Arles and Aix-en-Provence. As their Greek cognomina suggest, the role
of these two persons as seviri had to compensate for the fact that—as freedmen—
they were ineligible for municipal magistracies.16 These former slaves, however,

13
An anonymous individual from Saint-Gilles (no. 18) probably had the same professional
status: he was a curator of the utricularii Arelatentes and of the nautae shipping on the rivers
Ouvèze and Archèche. On this character, see Burnand (1971); Provost (1999: 625).
14 15 16
Christol (2010a: 536–46). Christol and Tran (2014). Duthoy (1974).
 Commerce and Crafts in Ancient Arles 259

were rich: access to the function of sevir supposed, effectively, the payment of a
summa honoraria of several thousands of sesterces, and, on top of that, the
undertaking of acts of generosity.17 Besides, the shippers of the Durance and the
utricularii of Ernaginum must have expected the generosity of their patron.18 As
protector of these two communities, Marcus Frontonius Euporus reversed
the patronage relationship that tied him to his former master—and compensated
for it. Yet whose dependant was he? Michel Christol has discussed this question
in an article that, forty years after it was published, still remains fundamental.19
Secundius and Frontonius correspond to two gentilicia of patronymical origin,
created in Gaul from the cognomina Secundus and Fronto, which were very
popular amongst the Celtic population. In fact, these nomina were common
among notables in the eastern part of the territory of Nîmes. Simultaneously, it is
possible that the ties of Euporus with Aix-en-Provence stemmed from the
origin of the familia where he was born. From these observations, Christol
built a convincing hypothesis according to which Lucius Secundius Eleuther
and Marcus Frontonius Euporus must be regarded as representatives of a
harbour elite with ties to the powerful families of the hinterland. Through such
individuals, the landowners of the countryside invested part of their wealth in the
harbour economy of Arles. As patrons, they probably granted loans to their
freedmen, as happened in the imaginary case of Trimalchio, or formed a joint
venture with them.20 Some will have profited abundantly. However, the marine
shippers of Arles collectively do not fit into this picture. Quintus Capitonius
Probatus was probably one of them, even if his epitaph from Lyons refers to
him as a navicularius marinus without referring to Arles (no. 8).21 Originally
from Rome, he was sevir at Lugdunum and at Puteoli. In fact, the links that
Capitonius Probatus constructed between his native Italy and the Rhône valley
necessarily passed through Arles, where the naviculari are the only ones known
as marini. As a sevir in more than one city, like Euporus, Probatus was a rich
businessman.
The social origins and the wealth of the marine navicularii seem to have had
a profound influence on the organization of their corpus. As far as we know,
the navicularii were the only corporati that were epigraphically visible in
public space. At an unspecified date in the first or second century, they erected
a statue for Cnaeus Cornelius Optatus, duumvir and a priest of the colony, on
the forum adiectum of the city (no. 2). Subsequently, probably around the turn
of the second and third centuries, they honoured an eques in charge of the
annona for the province of Gallia Narbonensis and Liguria: Cominius Bo[---]

17
Duncan-Jones (1974: 152–4, 215–17); Duthoy (1981: 1267, 1279).
18 19
Tran (2006: 450–9). Christol (1971).
20
On the credit, see Petronius, Sat. 76.9. On joint ventures, see the attitude of Crassus in Cic.
Par. 6.46.
21
According to Rougé (1965), Probatus exercised his occupation from Lyons, which is
improbable.
260 Nicolas Tran

Agricola Aurelius Aper (no. 1).22 The base was installed in the harbour quarter
of Trinquetaille. In this way, because of the status of their occupation and the
fact that they signed contracts with the state, the navicularii formed the only
known professional association from Arles that developed patronage ties
outside the world of labour.23 The shippers of the Durance and the utricularii
of Ernaginum had to make do with a patron who was distinguished, but who
was no member of the leading class (no. 7). Likewise, the lenunculari benefited
from the generosity of a probable freedman, Publius Petronius Asclepiades,
who must have belonged to the harbour society rather than to the local elite
(no. 22). By contrast, the navicularii were in a position to extend their relations
to networks outside their milieu. Their corpus was wealthy: it seems to have
possessed a statio where an apparitor (no. 5) operated as well as some slaves:
Quintus Navicularius Victorinus could have been their freedman (no. 6): the
official right to use the ius manumittendi was granted to a minority of
privileged colleges by Marcus Aurelius.24
The wealth and prestige of the navicularii resulted most notably from the
services they performed for the annona. This is attested by two, probably
contemporary, documents: the honorific base dedicated to Cominius Bo[---]
ius Agricola Aurelius Aper and a famous inscription from Beirut, preserved on
the back of a small, chiselled slab (nos 1, 3). Between 198 and 203, the
navicularii complained before the prefect of the annona Claudius Iulianus
about fraud, of which they claimed to have become victims.25 Part of the fiscal
grain that they transported to the capital was returned. Threatening to break
their contracts, and following the example of other professionals serving the
annona, they obtained an agreement from the prefect that public authorities
would regain control over the organization of Rome’s supply. Firm orders
were given to a procurator placed under command of the prefect. It is precisely
in the light of the relations with the annona, at the turn of the second and third
centuries, that the shippers from Arles were described as the navicularii
marini Arelatenses quinque corporum. On his epitaph, engraved during the
second century, Marcus Frontonius Euporus was described as the curator of a
single corpus (no. 7). Starting from this, two hypotheses are possible. Either the
corpus was divided later, according to the model of the five corpora of
lenuncularii of Ostia.26 However, it is difficult to know the logic behind this
division. The alternative is that, at the turn of the second and third centuries,
the corpus of Arles and four other corpora, having their seat elsewhere in the
empire, were treated as a joint unit by the Roman state. I would favour the

22
The date is discussed by Christol and Demougin (1984) and Alföldy (1986). I follow the
opinion of the former.
23
On the diverse background of patrons of professional associations, see Clemente (1972).
24 25
Dig. 40.3.1 (Ulpian, ad Sab. 5). Virlouvet (2004); Corbier (2006: 233–56).
26
This is, for example, the hypothesis of Salvo (1992: 409), based on CIL 14.352 and 4144.
 Commerce and Crafts in Ancient Arles 261

second possibility, following Michel Christol, and would add an argument in

support of his position. A problem is posed by the eastern origin of the slab. As
Catherine Virlouvet has shown, it is likely that the plaque, from which the slab
comes, was set up in an eastern Mediterranean port.27 Beirut appears too
small; Tyre would be more suitable; Alexandria still more so. The transport of
a decorative object from Egypt to Lebanon, during the early modern era, is not
at all implausible. Indeed, if they had formed one of the five corpora that stood
in direct contact with the state, the navicularii of Alexandria would have had
an interest in putting up the decisions of Claudius Iulianus publicly: the
problems raised by the navicularii of Arles ended in a general resolution
that the five corpora could claim.28 This reconstruction allows us to make
the best of a hypothesis that, though it cannot be excluded, is still fragile: there
is no indication that ports in the Greek-speaking East were frequented by ships
from Arles. The radius of action of the navicularii was probably not as large as
some historians have supposed, but the Roman state had placed the shippers
from Arles on an equal footing with those of four other ports or regions of the
empire that were essential to the supply of Rome.
Because of their modest character, the other professional associations of
Arles were far from getting the attention of imperial authorities. For example,
the lapidarii Almanticenses were, in the social hierarchy of the colony,
undoubtedly in a very different position from the corpus naviculariorum
(no. 24). Probably originating from a still unknown place called Almanticum
or Almantica, the stone-cutters paid the final honour to Sextus Iulius Valenti-
nus. The amount of money from the dues paid by their colleague permitted
the purchase of a sarcophagus. The lapidarii Almanticenses thus formed a
funerary association based on a professional network. Yet, 300 kilometres
away, at Cimiez, they dedicated an altar to Hercules.29 This mobility allows
us to discuss the way in which lapidarii conducted their work, and the nature
of the group that they formed. Were they itinerant workers, travelling from
building site to building site? This seems possible.30 In that case, they would
have formed a group of limited size that played a marginal role within the
society of Arles. By contrast, the navicularii, the utricularii, and the fabri
tignuarii often made sure that they described themselves as corpora colonia
Iulia Paterna Arelate. This description was not insignificant: it served to
emphasize not only the roots, but also the official existence and the legal

27
Virlouvet (2004: 356–7, particularly n. 82 on a possible provenance from Alexandria).
28
The prefect alludes to an enlargement of the conflict (cum eadem querella latius procedat,
ceteris etiam implorantibus auxilium aequitatis) and takes measures valid for the homines qui
annonae deserviunt, in general: the ceteri mentioned could correspond to four other corpora.
A large part of the text is lost, and the reason for which the plaque was put up in the eastern part
of the empire will forever remain unknown to us.
29
CIL 5.7869.
30
The opifices lapidarii, known at Vaison through one simple epitaph, could have worked in a
similar way (CIL 12.1384).
262 Nicolas Tran

recognition of the corpora in the city.31 Possibly, the lapidarii Almanticenses

and the partiarii (no. 26; Fig. 11.1) did not belong to the same juridical
category. In any case, they give the impression that they place themselves in
the lowest echelon of the hierarchy of associations at Arles. It is that humble
image that the lenuncularii must have wanted to get rid of when they decor-
ated a gathering place with all possible care and made it into a matter of honor
(nos 22–3).32 Through social activities, these boatmen, who maintained the
contacts between Arles and its pre-ports, as well as the crossing of the Rhône,
tried to show and confirm their respectability. Socially, however, they placed
themselves clearly below the seafarers and traders.

ABOVE AND BEYOND THE ASSO CIATIONS: T HE

HIGHEST LEVELS O F HARBOUR SOCIETY

A curiosity of the epigraphic record of Arles has to do with the absence of

traders and merchants of the highest level, usually referred to as negotiatores
and mercatores. The only exception confirms the rule. As negotiator familiae
gladiatoriae, Marcus Iulius Olympus was the impresario of a group of gladi-
ators, not a professional involved in long-distance trade.33 In this respect,
Arles contrasts with Lyon and Ostia. The harbour society of Ostia was
dominated by mercatores and negotiatores, united in various associations—
examples are the mercatores frumentarii and the negotiatores olearii ex Bae-
tica.34 At Lyon, the negotiatores vinarii formed the most prestigious colle-
gium.35 Other specialized negotiatores stood alongside them.36 Parallels from
the abundant documentation of Narbonne are less numerous, but still, one
negotiator and two mercatores are known; two people whose occupation has
partially been erased could perhaps be added.37 How then are we to explain
the silence of the epigraphy at Arles? One could, legitimately, explain it from
the coincidental nature of archaeological finds or from the loss of inscriptions;
texts mentioning the occupations in question may perhaps be found at Arles
in the future. However, about 800 inscriptions have already been discovered.

31
In Lyons, the Omnia corpora licite coeuntium formed an official category that without
doubt did not include all associations (CIL 13.1921 and 1974).
32
Christol and Tran (2014).
33
CIL 12.727.
34
Mercatores frumentarii: CIL 14.161, 4142, 4620. Negotiatores olearii ex Baetica: CIL
14.4458.
35
CIL 13.1911, 1921, 1954, 11179; ILGN 423.
36
AE 1982, 702, 709; CIL 13.1906, 1948, 1966, 1972, 1996, 2018, 2023, 2025, 2029, 2030,
2033, 2035.
37
CIL 12.4496, 4492; ILGN 586. The two uncertain cases correspond to CIL 12.5971, 5973.
 Commerce and Crafts in Ancient Arles 263

The absence of these professionals from the epigraphic record is thus strange,
since it is beyond doubt that mercatores and negotiatores operated at Arles.
A fragment from the Digesta refers to a slave merci oleariae praepositus
Arelate: anyone would associate him with a mercator.38
The hypothesis that can be proposed is that, besides the three known navicu-
larii, major businessmen are already present in the epigraphy from Arles, but
remain unnoticed because they did not make their involvement explicit. Why?
Because they had a more valuable aspect of their identity to put forward. The
group of the seviri Augustales corporati from Arles consists of individuals that,
based on their name, seem to be freedmen; their wealth is sometimes palpable.
For example, in the course of the first century, Caius Fabius Hermes was able to
have a vast mausoleum built for himself, his family, his patron, and the brother
of his patron (no. 27). It is, of course, likely that the wealth of these freedmen
seviri came from their involvement in the harbour economy. This is, for example,
the case with a certain Lucius Pacullius Ephoebicus with a remarkable Greek
cognomen (no. 28). In short, among the people from Arles who, on their
tombstone, referred to themselves as only seviri Augustales, there were, in all
probability, negotiatores, mercatores, navicularii, and other professionals.
Several texts make it possible to corroborate this hypothesis. Gaius Paquius
Pardalas was described by his freedman Epigonus as patron not only of the
corpus of the seviri Augustales (of which he was a member), but also of the
shipbuilders, the utricularii, and the centonarii (no. 19; Fig. 11.2).39 These
three professional associations had turned to the rich freedman to find support
and protection, probably because, after the success of his activities in the port,
they considered him a distinguished person in the city. Similarly, as donor of a
statue of Neptune, Publius Petronius Asclepiades was without doubt socially
superior to the lenunculari that profited from his generosity (no. 22). He
probably did not belong to their corpus, because he did not refer to it on the
inscription recording the donation. However, he must have been close to the
lenuncularii in his professional activities—if he did not do business with them.
While they remained silent about their occupation and the provenance of
their wealth, certain seviri Augustales did emphasize that they had multiple
geographical ties. Indeed, mobility was one of the characteristic features of the
lives of Roman businessman: the itineraries of Marcus Frontonius Euporus
and Quintus Capitonius Probatus have already suggested this. Like Euporus,
Publius Sextius Florus was sevir Augustalis at Aix-en-Provence and Arles (no.
29). Marcus Silenius Symphorus obtained the same title at Lyons, at Arles, and
at Riez (no. 30). The tomb of the sevir Augustalis L. Subrius La[---] and his
wife, from Arles, stood in Cartagena, where the couple had probably installed
themselves for business reasons (no. 31). Thus, the corpus of the seviri

38
Dig. 14.13 pr. (Ulp. Ad Edict. 28).
39
The combination recalls the case of Cnaeus Sentius Felix at Ostia (CIL 14.409).
264 Nicolas Tran

Fig. 11.2. Funerary altar of G. Paquius Pardalas (inscription 19). Musée de l’Arles
antique. Photo: Nicolas Tran.

Augustales of Arles welcomed individuals whose social and, probably, profes-

sional horizon surpassed that of the city.40 At least some of them must have
organized long-distance transport.
This type of trade was sometimes so lucrative that it made it possible for
people to escape from their condition and advance in society. Cicero refers to
this practice: ‘[trade] even seems to deserve the highest respect, if those who
are engaged in it, satiated, or rather, I should say, satisfied with the fortunes
they have made, make their way from the port to a country estate, as they have
often made it from the sea into port.’41
Seen from the top of the social hierarchy, the mercatura was not really
honourable until a mercator had stopped being one, and reinvested his profit

40
In 2007, an epitaph of a sevir Augustalis of Lyons, V[er]ecun[dius - - -] and one other city
was extracted from the Rhône. If one accepts that the letters PA were part of the title of the
colonia Iulia Paterna Arelate, this was Arles (no. 32).
41
Cic. De off. 1.151.
 Commerce and Crafts in Ancient Arles 265

in landed property. This was probably the route of Aebutius Agatho in the
second half of the second century: his epitaph should not be read as a picture
of his social identity, but as a curriculum vitae (no. 33). Agatho died aged 70 in
a city inland, where the function of curator peculi rei publicae was entrusted to
him. The Glanici had recognized in him the expertise of a businessman that he
had built up as shipper on the Saone.42 This occupation of river transporter,
which he exercised at Lyon, perhaps with financial support from a patron from
that city, should have led him to nurture ties with the colonies of Arles and Apt,
where he obtained the title of sevir.43 Finally, he installed himself in Glanum,
maybe to enjoy his retirement from business. It is possible that Quintus
Cornelius Zosimus did the same, installing himself in a village in the eastern
part of the territory of Arles—the pagus Lucretius (no. 34).44 There, in the reign
of Antoninus Pius, he used his wealth and network to defend the interests of the
inhabitants with the imperial authorities: the pagani considered themselves
wronged, because they were denied free access to the public baths, contrary to
the rights that their status of inhabitants of Arles gave them. One cannot,
however, be certain that Zosimus fits within the model of Trimalchio’s retire-
ment from business: his presence in the pagus Lucretius could equally well
be explained through the fact that his patron Quintus Cornelius Marcellus
possessed land here. He could have been its procurator. Despite all this, the
epigraphic silence about the professional activity of Zosimus and the majority
of the seviri augustales of Arles seems revealing. Their civic status permitted
them, if not to live on landed property, at least to detach themselves from their
professional identity, while putting up a powerful civic identity.
As a whole, it is plausible to think that the top of the harbour society was
dominated by rich freedmen, who were at least partially financed by their patrons
and who were judged worthy enough by the civic authorities to be attributed the
function of sevir.45 Many owed their financial wealth to occupations in trade and
maritime transport, even if very few confirmed this on their epitaph. This idea
has an impact on our picture of the harbour society as a whole. That the seviri
augustales formed more prestigious communities than the professional associ-
ations is clear. Indeed, the sportulae that some received beyond the public
distributions privileged the former.46 Craftsmen and traders who highlight

42
See also the quaestura aerarii entrusted by the colony of Ostia to the mercatores frumentarii
Publius Aufidius Fortis (CIL 14.4620–1) and Marcus Iunius Faustus (CIL 14. 4142). Agatho was
responsible for the kalendarium at Glanum (for the account book of the city).
43
Seneca dedicated his De Beneficiis ad Aebutium Liberalem, to a friend from Lyons (Ep. 91).
44
Jacques (1990: 63–4); Gascou (2000); Christol (2008: 133–4).
45
P. Publicius Eutychus was probably a freedman of the colony.
46
Van Nijf (1997: 253–4). At Cimiez, the decurions and the seviri were invited to a banquet,
while the members of the professional collegia had to content themselves with a distribution of
oil (CIL 5.7905). During another distribution, the former received sportulae of two denarii, the
latter received sportulae of only one (CIL 5.7920).
266 Nicolas Tran

only their memberships or their responsibilities in professional corpora were at

an inferior level of the social hierarchy. Indeed, a thorough analysis leaves the
feeling that at Arles the two groups had different characteristics, notably regard-
ing onomastics. Excluding Lucius Secundius Eleuther and Marcus Frontonius
Euporus, one has to observe that two out of fourteen members of the professional
associations had Greek cognomina (Table 11.1).47 Among the corporati with a
Latin cognomen, there is the son of a certain Lucius Iulius Trophimus. However,
the list of seviri Augustales gives exactly the opposite impression. Seven people of
the eleven whose cognomen has been preserved had a Greek one.48 The historian
has to make do with fragile hypotheses, since the explicit mention of legal status
was no longer the norm in the epigraphy of the second and third centuries, and
since onomastic properties give only a very crude indication, and the sample is
small. Yet ordinary freeborn people, exercising occupations that were less lucra-
tive than trade—public construction or carpentry, for example—probably occu-
pied a lower rank than did the rich seviri Augustales, of whom the majority were
freedmen. As it was often related to the self-serving financial support of a patron,

Table 11.1. Names of the corporati of Arles known from inscriptions

Members of professional associations Seviri Augustales of Arles

L. Secundius Eleuther (no. 4)

M. Frontonius Euporus (no. 7)
L. Avent. Avitianus (no. 9) Aebutius Agatho (no. 32)
Aur. Septimius Demetrianus (no. 24) Sex. Alfius Vitalis Florens (no. 35)
Caecilius Niger (no. 20) Q. Cornelius Marcelli lib. Zosimus (no. 33)
Q. Capitonius Probatus (no. 8) C. Fabius C. lib. Hermes (no. 26)
Q. Candidius Benignus (no. 10) C. Iulius Fortunatus (no. 36)
T. Flavius Titus (no. 11) L. Pacullius Ephoebicus (no. 27)
Iulius Eumenes (no. 15) G. Paquius Optati lib. Pardalas (no. 18)
C. Iulius Pomp[—] (no. 19) P. Publicius Eutychus (no. 34)
L. Iulius Secundus (no. 16) P. Sextius Florus (no. 28)
Sex. Iulius Valentinus (no. 23) M. Silenius Symphorus (no. 29)
M. Iunius Messianus (no. 17) L. Subrius La[—] (no. 30)
L. Iulius Augustalis, son of L. Iulius
Trophimus (no. 12)
Pompeius Lucidus (no. 13)
C. Publicius Bellicus (no. 14)

Note: Greek cognomina are highlighted in bold.

47
I have excluded from the count the partiarius Hermias and Marcus Atinius Saturninus, the
apparitor of the statio of the navicularii. One cannot know if he was employed by the navicularii
or was a member of the corpus nauiculariorum marinorum. The names of their wives did not
feature any Greek cognomen, which reinforces the impression that we are here dealing with a
milieu of freeborn plebeians.
48
Though it did not have a strong social connotation, the cognomen Fortunatus was carried
by slaves and freedmen (Kajanto 1965: 273); in Gallia Narbonensis: CIL 12.752 (Arles); 1370
(Vaison); 1525 (near Sisteron); 3277 (Nîmes); ILGN 548 (western territory of Nîmes).
 Commerce and Crafts in Ancient Arles 267

Fig. 11.3. Relief with harbour workers. Musée de l’Arles antique. Photo: Nicolas Tran.

the position of a freedman was thus not always synonymous with marginality
and social inferiority—quite the opposite. Though they certainly formed a small
minority of the original servile population, freedmen could live in wealth and
with prestige, not despite, but because of, their dependence. Hence, this case
study of Arles confirms one of the strongest elements in Mouritsen’s analysis of
the Roman freedman. On the one hand, most of the liberti would have con-
sidered a break of social and economic ties with their patron as a disadvantage,
and not as an asset.49 On the other hand, when they wanted to invest a part of
their wealth in port economies, notables sought the help of reliable persons. Most
of the time, they turned to their deserving former slaves, because they had
confidence in their loyalty and skills.50
To conclude, neither the members and dignitaries of the professional
associations, nor the seviri Augustales who based their wealth on activities
related to the harbour, lived on the margin of the society in Arles in the second
and early third centuries. On the contrary, the majority of the people we know
seem to have belonged to the highest echelons of the world of work—they

49 50
Mouritsen (2011: 228–34). Mouritsen (2011: 219).
268 Nicolas Tran

were part of a ‘plebeian’ elite that was well integrated in the city. However, the
identity of these people and the economic activity that supported them
cannot be understood in the same vein. In a certain way, the world of
work was less marginal than the work itself, in the sense that the civic
integration of craftsmen and traders was effected through their participation
in more or less prestigious institutions of the city or recognized by the city.
Indeed, the primary goal of these institutions was not to organize economic
life in the strict sense of the word. It remains to note that epigraphy
sheds only very partial light on this milieu. An epitaph like that of Hermias
lifts, covertly, the veil from the slaves, who are so rare in the epigraphy
of harbours in the Roman west, though their labour force must have been
crucial. Free or slaves, the unskilled workers employed as labourers escape
the view of the epigrapher, and appear in rare depictions on reliefs or
paintings (Fig. 11.3).51 The epigrapher perceives only the top of the iceberg,
and this, in itself, is already enough to illustrate the fundamental heterogen-
eity of the world of work.

Appendix: Catalogue of Inscriptions

1. Arles: Tribute of the navicularii marini to an equestrian procurator

Block of hard limestone, discovered in an ‘old wall of the Commanderie’, in the quarter
of Trinquetaille. CIL 12.672.
[.] Cominio [-- - f(ilio)] | Claud(ia) Bo[-- -]io | Agricola[e Aur]elio | Apro, praef(ecto)
cohor[t(is)] | tert(iae) Bracaraugustanor(um), | tribun(o) leg(ionis) I Adiut(ricis),
procur(atori) | Augustorum ad annonam | prouinciae Narbonensis | et Liguriae,
praef(ecto) a[lae] miliariae | in Mauretania Caesariensi, | nauic(ularii) marin(i)
Arel(atenses) | corp(orum) quinq(ue) patro[no] | optimo et innocentis | simo.

2. Arles: Tribute of the navicularii marini to a local notable

Statue base of marble, with moulded frame. Discovered in 1850 in an excavation
conducted in a yard in the Jesuits’ college, on the location of the forum adiectum. CIL
12.692.
Cn(aeo) Cornel(io) | Cn(aei) fil(io) Ter(etina) | Optato, | IIuir(o), pontific(i), | flamini,
| nauiculari(i) marin(i) | Arel(atenses) patrono.

51
The only attestation of dockers at Arles is a relief representing two people who are busy
packing a bundle depicted on Fig. 11.3. The sculpture comes from a mausoleum, and was thus
commissioned by an individual of a higher status than that of the dockers. Cf. Espérandieu
(1907–66: i. 164); Rothé and Heimans (2008: 619 (Fig. 890)).
 Commerce and Crafts in Ancient Arles 269

3. Beirut: Letter of the prefect of the annona to the

navicularii marini of Arles
Bronze plaque stemming from the cutting up (in modern times) of an inscribed tablet
of the Roman era. CIL 3.14165, 8.
[Cl(audius) I]ulianus nauiculariis | [mar]inis Arela[t]ensium quinque | [c]orporum
salutem. | [Qui]d lecto decreto uestro scripserim | [[- - -]] proc(uratori) Augg(ustorum),
e(gregio) u(iro), subi|ci iussit. Opto felicissimi bene ualeatis. | E(xemplum) e(pistulae). |
Exemplum decreti nauiculariorum ma|rinorum Arelatensium quinque cor|porum, item
eorum quae aput me acta|sunt, subieci. Et cum eadem querella la|tius procedat, ceteris
etiam imploranti | bus auxilium aequitatis, cum quadam de|nuntiatione cessaturi
propediem obsequi | si permaneat iniuria peto, ut tam indemni|tati rationis quam
securitati hominum | qui annonae deseruiunt consulatur, | inprimi charactere regulas
ferreas et | adplicari prosecutores ex officio tuo iu|beas qui in Vrbe pondus quo
susce|perint tradant.

4. Arles: Epitaph of Lucius Secundius Eleuther, navicularius from Arles

Marble altar, decorated on the sides with a patera and an ewer. Seen in the eighteenth
century at Saint-Pierre-de-Galigan. Lost. CIL 12.704.
D(is) M(anibus), | L(ucio) Secundio | Eleuthero, | nauicular(io) Arel(atensi), | item
IIIIIIuir(o) Aug(ustali) | corpor(ato) c(olonia) I(ulia) P(aterna) A(relate), | Secundia
Tatiana{e} fil(ia) | patri pientissim(o).

5. Arles: Epitaph of Marcus Atinius Saturninus, apparitor of the

statio of the navicularii
Sarcophagus seen in the second half of the seventeenth century at the Pointe de
Trinquetaille. Lost since at least 1766. CIL 12.718.
- - - | et quieti aeternae | M(arci) Atini Saturnin(i), [ap]|paritor(is) nauicular(iorum) |
station[is - -].

6. Arles: Epitaph of Quintus Navicularius Victorinus

Sarcophagus seen in 1574 ‘at the cemetery of the Aliscamps’. Lost. CIL 12.853.
Q(uintus) Nauicula|rius Victori|nus, Val(eriae) Seue|rinae, coniugi sanctissimae.

7. Saint-Gabriel: Epitaph of Marcus Frontonius Euporus,

sevir Augustalis and navicularius marinus
Marble altar, discovered in the sixteenth century. Preserved in the chapel. CIL 12.982.
[D(is)] M(anibus), || M(arci) Frontoni Eupori, | IIIIIIuir(i) Aug(ustalis) col(onia) Iulia
| Aug(usta) Aquis Sextis, nauicular(ii) | mar(itimi) Arel(atensis), curat(oris) eiusd(em)
corp(oris), | patrono nautar(um) Druen|ticorum et utric(u)larior(um) | corp(ora-
torum) Ernaginens(i)um, Iulia Nice, uxor, | coniugi carissimo.
270 Nicolas Tran

8. Lyon: Epitaph of Quintus Capitonius Probatus, navicularius marinus

Altar discovered in 1718 in the river bed of the Rhône. CIL 13.1942.
D(is) M(anibus) | Q(uinti) Capitoni Probati | senioris, domo Rom(a), | IIIIIIuir(i)
Aug(ustalis) Lugudun(i) | et Puteolis, | nauiculario marino, | Nereus et Palaemon, |
liberti, patrono | quod sibi uiuus insti|tuit posterisq(ue) suis | et sub ascia
dedicau(erunt).

9. Arles: Epitaph of Lucius Aventius Avitianus, magister

of the fabri tignuarii
Fragment of a marble sarcophagus, seen in the second half of the seventeenth century,
at the Pointe de Trinquetaille. Lost since at least 1766. CIL 12.719.
D(is) M(anibus) || L(ucii) Auent(i ou -ini) Auit|iani, fabr(o) ti | gnuar(io) c(olonia) I(ulia)
P(aterna) Arel(ate), | mag(istro) eiusdem | corp(oris), primo art(ificum), T(itus) Tossiu|s
Marcus CO|SNA merenti.

10. Arles: Epitaph of Quintus Candidius Benignus, magister

of the fabri tignuarii
Sarcophagus representing on its front side, in the middle of a frame, a metrical
inscription that nowadays is illegible. Around the frame are depicted, to the left, an
ascia and, to the right, a set square with plumb bob. Seen in the seventeenth century by
F. de Rebatu, who saw it on the banks of the Rhône ‘at La Ponche’. Found around 1725 at
the Pointe de Trinquetaille; subsequently brought to mas d’Euminy, in Camargue, and
brought back to the Alyscamps in 1844. CIL 12.722.
D(is) M(anibus), || Q(uinti) Candi[di] Benigni, fab(ri) tig(nuarii) c|orp(orati)
Ar(elate), ars cui summa fuit | fabricae, studium, doctrin(a), | pudorque, quem
magni | artifices semper dixsere | magistrum, doctior hoc ne|mo fuit, potuit quem
uinc|ere nemo, organa qui nosse|t facere, aquarum aut duce|re cursum, hic co[n]-
uiua fui|t dulcis, nosset qui pasce|re amicos, ingenio studio | docilis animoque
benig|nus, Candidia Quintina, | patri dulcissimo et Val(eria) | Maxsimina, coniugi
kar(issimo).

11. Arles: Epitaph of Titus Flavius Titus, faber tignuarius

Sarcophagus in limestone, seen in 1765 by Gaillard, when he was at the Pointe de
Trinquetaille. CIL 12.726.
D(is) M(anibus), || Tit(o) Fl(auio) Tito, corp(oris) | fabror(um) tig|narior(um)
corp(orato) | Arel(ate), Tit(us) Fl(auius) In|uentus, pa|trono pient(issimo).

12. Arles: Epitaph of Lucius Iulius Augustalis, faber tignuarius

Marble altar, found underneath the altar of the Virgin of the monastery of Saint-Césaire.
It is probable that a dedication to the Manes was inscribed on the pulvini of the
entablature, which are currently rather damaged. CIL 12.728.
[D(is) M(anibus)] | L(uci) Iuli Augus|talis, fabri | tign(ariorum) corpor(ati) | Arel(ate),
L(ucius) Iulius Trophimus, | pater infe|licissimus.
 Commerce and Crafts in Ancient Arles 271

13. Arles: Epitaph of Pompeius Lucidus, faber tignuarius

Unclear when and by whom it was seen. Provenance unknown. Lost. CIL 12.735.
D(is) M(anibus) | Pompei Lucidi, | fabri tignuari | corporati Arelat(e), | e funeraticio eius.

14. Arles: Epitaph of Caius Publicius Bellicus, magister

of the fabri tignuarii
Sarcophagus seen at Pointe de Trinquetaille from 1765 onwards. Decorated on its
small sides with an ascia and a plumb-line.
D(is) M(anibus), || C(aio) Publ(icio) Bellico, corp(orato) col(oniae) | Iul(iae) Paterne
Arel(ate) fabror(um) | tignuarior(um), item magistro, | Venucia Priscilla co(n)iugi |
incomparabili item | Venucia Priscilla | uiua sibi fecit.

15. Arles: Epitaph of Julius Eumenus, utricularius

Altar of hard limestone, with moulded base and top; on top, two pulvini carrying the
letters D and M. Between these, a roof shaped like a pediment. On the left, an ascia, to
the right a plumb-line. Found in 1875 at Trinquetaille. CIL 12.729.
D(is) M(anibus), | Iulius Eumenes | uixit ann(os) XXII, | Iulia Agrippina, | patron(a)
alumno | et corporato | utriclariorum, | quot tu nobis | debuisti facere, | et mater
infelicissimae | posuerunt.

16. Arles: Epitaph of Lucius Iulius Secundus, utricularius

and shipper of the Durance
Altar found in 1587, under the old church of Saint-Laurent. Broken in eighteenth
century, lost since. Underneath the text was an ascia. CIL 12.731.
L(ucio) Iul(io) Secundo, | utric(u)lario corp(orato) | c(olonia) I(ulia) P(aterna)
A(relate), qui legauit | eis testamento suo | (denarios) CC, ut ex usur(is) eor(um) |
omnibus annis sacri|ficio ei parentetur, | item naut(ae) Druentic(o) | corpor(ato),
Mogituma | Epipodius, filius nat(uralis), | patri pientissimo.

17. Arles: Epitaph of Marcus Iunius Messianus,

magister of the utricularii
Sarcophagus in limestone. The epitaph is on the front side, in a frame held up by two
cupids. The two small sides are decorated with a garland. Seen in the sixteenth century
‘next to the ruined church of Saint Césaire and near the chapel of Brau’. CIL 12.733.
D(is) M(anibus), || M(arco) Iunio Messiano, | utricl(ario) corp(orato) Arelat(e), |
eiusd(em) corp(oris) mag(istro) IIII f(acto), | qui uixit ann(os) XXVIII, | m(enses) V,
d(ies) X, Iunia Valeria | alumno carissimo.
272 Nicolas Tran

18. Saint-Gilles: Epitaph (?) of a curator of the utricularii, curator of the

nautae of the Ardèche and of the Ouvèze
Fragment seen in the eighteenth century, reused above the door giving access to the
Chapelle Saint-Pierre. CIL 12.4107.
- - - | naut(a) Atr(icae) et Ou(idis), curator | eiusdem corporis, item | utric(u)lar(ius)
corp(oratus) Arelat(e), | eiusdemq(ue) corp(oris) curat(or), | - - -

19. Arles: Epitaph of Gaius Paquius Pardalas, sevir Augustalis

Marble altar with moulded base and top. Preserved in 1655 underneath the ‘high altar’
of the church ‘of Saint Jean l’Évangeliste of the monastery of Saint-Césaire’. CIL
12.700.
D(is) M(anibus) | G(ai) Paqui Optati | lib(erti) Pardalae, IIIIII(uiri) | Aug(ustalis)
col(onia) Iul(ia) Pat(erna) Ar(elate), | patron(i) eiusdem | corpor(is), item patron(i) |
fabror(um) naual(ium), utric(u)lar(iorum) | et centonar(iorum), C(aius) Paquius |
Epigonus cum liberis suis | patrono optime merito.

20. Arles: Epitaph of Caius Iulius Pom[---], curator

of the fabri navales
Sarcophagus seen after 1739, supposedly preserved in the archbishop’s palace, but lost
since. CIL 12.730.
C(aius) Iul(ius) Pom[- - -] | collega fab[rum] | naualium c[orp(oris) Arel(ate)], |
curator eius[dem] | corporis, et | Seuera uiui s(ibi) | posuerunt et s(ub) [ascia] |
dedicauer(unt).

21. Arles: Epitaph of Caecilius Niger, faber navalis

Fragment of a marble altar. Discovered in 1886 in the old ‘Rue Wauxhall’ (now
‘Rue Jean-Jaurès’). CIL 12.5811; ILGN 108.
[Cae]c[ilio] | Nigro, fa[bro nauali. | Praete]riens quicumque leges h[aec carmina
nostra, | qu]ae tibi defuncti nomina uer[a dabunt, | incomptos] elegos ueniam peto
ne ue[rearis] | perlegere, et dicas carmen ha[bere bene (?). | C]aecilius Niger est hic ille
[sepul]tu[s eundem] | quo cernis titulum sta[re, habet ecce locum]. | Nunc tibi nauales
pauci damus ul[tima uota]: | hoc et defuncto corpore munus [habe]. | Ossa tuis urnis
optamus dulce quiesc[ant] | sitque leuis membris terra mo[lesta tuis]. | Arti[f]ic[i]
artifices Nigro damus ista s[odali] | carmina, quae claudit iam res[oluta salus].

22. Arles: Dedication to the Numina of the emperors and

to the honour of the association of lenuncularii
Inscription on the socle of a statue of Neptune found in the Rhône in 2007. Christol
and Fruyt (2009: 104–9); AE 2009, 822.
 Commerce and Crafts in Ancient Arles 273
Numinibus Auggg(ustorum) nnn(ostrorum), | honori corporis renunclariorum,
P(ublius) Pe|tronius Asclepiades donum dedit.

23. Arles: Dedication to the Genius of the association of lenuncularii

Fragmentarily preserved altar, found in the Rhône in 2007. Christol and Fruyt (2009:
104–9); AE 2009, 823.
[Neptuno?, | Genio cor|po]ris, len|[u]nclari | sacrum.

24. Arles: Epitaph of Sextus Iulius Valentinus, dedicated

by the lapidarii Almanticenses
Sarcophagus seen after 1588 near Saint-Honorat, now lost. CIL 12.732.
D(is) M(anibus) || Sex(ti) Iul(i) Valen|tini, lapida|ri(i) Almanti|censes ex fu|nere eius
et || Pomp(eiae) Gra | tiniae co(n)iugi | inconpara | bili posuer(unt).

25. Arles: Epitaph of Aurelius Septimius Demetrianus,

shipper of the Durance
Sarcophagus seen in 1739, discovered ‘close to Trinquetaille’. CIL 12.721.
Aur(elius) Septimius Demetrianus, | nauta Druenticus, uiuus | sibi posuit.

26. Arles: Epitaph of Hermias

Marble altar with moulded socle and top. Discovered in 2001 in the Rhône, near
Trinquetaille. Underneath the inscription a ship has been depicted, a plumb-line, and
an ascia. AE 2003, 1079.
D(is) M(anibus) | Ermie | partiari | college | posuerunt.

27. Arles: Epitaph of Caius Fabius Hermes, sevir Augustalis

Limestone plaque with moulded frame, probably intended to decorate a mausoleum.
Discovered in 1763 ‘in the quarter of Molleirez’. CIL 12.694.
C(aius) Fabius C(ai) lib(ertus) Hermes, | IIIIIIuir Aug(ustalis) c(olonia) I(ulia)
P(aterna) Arel(ate), | uiuos fecit sibi et suis et | C(aio) Fabio L(uci) f(ilio) Secundo,
patron(o), | et L(ucio) Fabio L(uci) f(ilio) Primo, fratri | eius. | H(oc) m(onumentum)
h(eredem) m(eum) n(on) s(equetur).

28. Arles: Epitaph of Lucius Pacullius Ephoebicus, sevir Augustalis

Altar seen in 1783 ‘in the middle of a small garden at Mas des Passerons’, in Camargue.
CIL 12.699.
D(is) M(anibus), | IIIIIIuir(o) Aug(ustali) | Arelate | L(ucio) Pacullio Ephoe|bico,
Aurelia | Eutychia uxor.
274 Nicolas Tran

29. Arles: Epitaph of Publius Sextius Florus, sevir Augustalis

Marble tablet, discovered in 1543 ‘near the church of Saint-Antoine’. Comes from the
necropolis of Trébon. Lost. CIL 12.705.
P(ublius) Sextius Florus, IIIIIIuir Aug(ustalis) | col(onia) Iul(ia) Aquis et col(onia) Iul(ia)
P(aterna) Arel(ate), | Valeriae Spuri f(iliae) Lassinae, | uxori pientissimae, | Sex(to)
Valerio Proculino et suis.

30. Lyons: Epitaph of Marcus Silenius Symphorus, sevir Augustalis

Altar discovered, in the Chemin de la Favorite; decorated with an ascia and a plumb-
line. ILTG 241; AE 1935, 17.
D(is) M(anibus) | M(arci) Sileni Symphori, | IIIIIIuir(i) Aug(ustalis) | Lug(duni),
Arelate, Reis, | Silenia Latina, | liberta idemque uxor, | patrono et marito | erga se
optimo | et sibi uiua posuit.

31. Cartagena: Epitaph of Lucius Subrius La[---], sevir Augustalis

Marble plaque discovered in 1975, in a necropolis in the north-west part of the city. AE
1975, 523.
L(ucius) Subrius La[- - -], | IIIIIIuir Au[g(ustalis) - - -] | c(olonia) I(ulia) P(aterna)
Arela[te], | et Subria L(uci) l(iberta) [- - -] | DA.

32. Arles: Epitaph of Verecundius [---], sevir Augustalis

Altar found in the Rhône in 2007. Heijmans (2009: 340, n. 1); AE 2009, 816.
- - - ? | V[er]cun[dio] | C[.]V[.]TO[..] IIII[IIuiro Aug(ustali)] | co[rp(orato)] col(oniae)
C[laud(ia) | L]u[gud(uni)] item [- - - | - - -] PA [- - - | - - -] AV[- - - | - - - | - - - | - - -] XX F
[- - - | - - -] dulci[ssimo].

33. Saint-Rémy-de-Provence. Epitaph of Aebutius Agatho,

sevir Augustalis
Altar, discovered in 1740 in the ruins of the chapel of Sainte-Trophime. CIL 12.1005.
[D(is) M(anibus) et | me]mori(a)e aeterna[e] | Aebuti Agathon[is], | [IIIIII]uiro
Aug(ustali) corp(orato) [col(oniae) Iul(iae) | Pat]er(nae) Arel(ate), curat(ori)
eius|[de]m corp(oris) bis, item IIII[II|ui]ro col(oniae) Iul(iae) Aptae, nau|[t]ae
Ararico, curator[i] | peculi r(ei) p(ublicae) Glanico(rum), qui | uixit annos LXX, |
Aebutia Eutychia patro|no erga se pientissimo.

34. Géménos: Tribute of the pagani of the Pagus Lucretius

to Quintus Cornelius Zosimus, sevir Augustalis
Moulded marble table, seen for the first time in the seventeenth century. It served as
altar table in the chapelle Notre-Dame. CIL 12.594; AE 2000, 883; AE 2006, 783.
 Commerce and Crafts in Ancient Arles 275
[P]agani pagi Lucreti qui sunt fini|bus Arelatensium, loco Gargario, Q(uinto)
Cor(nelio) | Marcelli lib(erto) Zosimo, IIIIIIuir(o) Aug(ustali) col(onia) Iul(ia) |
Paterna Arelate, ob honorem eius. Qui notum fecit | iniuriam nostram omnium
saec[ulor]um sacra|tissimo principi T(ito) Aelio Antonino [Aug(usto) Pio te]r
Romae | misit per multos annos ad praesides pr[ouinci]ae perse|cutus est iniuriam
nostram suis in[pensis e]t ob hoc | donauit nobis inpendia quae fecit ut omnium
saecu|lorum sacratissimi principis Imp(eratoris) Caes(aris) Antonini Aug(usti) Pii |
beneficia durarent permanerentque, quibus frueremur | [aquis] et balineo gratuito
quod ablatum erat paganis, | quod usi fuerant amplius annis XXXX.

35. Arles: Epitaph of Publius Publicius Eutychus,

sevir Augustalis
Moulded marble tablet, perhaps from the Alyscamps (?). CIL 12.702.
D(is) M(anibus), | P(ublio) Publicio | Eutycho, | IIIIIIuir(o) Aug(ustali) | c(olonia) I(ulia)
P(aterna) Arel(ate).

36. Arles: Epitaph of Sextus Alfius Vitalis Florens

Sarcophagus discovered in 1809 in the ancient circle of Vauxhall. CIL 12.689.
D(is) M(anibus), || Sex(tus) Alfius Vitalis Forens, | IIIIIIuir Aug(ustalis) corp(oratus)
c(olonia) I(ulia) P(aterna) Arel(ate), et | Alfiae Epauxesi, lib(ertae) uxoriq(ue) | suae,
quae uixit ann(os) XL, m(enses) II, | d(ies) XXVII, sibi posterisquae | suis uiuus fecit.

37. Arles: Epitaph of Veria Filtata and of Caius Iulius Fortunatus,

sevir Augustalis
Marble altar decorated with foliage; seen in the eighteenth century at Saint-Honorat-des-
Alyscamps. CIL 12.709.
[D(is)] M(anibus), | Veriae Filtate, | amica dolens | posuit in honorem | C(ai) Iuli
For|tunati, IIIIIIuir(i) | Augustalis, | uxori.

REFERENCES

Alföldy, G. (1986). ‘Zur Nomenklatur und Laufbahn eines römischen Ritters aus
Concordia’, ZPE 63: 173–80.
Andreau, J. (1985). ‘Modernité économique et statut des manieurs d’argent’, MEFRA
97/1: 373–410.
Burnand, Y. (1971). ‘Un aspect de la géographie des transports dans la Narbonnaise
rhodanienne: Les Nautes de l’Ardèche et de l’Ouvèze’, RAN 4: 149–58.
Clemente, G. (1972). ‘Il patronato nei collegia dell’impero romano’, SCO 21: 142–229.
Christol, M. (1971). ‘Remarques sur les naviculaires d’Arles’, Latomus 30: 643–63.
Christol, M. (2008). ‘Colonie de vétérans et communautés indigènes’, in J.-M. Rouquette
(ed.), Arles: Histoire, territoires et cultures. Arles, 125–34.
276 Nicolas Tran
Christol, M. (2010a). Une histoire provinciale: La Gaule narbonnaise de la fin du IIe
siècle av. J.-C. au IIIe siècle ap. J.-C. Paris.
Christol, M. (2010b). ‘Formes de la vie économique et formes de la vie sociale à Arles
au IIe et au IIIe siècle: Sources et travaux récents’, in M. Silvestrini (ed.), Le Tribù
romane: Atti della XVIe Rencontre sur l’épigraphie. Bari, 405–16.
Christol, M., and Demougin, S. (1984). ‘Notes de prosopographie équestre’, ZPE 57: 163–78.
Christol, M. and Fruyt, M. (2009). ‘Neptune: Étude épigraphique’, in L. Long and
P. Picard (eds), César: Le Rhône pour mémoire. Arles, 104–9.
Christol, M., and Tran, N. (2014). ‘Tituli et signa collegiorum en Gaule méridionale et
ailleurs: Réflexions sur le décor des sièges de collèges à partir du cas arlésien’, École
antique de Nîmes 31: 15–31.
Clemente, G. (1972). ‘Il patronato nei collegia dell’impero romano’, SCO 21: 142–229.
Corbier, M. (2006). Donner à voir, donner à lire: Mémoire et communication dans la
Rome ancienne. Paris.
D’Arms, J. H. (1981). Commerce and Social Standing in Ancient Rome. Cambridge and
London.
Duncan-Jones, R. P. (1974). The Economy of the Roman Empire: Quantitative Studies.
Cambridge.
Duthoy, R. (1974). ‘La Fonction sociale de l’augustalité’, Epigraphica 31: 134–54.
Duthoy, R. (1981). ‘Les *Augustales’, in ANRW II 16/2: 1254–309.
Espérandieu, E. (1907–66). Recueil général des bas-reliefs, statues et bustes de la Gaule
romaine. Paris.
Finley, M. I. (1985). The Ancient Economy. 2nd edn. Berkeley and Los Angeles.
Garnsey, P. (1981). ‘Independent Freedmen and the Economy of Roman Italy under
the Principate’, Klio 63: 359–71; repr. in Garnsey (1998), 28–44.
Garnsey, P. (1998). Cities, Peasants and Food in Classical Antiquity: Essays in Social
and Economic History. Cambridge.
Gascou, J. (2000). ‘L’Inscription de Saint-Jean-de-Garguier en l’honneur du sévir
augustal Q. Cornelius Zosimus’, MEFRA 112/1: 279–95.
Heijmans, M. (2003). ‘Récentes découvertes épigraphiques dans le Rhône à Arles
(Bouches-du-Rhône)’, RAN 36: 377–81.
Heijmans, M. (2009). ‘Les Inscriptions lapidaires et diverses’, in L. Long and P. Picard
(eds), César: Le Rhône pour mémoire. Arles, 340–3.
Jacques, F. (1990). Les Cités de l’Occident romain. Paris.
Kajanto, I. (1965). The Latin Cognomina. Helsinki.
Laubry, N. (2012). ‘Ob sepulturam: Associations et funérailles en Narbonnaise et dans
les Trois Gaules sous le Haut-Empire’, in M. Dondin-Payre and N. Tran (eds),
Collegia: Le Phénomène associatif dans l’Occident romain. Bordeaux, 103–33.
Mouritsen, H. (2011). The Freedman in the Roman World. Cambridge.
Nijf, O. M. van (1997). The Civic World of Professional Associations in the Roman East.
Amsterdam.
Provost, M. (1999). Carte archéologique de la Gaule. Le Gard. 30/3. Paris.
Rostovtzeff, M. I. (1957). Social and Economic History of the Roman Empire. 2nd edn.
Oxford.
Rothé, M. P., and Heijmans, M. (2008). Carte archéologique de la Gaule. Arles, Crau,
Camargue. 13/5. Paris.
 Commerce and Crafts in Ancient Arles 277
Rougé, J. (1965). ‘Les Relations de Lyon et de la mer: À propos de CIL XIII, 1942’, in
Actes du 89e Congrès des Sociétés savantes (Lyon 1964). Paris, 137–52.
Salvo, L. de (1992). Economia privata e pubblici servizi nell’impero romano. I corpora
naviculariorum. Messina.
Tran, N. (2006). Les Membres des associations romaines: Le Rang social des collegiati en
Italie et en Gaule sous le Haut-Empire. Rome.
Verdin, F. (2005). ‘Encore les utriculaires . . . ’, in A. Bouet and F. Verdin (eds), Territoires
et paysages de l’âge du Fer au Moyen Age: Mélanges offerts à Philippe Leveau. Bordeaux,
275–82.
Virlouvet, C. (2004). ‘Les Naviculaires d’Arles: À propos de l’inscription de Beyrouth’,
MEFRA 116/1: 327–70.
Waltzing, J.-P. (1895–1900). Étude historique sur les corporations professionnelles chez
les Romains. Louvain.
 12

Hierapolis and its Professional

Associations: A Comparative Analysis
Ilias Arnaoutoglou

Since the middle of the 1990s, there has been a clear shift in the way
associations are treated by scholars. In sharp contrast to the holistic approach
of late-nineteenth-century scholarship, there is a growing body of literature
examining associations in their geographical and historical context.1 I have
attempted a similar approach, underlining the duality in the associative
experience in Roman Lydia through the examination of the relevant epigraph-
ic testimonia from Thyateira and Saittai.2 The aim of this chapter is to enlarge
the frame and add the nearby Phrygian city of Hierapolis to the picture.3
Hierapolis provides an interesting comparison, not so much because of its
geographic proximity, but rather because, as is the case in Thyateira and
Saittai, its epigraphic record provides an above-average concentration of
evidence for professional associations.4 After some introductory remarks on

1
Earlier scholarship: Ziebarth (1896); Poland (1909); Waltzing (1895–1900). More recently:
Ritti (1995); Labarre and Le Dinahet (1996); Dittmann-Schöne (2000); Zimmermann (2002);
Sommer (2006). On the historiography of Roman associations, see Perry (2006) and Dissen (2009).
2
Arnaoutoglou (2011).
3
Ancient Hierapolis, near modern Denizli in Turkey, has become a tourist destination thanks
to the impressive travertine cascades, i.e. deposits formed as a result of the flow of water from
thermal springs (35o C). Reduction of pressure results in loss of carbon dioxide and deposits of
calcium carbonate in a thin layer at a rate of 3 cm/annum. For the excavations and the
restoration on the site, see the report of D’Andria (2006).
4
The unique nature of the Hierapolitan record becomes apparent when it is compared with
associations attested in the nearby towns of Colossae and Laodikeia on the Lykos. For Colossae, see
MAMA 6.47 (= BE (1939: 392); second/third century AD): honorary inscription for Glykon referring
to οἱ ἑταίροι; MAMA 6.48 (= Ritti 2007, 73; BE (1939: 392); second/third century AD): dedication to
Tatianos Bartos referring to συνγενικὸν νεώτερον. For Laodikeia on the Lykos, see IK 49.32, 33 (= BE
(1997: 585); imperial period): seats in theatre mentioning an [ἐ]ργασίας κλ—. Cf. Dittmann-Schöne
(2001: 232); IK 49.50 (= IGR 4.863, BE (1938: 447); third century AD): fragment of an honorary decree
mentioning [ἡ ἐργασία] τῶν γναφέ[ων καὶ] ἁπλουργῶν, cf. Waltzing (1895–1900: 3, no. 129), Labarre
and Le Dinahet (1996: no. 55), Dittmann-Schöne (2001: 232), Zimmerman (2002: 191), Ramsay
 Hierapolis and its Professional Associations 279

the history of Hierapolis, the comparison with the fenomeno associativo in

Lydia will develop around five questions.5 First, who were forming associ-
ations? Second, how did craftsmen name their groups? Then, what was the
social positioning of professional associations and what was their social
composition? This leads to the question of what was the image of an associ-
ation of craftsmen in society. Finally, the analysis will focus on the issue of
memory management, and the role of professional associations in it.

INTRODUCING HIERAPOLIS

Hierapolis lies in the south-west corner of Phrygia, close to Colossae and Laodi-
keia on the Lykos, on a plateau looking out over the valley of the river Lykos,
which is a tributary of the Maeander. Because of its location, Hierapolis controlled
the route into Lydia from the south-east, as well as the routes connecting inner
Anatolia to the coastal towns of Ephesos and Miletos and to north-west Caria.
Herodotus refers to a Phrygian border town called Kydrara, which was
situated in the area of Hierapolis.6 Yet Hierapolis itself was founded only
sometime in the third century BC, most probably by the Seleucids.7 The area
was the centre of a local cult of Cybele, but, in the Greek polis, the cult of
Apollo Archegetes was dominant.8 After the peace of Apameia in 188 BC,
Hierapolis became part of the Attalid kingdom and followed the latter’s
destiny when it passed under Roman control in 133 BC. Roman merchants
were probably well established in the city: two inscriptions attest the existence
of an association (conventus) of Romans.9 From the second century BC to the

(1890: i. 74, 8), Poland Z70; MAMA 6.11 (= IGR 6, 858; BE (1939: 393); second/third century AD):
honorary inscription mentioning a [ἡ θ]ρεμματι[κή ἐργασία?]; MAMA 6.24 (= BE (1939: 393);
second/third century AD). Tombstone of Meltine referring to τὸ συνγενικὸν; Olba 19 (2011),
186 (= Simsek 2013: 237; imperial period): seats in the northern theatre mentioning a ἐργασία
ὀρθοκουρων?; Thonemann (2011: 188; imperial period): inscription found in forum of Laodikeia
mentioning a τόπος βαφέων.
5
See already Labarre and Le Dinahet (1996: 67), who note the differences between the
testimonies on professional associations from Lydia and Hierapolis. Sanidas (2011) compares
textile production in mainland Greece and Asia Minor.
6
Hdt. 7.30 describes how the army of Xerxes leaves Colossae and proceeds through the
border region between Phrygia and Lydia in the direction of Sardis: ‘ἐκ δὲ Κολοσσέων ὁ στρατὸς
ὁρμώμενος ἐπὶ τοὺς οὔρους τῶν Φρυγῶν καὶ Λυδῶν ἀπίκετο ἐς Κύδραρα πόλιν, ἔνθα στήλη
καταπεπηγυῖα, σταθεῖσα δὲ ὑπὸ Κροίσου, καταμηνύει διὰ γραμμάτων τοὺς οὔρους.’
7
Kolb (1974) and Debord (1997).
8
On cults in Hierapolis, see Ramsay (1895–7: i. 85–90); Cichorius (1898: 42–3). On the
sanctuary at Hierapolis: Ritti (1989–90). On evidence for the cult of Apollon archegetes: Judeich
(1898: 4); SEG 56.1500. For the oracle, see Pugliese Carratelli (1963–4); Ruffing (2008: 260–1). In
the 1960s Italian excavators identified a cavity to the right side of the temple of Apollo with what
was interpreted as the Ploutoneion (cf. Strabo, Geogr. 13.4.14).
9
See Judeich (1898: 32, 15–17): κονβεντα ̣[ρ]/χήσαντα τῶν Ῥωμα ̣[ί]/ων, and SEG 53.1464 (AE
2007: 1696; SEG 58.1510), Ῥωμαίων κωουένταρχον, with Hatzfeld (1919: 165).
280 Ilias Arnaoutoglou

first century AD there is a clear development of manufacturing activities,

particularly wool-working and dyeing of textiles.10 Strabo refers to the hot
springs in the city, emphasizing the property of fixing colours in woollen
textiles; the plant used, madder root (robia tinctorum), produced a red dye,
similar to purple but a lot cheaper: ‘The water at Hierapolis is remarkably
adapted also to the dyeing of wool, so that wool dyed with roots rivals that
dyed with the coccus or with the marine purple. And the supply of water is so
abundant that the city is full of natural baths.’11
In AD 60, an earthquake destroyed the city as well as the neighbouring cities of
Colossae and Laodikeia.12 Reconstruction started under Domitian (AD 81–96).13
The city prospered under Hadrian, Antoninus Pius, Marcus Aurelius, and
Septimius Severus. At some point between AD 218 and 222, Hierapolis was
granted the title neokoros.14
Most of the epigraphic testimonia of the Hierapolitan associations can be
dated on the basis of letter style to the period between the mid-second and
mid-third century AD.15 Despite the lack of an up-to-date epigraphic corpus,16
two salient features emerge. First, we are short of information about their
internal structure and life. Secondly, associations were involved in the man-
agement of the memory of the deceased: several inscriptions mentioning
associations are in fact part of funerary monuments, and they specify that
fines for the violation of a tomb or endowments are to be paid to a particular
professional group.17

10
Ritti (1995: 71–2) and D’Andria (2001). The rapid development of Hierapolis as an urban
centre after AD 60 may be reflected in the prosperity of craftsmen in the period between the mid-
second and mid-third century, as suggested for towns in the western part of the empire by
Patterson (2006).
11
Strabo, Geogr. 13.4.14: ἔστι δὲ καὶ πρὸς βαφὴν ἐρίων θαυμαστῶς σύμμετρον τὸ κατὰ τὴν
Ἱερὰν πόλιν ὕδωρ, ὥστε τὰ ἐκ τῶν ῥιζῶν βαπτόμενα ἐνάμιλλα εἶναι τοῖς ἐκ τῆς κόκκου καὶ τοῖς
ἁλουργέσιν· οὕτω δ’ ἐστὶν ἄφθονον τὸ πλῆθος τοῦ ὕδατος ὥστε ἡ πόλις μεστὴ τῶν αὐτομάτων
βαλανείων ἐστί, trans. H. L. Jones. Cf. Ritti (1995: 67). On the plants producing the natural dye in
the area, see now Huttner (2009) and Ruffing (2009).
12
See Ritti (1997: 346–8).
13
See, e.g. the bilingual inscription AE 1969/1970, 593, AD 86/87): [Αὐτοκράτορι [[Δομιτιανῷ]]
Καίσαρι Σεβασ]τῷ Γερμανικῷ, ἀρχιερεῖ με[γίσ]τῳ, δημαρ[χι]/[κ]ῆς ἐ[ξουσίας τὸ ζʹ, αὐτοκράτορι
τὸ ιδʹ, ὑπάτῳ τὸ ιβʹ, πα]τρὶ πατρίδος, τὴν πύλην καὶ τοὺς πύ[ργους ἐποίη]σεν Σέξτος [Ἰ]ούλιος
Φρον[τῖνος ἀνθύπατος — — —].
14
See Ritti (2003) and Burrell (2004: 135–41), and the two letters of Hadrian to Hierapolis
sent in AD 117 (SEG 55.1415) and AD 130 (SEG 55.1416).
15
See Ritti (1995: 66, 70), and, more generally, Dittmann-Schöne (2000: 13).
16
Judeich (1898) and more recently Ritti (1995, 2004). See also the overview of epigraphic
research in Ritti, Miranda De Martino, and Guizzi (2007).
17
The question of the origin of these associations cannot be dealt with in this chapter. On this
topic, see Debord (1982) and Ritti (1995). According to Debord (1982: 15, 305), these associations
continue the pre-Hellenistic (i.e. Achaemenid) tradition of guild organization and activities. Ritti
(1995: 68–9) claims that the use of hot water from the sacred water springs, controlled by the
priesthood, for dyeing the wool reinforces the link between the priesthood and the craftsmen, the
latter perhaps being descendants of earlier sacred slaves working in the sanctuaries.
 Hierapolis and its Professional Associations 281

In Lydia, the epigraphic evidence for professional associations peaks at

roughly the same time, but it spans a period from the first to the end of the
third century AD. The existence of a corpus for Thyateira and Saittai facilitates
analysis. As in Hierapolis, we are badly informed about the internal structure
of professional associations; while in Thyateira they do not seem to have been
involved in funerary activities, associations in Saittai are attested only in
funerary contexts, but not in the same way as in Hierapolis, as it will become
clear below.18 The funerary monuments themselves are also different: in
Hierapolis, funerary inscriptions can be found on a variety of monuments;
they frequently appear on sarcophagi placed upon a base, sometimes upon a
building with a room and funerary beds, or upon a mortuary chapel with a
gabled roof.19 By contrast, in Saittai, funerary inscriptions are rather formu-
laic, inscribed on simply decorated stelae. In Thyateira, most epigraphic
evidence is found on statue bases.

W H O WERE F O R M I N G P R O F E S S I O N A L
ASSOCIATIONS?

Of the thirty-three testimonies of associations in Hierapolis, twelve concern

religious associations or associations of a different kind; the remaining twenty-
one inscriptions mention professional associations of craftsmen. Together, these
texts contain twenty-eight references to craftsmen groups, of which seven are
to πορφυραβάφοι (purple-dyers),20 three to βαφεῖς (dyers) and ἐριοπλύται (wool-
washers), two to ἐργαστηριάρχαι (manufacturers), θρεμματικὴ (cattle-breeders),21
κηπουροὶ (gardeners), and χαλκεῖς (coppersmiths), and one each to ἀκαιροδαπισ-
ταὶ (carpet-weavers), ἠλοκόποι (nail-makers), κοπιδερμοὶ (cutlers), λινωταὶ (linen-
weavers?), ἀρτοποιοὶ (bread-makers), and ὑδραλέται (water-mill operators).22

18
On the importance of the funerary evidence for questions of status and social influence, see
van Nijf (2010: 165–71).
19
See the discussion of monumental tombs and their overall environment in Roman Asia
Minor by Cormack (2004).
20
For an overview of the evidence for this profession, see Drexhage (1998).
21
Suggested already by Cichorius (1898: 48). Judeich (1898: 143), followed by van Nijf (1997:
82 n. 152), disagrees. Ramsay (1895–7: i. 119) suggested the far-fetched explanation that it was an
organization for looking after foundlings (from threptoi).
22
Ritti (1995: 72–3; 2004: 486–7) refers to unpublished inscriptions mentioning λινουργοί,
πιλοποιοί, κοπιδερμοί, ἀρτοποιοί, and a σιλιγνειτοπώλης. Evidence for the various crafts is collected
by Ruffing (2008): ἀκαιροδαπιστὴς (p. 400), βαφεῖς (pp. 453–9), ἐργαστὴς (pp. 522–3), ἐριοπλύτης
(p. 525), ἠλοκόποι (p. 537), κοπιδερμοί (p. 601), λινωταί, λινουργοί (pp. 640–7), πορφυροβάφος
(p. 725–7), χαλκεῖς (pp. 814–19). Earlier but still useful is Zimmermann (2002). For κοπιδερμὸς,
see also IK 36.250, II 21 (Tralleis, third/fourth century AD): τόπος ἐν Παρκάλλοις Ἀλεξάνόρου
Κοπιδέρμου ζυ(γὸν) κ(εφάλαιον). Κοπιδέρμου should be understood as a profession; cf. also
MAMA 3.573 (Korykos, third/fourth century AD) for a κοπιδᾶς. For other professions in Hierapolis,
282 Ilias Arnaoutoglou

The prominence of the association of purple-dyers is remarkable: about a third of

the inscriptions mentioning professional associations refer to it. In general,
associations of craftsmen involved in the clothing industry seem to hold a
dominant position: fifteen out of twenty-eight references are related to this branch
of the economy.23
In Lydia, professional associations also brought together craftsmen involved in
textile production—more than half of the inscriptions from Thyateira and the
area of Saittai (thirty-two our of fifty-three) pertain to professional groups in this
branch of the economy. There are linen-workers and linen-weavers (λινουργοί,
λινύφοι)—linen-workers seem prominent in the area of Saittai (fourteen out of
thirty-two references)—wool-workers, wool-sellers (λανάριοι, ἐριουργοί, ἐριοπῶ-
λαι), felt-workers (πιλοποιοί ), fullers (γναφεῖς), weavers (ὑφανταί ), tow-makers
(σιππινάριοι), and dyers (βαφεῖς). Dyers appear particularly active in Thyateira
(nine out of twenty-one mentions). Beyond textiles, there are organizations of
people involved in leather-processing—tanners (βυρσεῖς) and shoemakers (σκυ-
τεὺς, σκυτοτόμοι)—bread-making—bakers (ἀρτοκόποι, ἀρτοποιοί ) and ‘winter-
wheat’ sellers (σιλινάριοι)—doll-makers (κοραλλιοπλάσται), coppersmiths
(χαλκεῖς, χαλκοτύποι), carpenters (τέκτονες), carpet-cleaners? (ψιλαγνάφοι),
potters (κεραμεῖς), gardeners (κηπουροί ), musicians (μουσικοί ), and people
involved in the slave trade (προξενηταὶ σωμάτων, ἐργασταὶ σταταρίου).

HOW DID CRAFTSMEN NAME THEIR GROUPS?

In Hierapolis, craftsmen, in order to denote an association, used two categories

of terms. The most widely employed terms underline the craft element behind
the organization: ἐργασία (ten inscriptions), συντεχνία (three inscriptions),
τέχνη (three inscriptions); less frequently used were more generic terms such
as συνέδριον (three inscriptions), and κοινόν. It is important to emphasize that
these designations do not reflect any difference in the nature of the organiza-
tions; all refer to associations of craftsmen.24 Take, for example, the ‘most

see AAST 101 (1966–7): 313, no. 37; Judeich (1898: 156): πορφυροπώλης; Judeich (1898: 75): κάπηλος;
no. 135: ζωγράφος; no. 222: παστιλλᾶς; no. 262: μυροπώλης (cf. SEG 54.1302); no. 274: τεχνίτης; SEG
33.1139: ἀνδριαντουργός. Note also SEG 57.1385 (Colossae, second/third century AD): διφθεροπὺς for
διφθεροποιὸς.
23
A similar assessment was made by Ritti (1995: 71–2).
24
Ritti (1995: 68–9) and Dittmann-Schöne (2000: 25). For remarks on the nomenclature of
the professional associations, see Labarre and Le Dinahet (1996: 59). Cf., for Phrygia, Dittmann-
Schöne (2000: 18), and, for Lydia, Arnaoutoglou (2011: 263–4). However, such a division does
not appear in Hierapolis: Ritti (1995: 66).
 Hierapolis and its Professional Associations 283

revered’ (σεμνοτάτη) association of purple-dyers. This association has been

described in two different ways, ἡ ἐργασία and ἡ τέχνη τῶν πορφυραβάφων.25
Similarly the dyers are designated both as ἡ τέχνη and as ἡ ἐργασία τῶν
βαφέων. By contrast, the association of wool-washers is called in both in-
stances ἡ ἐργασία τῶν ἐριοπλυτῶν; similarly, gardeners are described as ἡ
ἐργασία τῶν κηπουρῶν.
A few words are needed to interpret the terms proedria and synedrion,
which are unusual for associations.26 The term synedrion had a wide range of
meanings such as ‘gathering’ and ‘assembly’; by extension, it could also mean
the association as such, especially in the context of crafts.27 More puzzling is
the case of the term proedria: in an epigraphic context, this term denotes
almost without exception the privilege to be seated in the front rows during
theatre performances and musical or athletic contests, granted to distin-
guished citizens and foreigners.28 However, it is also attested in the nomen-
clature of the association of purple-dyers of Hierapolis.29 The term features in
three inscriptions; two of these do not shed any light on its meaning. The
third, nonetheless, provides some clues; it reads:
εἴ τις δὲ | ἀποκορακώσει ἤ τε κληρονόμος ἤ τι συνγενὴς | θήσει τῇ προεδρίᾳ τῶν
πορφυραβάφων ἤ τοῖς | κατὰ ἔτος ἐπειμεληταῖς προστείμου * υ’
If anyone opens [the sarcophagus’ lid], either heir or relative, he shall pay to the
proedria of the purple-dyers or to the annually appointed stewards a fine of 400
denarii.30

25
Cf. Cichorius (1898: 51): ‘Der Name der Zunft ist auf den Inschriften regelmassig ἡ ἐργασία
τῶν πορφυροβάφων.’
26
Synedrion: τὸ συνέδριον τῶν ἀκαιροδαπιστῶν in SEG 46.1656, τῷ συνεδρίῳ τῶν κοπιδέρμων
in Ritti (2004: 544 [ineditum]). Combination of both in Judeich (1898: 227): τὸ συνέδριον τῆς
προεδρίας τῶν πορφυραβάφων.
27
For example, συνέδριον τῆς γερουσίας: Judeich (1898: nos 73, 312); AAST 101 (1966–7),
296, no. 4; συνέδριον τῶν νέων: Judeich (1898: 117). Other examples: IG 12.8, 388, IPriene 246
(συνέδριον τῆς γερουσίας); IG 12.7, 271, 14; TAM 2.225 (Sidyma, Lycia, imperial) (συνέδριον τῆς
βουλῆς); IPerinthos 55 (τὸ συνέδριον τῶν φιλαπαμέων); for professional associations see IMilet ii
939 (συνέδριον τῶν λινουργῶν), IK 13 (Ephesos) 636 (τὸ ἱερὸν συνέδριον τῶν ἀργυροκόπων), IK 16
(Ephesos) 2212 (τὸ συνέδριον τῶν ἀργυροκόπων). See Poland (1909: 156–8), who considers the
term as highly significant, since it was used by heavyweight groups in the cities, such as the
gerousia. Synedrion can also mean the meeting place, as in Ath. Agora 16, 181.39–40 (Athens,
282/1 BC): καὶ στῆσαι ἔμπροσθε τοῦ συνε/δρίου; Petrakos (1999: 10, 11) (248/7 BC): ἀνατέθηκε καὶ
συνέδριον πρὸς τῶι στρατηγίωι.
28
e.g. in Phrygia, IK 49.1.22 (Laodikeia on Lykos, 267 BC); 2.17 (third century BC).
29
Judeich (1898: 227), SEG 46.1656, and SEG 54.1323. Recently another instance of proedria,
this time proedria erioplyton, appeared in a fragmentary funerary inscription published by
Guizzi, Miranda De Martino, and Ritti (2012: 659, no. 15) and dated on the basis of the letter-
forms to the second or third century AD; in particular, the proedria of the wool-washers is entitled
to use the amount of money bequeathed to perform a stephanotikon, perhaps for a different
association.
30
SEG 54.1323, 5–8.
284 Ilias Arnaoutoglou

Here, proedria does not designate the association as such but most probably
refers to the board of the group.31 In particular, proedria is put on a par with the
annually elected stewards (epimeletai): the text refers to a fine for tomb violation
that will be paid either to the proedria of the group or to its epimeletai. Therefore,
proedria and epimeletai enjoy some sort of equal status. I think it is unlikely that
the clause refers to a payment to the group (that is, proedria) distinct from its
epimeletai. In a sense proedria should be understood as something similar to the
expression recorded in another inscription from Hierapolis, which refers to ‘οἱ
τῆς ἐργασίας τῶν ἐριοπλύτων οἱ μετεχόντες τῶν ἐπιμελημένων’—‘those partici-
pating in the administration of the ergasia of wool-washers’.32 In other words,
proedria cannot denote something entirely different from epimeletai: it should be
exercising a similar function. Nevertheless, there was a distinction based on
status enjoyed by those included in the proedria. Whether epimeletai enjoyed
the privilege of proedria ex officio remains difficult to substantiate. The proedria
of purple-dyers may originate in the honorary places at the theatre allocated
to leading and/or prominent members of the association, such as the seats of
linourgoi in the theatre of Saittai or those of the undefined ergasia in that of
neighbouring Laodikeia on the Lykos. In practice, this would imply some sort
of hierarchy among purple-dyers. If this interpretation is correct, it gives an
interesting view of the administrative structure of this particular association: it
means that there were the annually selected stewards and, alongside it, a body
of officials—perhaps ex-epimeletai—constituting the proedria, which could
convene perhaps as a separate body.33 It is, therefore, possible to consider the
proedria as some sort of executive committee, comprising all ex-magistrates of an
association; yet, as far as I know, there is no evidence for a comparable phenom-
enon among the other associations in Hierapolis, or in the Greek-speaking
Roman East in general. An alternative interpretation would see in the proedria
a board of senior, prestigious, wealthy, and distinguished members of the group.
In Thyateira and Saittai, the character of professional associations is
expressed by a variety of terms, such as θίασος, ὁμότεχνον, πλατεῖα, πλῆθος,
συμβίωσις, συνεργασία, σύνοδος, and φυλή. While there is no clear pattern in
the adoption and use of a particular name, two main categories of terms can be
distinguished. First, there are the terms borrowed from other areas of

31
See also Ziebarth (1896: 107); Cichorius (1898: 49); Poland (1909: 126, 415); Labarre and Le
Dinahet (1996: 59); Dittmann-Schöne (2000: 35); Zimmermann (2002: 37–9); and Harland
(2006a: 237). The term proedria appears with a different meaning in IK 12.217.3–5 (Ephesos):
καὶ νῦν φανερωθῇ ἥ τε ἐξ ἀρχῆς ἔκ τε τῶν νόμων καὶ τῶν θείων | διατάξεων καὶ τῶν δογμάτων τῆς
ἱερᾶς συνκλήτου τῇ ὑμετέρᾳ μητροπόλει | προεδρία προσνενεμημένη; ‘and in the honorary decree
of the council and the people of Attaleia in Lydia for Dionysios, son of Glykon’ (LGPN vA (490)),
TAM 5.2.829, 5–7 (AD 211–217?): με|τέχ ̣οντα κ(αὶ) τῆς προε|δρίας.
32
AAST 101 (1966–7), 317, no. 45, 5–6: ἀποτείσει|τοῖς τῆς ἐργασίας τῶν ἐριοπλύτων τοῖς
μετεχοῦ|σιν τῶν ἐπιμε<με>λημένων * τʹ: ‘he shall pay to those participating in the administration
of the ergasia of wool-washers 300 denarii.’
33
In this sense, to synedrion tes proedrias in inscription: Judeich (1898: 227).
 Hierapolis and its Professional Associations 285

associational activity, such as religious groups (θίασος and συμβίωσις), per-

formance groups (σύνοδος), or from public subdivisions (φυλὴ); secondly,
there are those that seem to be used by artisan groups exclusively. Some
emphasize a spatial dimension (πλατεῖα, six cases), or the quantity of people
involved (πλῆθος, two cases), but terms stressing craftsmanship (ὁμότεχνον: six
cases) or the cooperative aspect of the activity (συνεργασία, eighteen cases) are
much more numerous. As far as chronology is concerned, most associations
designated as ὁμότεχνον occur around the mid-second century AD; by contrast,
the term συνεργασία occurs exclusively in Saittai in the period from the middle
of the second century AD to the first decades of the third. Moreover, there are
several interesting geographical peculiarities. First, the term φυλὴ appears
almost exclusively in Philadelphia: the only exception is made up by the
inscribed seats in the stadion of Saittai. Secondly, at Saittai, the variety of
designations is rather limited: there is a substantial predominance of συνερ-
γασία. Finally, at Thyateira, designations are almost completely lacking.
A significant, but usually overlooked, difference between Hierapolis and the
Lydian towns of Thyateira and Saittai lies in the agent—the user of the terms
designating an association. In Hierapolis few inscriptions reveal the way
craftsmen groups were self-designated, while in the Lydian cities there is
hardly any document that has not been issued by the association itself. This
discrepancy makes comparison a delicate operation, since in Hierapolis the
denomination reflects the picture that outsiders had of associations, while in
Lydia the denomination is a reflection of how the members themselves defined
their own group.

THE S OCIAL P OSITIONING OF PROFESSIONAL

ASSOCIATIONS

Hierapolitan professional associations display a well-balanced record of social

activities in both public and private spheres. They pay the honours due to the
local council and local leaders, but also to the procurator of the emperor.34 In a
fragmentary inscription the association of purple-dyers honoured an un-
named ἐπίτροπος τοῦ Σεβαστοῦ (procurator Augusti), designated as their
benefactor on every occasion or affair (ἐν πᾶσιν).35 The text suggests that the
association had—or rather wished to demonstrate that it had—access to the

34
Procurator: Judeich (1898: 42). Council: SEG 41.1201 with Ritti (1989–90: 871 n. 7);
Harland (2006b: 40) and Ruffing (2008: 264–5). Local dignitaries: Judeich (1898: 40), with
Harland (2006b: 41) and SEG 56.1499; his funerary monument probably in Judeich (1898: 46).
35
Judeich (1898: 42).
286 Ilias Arnaoutoglou

Roman authorities and perhaps could exercise some leverage on their deci-
sions.36 The association of dyers honoured the council of Hierapolis with the
erection of a monument on a marble base, and the purple-dyers honoured
Tiberios Klaudios Zotikos Boas, a personality with a considerable career, since
he had been πρῶτος στρατηγὸς (leading? general),37 ἀγωνοθέτης, γραμματεὺς
of the emperors’ temples in the province of Asia, πρεσβευτής, and ἀρχιερεὺς
(chief priest).38 Tiberios Klaudios Zotikos was also honoured by the associ-
ation of wool-washers with an almost identically worded inscription, the only
difference being the committee of three individuals charged with supervising
the inscription of the decree on stone.39 The elevated socio-economic status of
the association of purple-dyers in Hierapolis is more clearly revealed in an
early third-century AD inscription, which informs us that the association had
contributed a substantial amount of money for the decoration of the ceiling of
the local theatre:
συνετέλεσεν δὲ καὶ πρὸς τὸν κόσμον τῆς τε πρώτης καὶ τῆς δευτέρας στέγης λίθου
Δοκιμηνοῦ ἀπηρτισμένου [καὶ πρὸς τὴν προ]σάρτησιν παρ' ἑαυτῆς πόδας ἑξακοσ
ίους πεντήκοντα τρεῖς, ἡ τέχνη τῶν πορφυραβάφων.
The association of purple-dyers have contributed towards the ornamentation the
ceiling of both the first and the second storey in worked Dokimeion marble and
to that part added to it, for 653 feet.40
This shows that the corporation of purple-dyers had access to capital funds
that enabled them simultaneously to contribute to the decoration of one of
Hierapolis’ urban hallmarks, display their attachment to civic ideology, and
elevate themselves above the status of the average craftsmen, and alongside the
ranks of local benefactors.
The highest number of honorary inscriptions issued by professional asso-
ciations in Asia Minor comes from Thyateira; among them the association of
dyers (9/20 cases) seems to have been particularly active. Associations of
craftsmen in Thyateira honoured male Romans, Romanized Greeks, or
Greeks.41 All these people were members of the local elite who had exercised

36
See also the remarks of Cichorius (1898: 49).
37 38
For strategos, see Ritti (1997: 344) and Dmitriev (2005). Judeich (1898: 41).
39
Judeich (1898: 40).
40
SEG 35.1369 (AD 206–9): trans. Ritti (1995: 75). See also Ritti’s assessment (1995: 76) that
the accumulation of capital by the professional associations involved in the manufacture of
luxury goods elevated them to the position of local euergetai. Note that the inscription recording
the contribution of the purple-dyers follows the dedication of the theatre by Hierapolis to the
patron deities and the imperial family. On who paid for the building of theatres, see Sturgeon
(2004). For the sculpted reliefs of the Hierapolitan theatre, see Cubak (2008).
41
With the exception of Kl. Ammion (LGPN vA (99)) in TAM 5.2.972, all the honoured
individuals are men. For office-holding by women, see van Bremen (1996). Roman or Romanized
Greeks: TAM 5.2.933, 936, 945, 965, 978, 1002, 1019, and SEG 49.1669. Greeks: TAM 5.2.932,
989, 991.
 Hierapolis and its Professional Associations 287

functions associated with local offices—ranging from the more modest offices
like the agoranomia to the more significant ones such as the function of
secretary of the boule and the demos, or that of strategos.42 Yet the professional
associations of Thyateira could aim at even higher echelons of power: there are
at least two texts in which persons of consular rank were honoured.43 While
Saittai provides the largest amount of testimonia for professional associations
in Lydia and among the most numerous in Roman Asia Minor, the informa-
tion they supply is limited to a date, the name of the deceased, and that of the
association, and does not refer to the role played by these associations in
civic life.
As noticed already by Ziebarth in the late nineteenth century, the extent to
which professional associations imitated the evolving discourse of the polis is
striking. This is particularly obvious in the procedures and the language of
honouring. In Hierapolis—as no doubt was the case in several other cities as
well—the adjective σεμνοτάτη, usually used to qualify the gerousia,44 is also
assigned to the purple-dyers.45 Yet it is not the purple-dyers themselves who
employed the adjective in the inscriptions they produced; rather, individuals
used this phrase when referring to the association of purple-dyers in the epitaphs
on their tombstones. In practical terms, it suggests that the association of purple-
dyers was seen as a revered, respectable collective body, (almost) equal in social
standing to the gerousia. Another example concerns the honorific language used
in some texts. In several inscriptions from Hierapolis the term στεφανωτικὸν
(stephanotikon) appears. This refers to the annual ritual of laying a wreath on the
tomb. One text reports a case in which the stephanotikon was to be performed by
the association of purple-dyers.46 Its costs were covered by the endowment made
to the association. In Saittai a garland is sometimes depicted on the otherwise
austere funerary monuments.47

42
e.g. TAM 5.2.932 (agoranomos), TAM 5.2.1002 (agoranomos, ambassador, legal consult-
ant). Cf. Arnaoutoglou (2011: 266–70).
43
TAM 5.2.935.986.
44
Judeich (1898: nos 67, 73, 98, 111, 146, 194, 209, 234, 290, 315); SEG 33.1123; AAST 101
(1966–7), 313, no. 38. The same collocation appears in Lycia (TAM 2.210: Sidyma; TAM 2.294,
325, FXanthos 7, 87: Xanthos), in Myra (Luschan and Petersen (1889: 2.2, 36, no. 57; 45, no. 82))
and in Pamphylia (IK 43.26: Side; 54.291: Perge). Σεμνοτάτη βουλὴ is attested in Athens, Rhodes,
and Ephesos, while the fiscus (ταμεῖον) is most often qualified as ἱερώτατον.
45
Associations: IG 22.1369: Attica, σεμνοτάτην σύνοδον; IK 60.63: Kibyra, σεμνοτάτη συνερ-
γασία; Judeich (1898: 42); SEG 54.1313; 56.1499: σεμνοτάτη ἐργασία; SEG 46.1656: σεμνοτάτη
προεδρία. See also Poland (1909: 170).
46
SEG 46.1656.
47
Van Nijf (1997: 63–4) and McLean (2002: 277). This honorary practice had already
appeared in the second century BC in Rhodes, when the cult association of Haliastai–Haliadai
decided to crown the tomb of its head eranistes and benefactor, Dionysodoros of Alexandreia. Cf.
IG 12.1.155, 66–9, with Fraser (1977: 62–8). In second-century BC Kyme in Aeolis (IK 5.13
VI. 45–8), in Priene (IPriene 99), and in Egypt (e.g. SEG 8.529, 42–4 (64 BC); P. Mich. 5.243(AD
14–37)). See the explicit reference to stephanosis in SEG 57.1212 (Saittai, Hellenistic period).
288 Ilias Arnaoutoglou

THE S OCIAL COMPOSITION OF PROFESSIONAL

ASSOCIATIONS

As far as the social composition of professional associations is concerned, it makes

sense to focus on the ethnic background of the individuals involved, and on their
socio-economic position with respect to local elites. However, it should first be
noted that it is not always straightforward to see, in texts, who actually was a
member of a professional association. In particular, there has been some debate
about the affiliation of the deceased who made an endowment to a professional
association or designated it as the beneficiary of the fine for tomb desecration.
While some have believed that the deceased as a rule were members, some
scholars, including Ritti and Harland, have recently argued that this is not
necessarily the case.48 In arguing that the situation was not so straightforward,
Ritti used an unpublished inscription in which a μολυβδουργός bequeaths prop-
erty to a θρεμματική association, an equal amount to an association of ἐργαστη-
ριάρχαι and a lesser sum to an association of χαλκεῖς.49 It is difficult to accept
individuals belonging to two different associations, as there was a roughly con-
temporary imperial edict prohibiting individuals from doing so.50 This opens up
the possibility for other scenarios. The μολυβδουργός in question could have had
business links and dealings with the ἐργαστηριάρχαι and χαλκεῖς, while his
preference for the θρεμματική association could be interpreted in social terms.
Harland is thus probably right that not all individuals referred to in funerary
inscriptions mentioning an endowment necessarily belonged to the corporate
groups.51 In the case of Hierapolis, none of the deceased endowing professional
associations is otherwise known; thus their relation with these professional
associations remains completely unclear, and they should not be included in
our analysis of the social composition of professional associations.
As to the ethnic background of the membership of professional associations, the
committee of four men elected from among the dyers of Hierapolis to supervise
the erection of the statue of the council consisted of two Greeks (Γλυκωνιανός
Μένανδρος τοῦ Γλύκωνος, Ἀντίοχος τοῦ Ζωσίμου), one Roman (Γάιος Ἰούλιος
Λούκιος Φαβίᾳ Κρήσκεντος), and one Romanized Greek (Γάιος Βαίβιος Γλύκων
β'), using their respective onomastic conventions.52 A similar committee of wool-
washers was made up of three Romanized Greeks (Μᾶρ(κον) Αὐρ(ήλιον) Ἀπολ-
λώνιον δὶς Πυλωνᾶ, Μᾶρ(κον) Αὐρ(ήλιον) Ἀμμιανὸν Ἀμμιανοῦ δὶς ̣ τοῦ Γλύκωνος,
Αὐρηλια ν̣ ὸν Ἑρμίππου [Ἀρ]ρουντιανόν).53 It may be premature to draw a conclu-
sion about the social composition of professional associations on the basis of only

48
Cichorius (1898: 49) believed the deceased were generally members.
49 50 51
Ritti (1995: 70–1). Dig. 47.22.1. Harland (2006a: 234).
52
SEG 41.1201. Ritti (1989–90: 871, no. 7); Labarre and Le Dinahet (1996: no. 56); Dittmann-
Schöne (2001: 222); Zimmermann (2002: 196); Guizzi, Miranda De Martino, and Ritti (2012: 648).
53
SEG 56.1499.
 Hierapolis and its Professional Associations 289

two inscriptions, but the evidence at Hierapolis so far suggests a majority of

Romanized Greeks. Remarkable is the tombstone of Poplios Aelios Glykon
Zeuxianos Ailianos, who endowed the association of purple-dyers and the assem-
bly of carpet-makers with 200 and 150 denarii respectively in order to perform
celebrations at two Jewish festivals—Passover and Pentecost—and at the Roman
festival of the Kalendae.54 While there is a certain amount of evidence for Jews in
Hierapolis, there is no compelling reason to assume that there were exclusively
Jewish professional associations of craftsmen: in Roman Asia Minor: professional
associations were very rarely constituted along clearly demarcated ethnic lines.55
As far as their socio-economic position is concerned, an instructive example of
upward social mobility is the case of Μᾶρκος Αὐρήλιος Ἀλέξανδρος Μοσχιανοῦ
βουλευτὴς πορφυροπώλης, whose tomb proudly features his profession—he was a
purple-seller—but also lists his membership of the local council.56 In the case of
certain craftsmen and traders, their occupation seems to have gained them social
prominence.57

MEMORY MANAGEMENT, L AW, AND ASSOCIATIONS

What is really unique in the professional associations of Hierapolis are the

different ways in which they were involved in the preservation of the memory
of the dead.58 Associations could be designated either as recipients of fines,59
or as beneficiaries of legacies, or both.60

54
SEG 46.1656. For the most recent discussion of the complexities of cultural identities
attested in the inscription, see Harland (2003a: 207–10; 2006a: 238–9).
55
See IJO II.187–209 and Harland (2006a).
56
Judeich (1898: 156) and another (anonymous) purple-seller in AAST 101 (1966–7), 313,
no. 37. See also the case of Titus Flavius Zeuxis in Judeich (1898: 51; with SEG 54.1304).
57
Pleket (1983: 141–2); Ritti (1995: 76); Levick (2004: 190); Ruffing (2004: 96–7). Tomb-
stones of other bouleutai: Judeich (1898: 145, 163); AAST 101 (1966–7), 298, nos 8–9. See also a
wealthy purple-seller in Mysia, IK 26.35 (Miletupolis; first/second century AD); a shoemaker
carrying the title of proboulos in Magnesia on the Meander, IMM 111 (first century BC).
Discussion: Zuiderhoek (2011: 190–1).
58
See the excellent discussion in van Nijf (1997: 57–64) about the different facets of profes-
sional associations’ involvement in memory management. Van Nijf (1997: 59) concludes: ‘We can
therefore see the inscriptions that mention fines as expressions of how men of middling wealth
and status perceived the social hierarchy, and of how they saw the place of collegia therein.’ The
evidence for Hierapolis was collected by Ritti (2004: 562–6). Yannakopoulos (2008: 306–7)
confirms that the gerousia was receiving endowments from individuals of high social status.
59
See van Nijf (1997: 57–8); the size of the fines intended for professional collegia lower end
of the scale. In particular in Hierapolis the fiscus/tameion is designated to receive fines ranged
between 50 and 10,000 denarii (usually 2,500), while gerousia is called to collect fines between
100 and 2,500 denarii (usually 500).
60
For a parallel case from imperial Ephesos, see IK 16.2446. In neighbouring Laodikeia on the
Lykos, professional associations are not designated as administrators of endowments.
290 Ilias Arnaoutoglou

The responsibilities of professional associations as recipients of fines for tomb

violation and as beneficiaries of endowments are characteristic of the central role
they (and especially the association of purple-dyers) played in the society of
Roman Hierapolis.61 In three cases, professional associations were designated as
recipients of fines for tomb desecration—together with the polis-treasury, in two
cases they were to use the fine as the capital of an endowment, while in five cases
they were beneficiaries of a legacy.62 This role not only had profound social effects;
it also highlights the way in which associations were legally treated. A decade ago,
I argued that Roman authorities followed a hands-off policy with regard to
associations in Asia Minor and intervened only when and where public order
required it. The evidence from Hierapolis confirms this impression: professional
associations performed quasi-juridical duties, since they could be designated as
beneficiaries of legacies or appointed as recipients of fines, and consequently, one
would imagine, they could appear in the local law courts or before the Roman
authorities to denounce a violator or to claim the fine.63 Professional associations,
with their prestige, participated in memory management, which allowed them to
build up a social reputation as decent and pious administrators, and therefore as
trustworthy partners. To a lesser extent, perhaps with the aid of a bit of creative
accounting, it also allowed them to fill the coffers of the group.
The evidence from Hierapolis is unique in one more respect: perhaps
because of a lack of capital, the amount of the fine is often added to the legacy.
A good example is provided by the inscription on the tomb of Aurelia Pakonia
Pauline.64 The fine for the violation of the tomb—300 denarii—is provided to
be paid to the association of wool-washers, and it is designated to be used as an
endowment to honour the already deceased son of the testator, Tatianos, by
laying a wreath on his tomb on a specific date. In this way the association has
two reasons to feel compelled to perform the task assigned to it: it was
encouraged to protect the tomb, not only to receive the fine but also to be
able to perform the ritual every year—thereby enhancing its prestige as a
reliable partner. However, what remains unclear is whether the posthumous

61
Out of eighteen inscriptions from Hierapolis mentioning endowments, seven involve
professional associations.
62
Van Nijf (1997: 59–60). For gerousia, see now Yannakopoulos (2008: 196–8 (in Hierapolis)
and 292–342). See also IK 16.3216 in which οἱ ἐν Ἐφέσῳ ἐργάται προπυλεῖται πρὸς τῷ Ποσειδῶνι
are endowed with 500 denarii ἐπὶ τῷ γείνεσθαι ἀπὸ τοῦ γεινομένου τόκου οἰνοποσίαν καὶ
κηριόλους καὶ στεφά(νους) * ι’ ποιήσουσιν δὲ καὶ τὴν εὐωχίαν μη(νὸς) Ποσ(ιδεῶνος); in case
they fail to do so, the measurers of the grain (prometrai) of Ephesos shall legally exact the amount
of money.
63
Of course, one may argue that, according to Digest 34.5.20, collegia were allowed to receive
legacies; in this context we have to treat accordingly the surge of related inscriptions in
Hierapolis; cf. Waltzing (1895–1900: 2.2.456–62) and Robertis (1970). There is also some similar
evidence from Ephesos: IK 16.2115; 2446.2 (imperial period). Yet, the practice had already been
widespread before the arrival of the Romans, as is shown by the Hellenistic material in Laum
(1914) and, recently, by SEG 58.1640 (Tlos [Lycia]; second century BC).
64
AAST 101 (1966–7), 317, no. 45.
 Hierapolis and its Professional Associations 291

honours would also have been performed if no one had violated the tomb or
when it was not possible to identify who had desecrated it, or even when the
desecrator could not (or would not) pay the fine.
Another question concerns the compulsory character of the ritual perform-
ance; what would have happened if the association was unable or unwilling to
perform the ritual? Five inscriptions show what could happen in such a
situation. In the text on the tomb of Aurelios Zotikos Epikrates the testator
designates three different associations as recipients of 150 denarii to perform
the laying of a wreath over his tomb; if the nail-smiths fail, then the copper-
smiths will take over, and if they also fail to perform the ritual, the purple-
dyers shall get the endowed amount of money.65 The inscription on the tomb
of Markos Aurelios Diodoros Koreskos stated that, if the purple-dyers did not
perform the ritual, the association of cattle-breeders would assume the task,
and would receive the money to perform it.66 The epitaph of Markos Aurelios
Amianos stated that, if the association of linen-workers failed to perform the
stephanotikon,67 they not only had to pass the money to the philoploi (‘lovers
of armed battles’) but also had to provide double the amount—as a penalty.68
Two recently published texts contain a clause stating that, if the association
encumbered with the performance of stephanotikon fails to comply, a different
association will be allowed to exact the bequeathed amount of money.69 The
clause implies that the originally designated association has already received
the amount of money for the performance of the ritual.
The question remaining is why we find such a concentration of evidence
for this particular practice in Hierapolis. Is this due to the quantity and
quality of our sources or can it be ascribed to localized patterns of euerget-
ism? Among the published inscriptions from Hierapolis, there are at least six
texts in which an endowment to the council, to the gerousia or to a compar-
able organization is mentioned.70 The sums provided range between 300 and
2,500 denarii.71 At the same time there are eight endowments to professional
associations in which the sums bequeathed range between 150 and 3,000
denarii, though in five of these the sum is less than 500 denarii.72 It is likely

65 66
Judeich (1898: 133). Judeich (1898: 227).
67
The same term for the ritual occurs only in neighbouring Laodicea on the Lycus: IK
49.84–5 (imperial period); see Yannakopoulos (2008: 314–18).
68
SEG 56.1501.
69
Guizzi, Miranda De Martino, and Ritti (2012: 659, no. 15): [τοῖς]/ παραγενομένοις καθ’ ἔτος
ἑκάστῳ τὸ αἱροῦν, εἰ δὲ μὴ δοθῇ τῇ ὠρισμένῃ ἡμέρᾳ,/ πραχθήσεται τούτου ὑπὸ τῆς προεδρίας τῶν
ἐριοπλυτῶν; p. 660: τὸν γεινόμενον κατ’ ἔτος τόκον δίδοσθαι τοῖς παραγεναμένοις ἐπὶ τὸν τόπον
στεφανωτικοῦ ὀνόματει τὸ αἱροῦν, ἅτινα ἐγδώσουσιν οἱ κατ’ ἔτος γραματεῖς μη(νὸς?). δ’ ἡ(μέρα?),
εἰ δὲ μὴ δοθήσεται πρὸς ἔτος, πραχθήσονται ὑπὸ τῆς ἐργασίας τῶν ἀρτοποιῶν.
70
Judeich (1898: nos 67, 209, 234, 278, 293, 336).
71
300 denarii: Judeich (1898: 209, 278, 293); 2,500 denarii: Judeich (1898, 234).
72
Judeich (1898: 133, 153, 195, 227, 342); AAST 101 (1966–7), nos 23, 45; SEG 56.1501; 150
denarii: Judeich (1898: 133); 3,000 denarii: Judeich (1898: 227).
Table 12.1. Hierapolis: professional associations and the management of memory
Inscription Amount Collectivity Purpose

SEG 54.1313 1,000 Association of gardeners, the denouncer Fine for violation
denarii 300
denarii
AAST 101 (1966–7), 500 denarii The treasury, the association of water-mill operators Fine for violation
297, no. 7 300 denarii
Judeich (1898: 218) 500 denarii The treasury, the association of ?gardeners Fine for violation
300 denarii
SEG 54.1323 400 denarii The association of purple-dyers Fine for violation, which will serve as an endowment to
the association. Interest: 144. Purpose: Feast on the grave
AAST 101 (1966–7), 300 denarii The association of wool-washers Fine for violation, which will serve as an endowment to
317, no. 45 the association. Cost of the ritual, fifteen denarii per year.
Purpose: stephanotikon
Judeich (1898: 227) 3,000 The associations of purple-dyers and thremmatike in Endowment; purpose: grave ritual (burning seeds)
denarii succession
SEG 54.1315 1,000 The association of dyers Endowment; purpose: stephanotikon
denarii per
year
SEG 56.1501 250 denarii The association of linen-workers; if the linen-workers Endowment; purpose: stephanotikon
fail to observe the ritual, they shall pay double that
amount to the philoploi (lovers of armed combat)
SEG 46.1656 200 denarii The association of purple-dyers, the association of Endowment; purpose: stephanotikon
150 denarii akairodapistai
Guizzi, Miranda De 200 denarii An unknown entity Endowment?; purpose: stephanotikon
Martino, and Ritti
(2012: 659, no. 15)
Judeich (1898: 133) 150 denarii The association of nailsmiths, coppersmiths, and purple- Endowment; purpose: stephanotikon
dyers in succession
Guizzi, Miranda De 100 denarii The association of linen clothes makers Endowment; purpose: stephanotikon
Martino, and Ritti
(2012 : 660)
 Hierapolis and its Professional Associations 293

that this cluster of evidence for endowments reflects a more widespread

diffusion of this particular practice, at least in Hierapolis, if not in Phrygia
as a whole. One could interpret it as an expression of middle-class euergetism
that parallels similar practices pertaining to the fiscus or the city’s council or
even the local gerousia.
Last but not least, two questions need to be addressed. First, why did
individuals turn to professional associations to safeguard the integrity of
their last residence and, second, what was the relation between the donor
and the professional association(s) encumbered with the receipt of the fine or
the endowment? It has been argued for a long time that the involvement of
associations in funerary rituals is a testimony of weakened family links during
the Hellenistic and Imperial period; individuals could not rely on (close or
distant) relatives for the performance of the customary honours over their
tomb. However, the material from Hierapolis points in another direction.
Besides the twelve inscriptions listed in Table 12.1, there are some forty
references to the gerousia as enforcing the fine for tomb violation or benefi-
ciary of an endowment. Yet, the great majority of funerary inscriptions
suggests that people still relied on their immediate family to pay the post-
mortem honours.73 The involvement of gerousiai and associations is due to
the desire of the owner of the tomb to display his social standing and prestige.
There was no need also to be a member of the group: socializing and
acquaintance with some of their members provided sufficient ground for
including the association among the beneficiaries. It should also be reminded
that having a respectable association performing funerary rites did not replace
the standard honours paid to the deceased by families but rather supplement-
ed them. Naming a respectable association to perform funerary rites over a
tomb can be regarded as a way to secure social distinction and social capital
even after death: the individuals who could afford to be buried in the lavish
sarcophagi in question were not members of the lower classes. For profes-
sional associations, taking care of a tomb or administering an endowment of a
distinguished, even middle-class, citizen provided a ready opportunity to
boost their respectability within the community.
Professional associations are conspicuously absent from any cult activity in
their own right; this is due, perhaps, to the kind of material we have, and it can
be argued that fulfilling the wishes of the deceased and performing stephano-
tikon on the grave should be included in the concept of cult activity.74

73
See, e.g. SEG 57.1212 (Saittai, Hell.); 1148A (Daldis-Charakipolis (Lydia) AD 11/12); 1153
(Daldis area, AD 152/3 or 206/7); 1178 (Iulia Gordos, AD 230/1).
74
See Ritti (1989–90), Dittmann-Schöne (2002). Cf. Harland (2003b: 23), who argues that
‘cultic honours for the imperial gods, which paralleled the sacrifices, mysteries and other rituals
directed at traditional deities, were a significant component within numerous associations’. See
Rebillard (2003: 53–5) for the priority of the family over the collegia in the funerary rituals.
294 Ilias Arnaoutoglou

CONCLUSIO N

The epigraphic evidence for professional associations in Hierapolis is more

varied than it is in Saittai or Thyateira. The majority of texts referring to these
groups comes from a funerary context, and were inscribed on sarcophagi; in
many cases, associations are designated as receivers of fines for tomb violation or
beneficiaries of a funerary endowment with ritual duties. This is a remarkable
contrast with the situation in Saittai, where professional associations honour
people who in all probability were their deceased members. It also contrasts with
the situation in Thyateira, where associations appear to have mostly occupied
themselves with honouring members of the local and provincial elite.
On a methodological level, the present analysis has highlighted the possi-
bilities of a comparative approach that involves clearly demarcated geograph-
ical entities and political units: it has shown that professional associations were
not only a defining phenomenon in urban communities—for which hardly
any confirmation was needed—but also that specific local circumstances play a
crucial role in their formation and development. This also sheds new light on
the varying social positions of craftsmen and their associations in the civic
constellations of Greco-Roman cities. There was definitely a bias against
craftsmen by some members of the elite, but this apparently did not impede
professional associations from assuming, in some socio-political settings at
least, a prominent place within the urban community, owing to the import-
ance of their trade for the local economy and to the wealth it generated, as was
the case with the dyers in Thyateira and the purple-dyers in Hierapolis.75

ACKNOWLEDGEMENTS

I wish to thank the organizers for the opportunity to test my approach, the
participants for their criticism, and the editors of the volume, Miko Flohr and
Andrew Wilson, for their assistance in clarifying various aspects of the argument.

REFERENCES

Arnaoutoglou, I. N. (2002). ‘Roman Law and collegia in Asia Minor’, RIDA 49: 27–44.
Arnaoutoglou, I. N. (2005). ‘Collegia in the Province of Egypt in the First Century AD’,
AncSoc 35: 197–216.

75
For the role of the textile industry in Asia Minor, see Pleket (1998: 122–7) on the export of
textiles; Pleket (2003: 87–95) on the development of textile industry independently of urbaniza-
tion; Poblome (2004); Erdemir (2011), and Thonemann (2011: 186).
 Hierapolis and its Professional Associations 295
Arnaoutoglou, I. N. (2011). ‘Professional Associations in Roman Lydia: A Tale of Two
Cities?’, AncSoc 41: 257–90.
Bremen, R. van (1996). The Limits of Participation: Women and Civic Life in the Greek
East in the Hellenistic and Roman Periods. Amsterdam.
Burrell, B. (2004). Neokoroi: Greek Cities and Roman Emperors. Leiden.
Cichorius, C. (1898). ‘Geschichte und städtische Verhältnisse’, in C. Humann (ed.),
Altertümer von Hierapolis. Berlin, 18–55.
Cormack, S. (2004). The Space of Death in Roman Asia Minor. Vienna.
Cubak, N. (2008). Hierapolis tiyatro kabartmalari. Istanbul.
D’Andria, F. (2001). ‘Hierapolis of Phrygia: Its Evolution in Hellenistic and Roman Times’,
in D. Parrish (ed.), Urbanism in Western Asia Minor: New Studies on Aphrodisias,
Ephesus, Hierapolis, Pergamon, Perge and Xanthos. Portsmouth, RI, 96–115.
D’Andria, F. (2006). ‘Hierapolis of Phrygia’, in W. Radt (ed.), Stadtgrabungen und
Stadtforschung im westlichen Kleinasien: Geplantes und Erreichtes (Internationales
Symposion, 6/7 August 2004 in Bergama (Türkei)). Istanbul, 113–24.
Debord, P. (1982). Aspects sociaux et économiques de la vie religieuse dans l’Anatolie
gréco-romaine. Leiden.
Debord, P. (1997). ‘Hiérapolis: Du sanctuaire-état à la cité’, REA 99: 415–26.
Dissen, M. (2009). Römische Kollegien und deutsche Geschichtewissenschaft im 19. und
20. Jahrhundert. Stuttgart.
Dittmann-Schöne, I. (2001). Die Berufsvereine in der Städten des kaiserzeitlichen
Kleinasiens. Regensburg.
Dittmann-Schöne, I. (2002). ‘Götterverehrung bei den Berufsvereinen im kaiserzeitli-
chen Kleinasien’, in U. Egelhaaf-Gaiser and S. Schaefer (eds), Religiöse Vereine in
der römischen Antike: Untersuchungen zu Organisation, Ritual und Raumordnung.
Tübingen, 81–96.
Dmitriev, Sv. (2005). City Government in Hellenistic and Roman Asia Minor. Oxford.
Drexhage, H.-J. (1998). ‘Der Porphyropoles, die Porphyropolis und der Kogchistes in
den Papyri und Ostraka’, MBAH 17: 94–9.
Erdemir, H. (2011). ‘Woollen Textiles: An International Trade Good in the Lycus
Valley in Antiquity’, in A. H. Cadwallader and M. Trainor (eds), Colossae in Space
and Time: Linking to an Ancient City. Göttingen, 104–29.
Fraser, P. M. (1977). Rhodian Funerary Monuments. Oxford.
Guizzi, A., Miranda De Martino, E., and Ritti, T. (2012). ‘Acquisizioni epigrafiche:
Iscrizioni ritrovate o studiate nel triennio 2004–2006’, in F. d’Andria, M. P. Caggia,
and T. Ismaelli (eds), Hierapolis di Frigia V: Le attività delle campagne di scavo e
restauro 2004–2006. Istanbul, 643–78.
Harland, P. A. (2003a). Associations, Synagogues and Congregations: Claiming a Place
in Ancient Mediterranean Society. Minneapolis.
Harland, P. A. (2003b). ‘Imperial Cults within Local Life: Associations in Roman Asia’,
AHB 17: 85–107.
Harland, P. A. (2006a). ‘Acculturation and Identity in the Diaspora: A Jewish Family
and “Pagan” Guilds at Hierapolis’, Journal of Jewish Studies 57: 222–44.
Harland, P. A. (2006b). ‘The Declining Polis? Religious Rivalries in Ancient Civic
Context’, in L. E. Vaage (ed.), Religious Rivalries in the Early Roman Empire and the
Rise of Christianity. Waterloo, 21–49.
296 Ilias Arnaoutoglou
Hatzfeld, J. (1919). Les Trafiquants italiens dans l’Orient hellénique. Paris.
Huttner, U. (2009). ‘Die Färber von Laodikeia und Hierapolis: Eine Nachricht aus dem
Corpus der Alchemisten’, MBAH 26: 139–57.
Judeich, W. (1898). ‘Inschriften’, in C. Humann (ed.), Altertümer von Hierapolis. Berlin,
67–180.
Kolb, Fr. (1974). ‘Zur Geschichte der Stadt Hierapolis in Phrygien: Die Phyleninschriften
im Theater’, ZPE 15: 255–70.
Labarre, G., and Le Dinahet, M.-T. (1996). ‘Les Métiers du textile en Asie Mineure de
l’époque hellénistique à l’époque impériale’, in Aspects de l’artisanat du textile dans
le monde méditerranéen (Égypte, Grèce, monde romain). Lyon, 49–115.
Laum, B. (1914). Stiftungen in der griechischen und römischen Antike: Ein Beitrag zur
antiken Kulturgeschichte. Leipzig.
Levick, B. (2004). ‘The Roman Economy: Trade in Asia Minor and the Niche Market’,
G&R 51: 180–98.
Liu, J. (2008). ‘The Economy of Endowments: The Case of the Roman collegia’, in
K. Verboven, K. Vandorpe, and V. Chankowski-Sable (eds), ‘Pistoi dia ten technen’.
Bankers, Loans and Archives in the Ancient World: Studies in Honour of R. Bogaert.
Leuven, 231–56.
Luschan, F. von, and Petersen, E. A. H. (1889). Reisen in Lykien, Milyas und Kibyratis.
Vienna.
McLean, B. H. (2002). An Introduction to Greek Epigraphy of the Hellenistic and
Roman Periods from Alexander the Great Down to the Reign of Constantine (323
B.C.–A. D. 337). Ann Arbor.
Nijf, O. M. van (1997). The Civic World of Professional Associations in the Roman East.
Amsterdam.
Nijf, O. M. van (2010). ‘Being Termessian: Local Knowledge and Identity Politics in a
Pisidian City’, in T. Whitmarsh (ed.), Local Knowledge and Microidentities in the
Imperial Greek World. Cambridge, 163–88.
Patterson, J. R. (2006). Landscapes and Cities: Rural Settlement and Civic Transformation
in Early Imperial Italy. Oxford.
Perry, J. S. (2006). The Roman collegia: The Modern Evolution of an Ancient Concept.
Leiden.
Petrakos, V. (1999). Ho dēmos tou Ramnountos: Synopsē tōn anaskaphōn kai tōn
ereunōn (1813–1998), II. Hoi epigraphes. Athens.
Pleket, H. W. (1983). ‘Urban Elites and Business in the Greek Part of the Roman
Empire’, in P. Garnsey, K. Hopkins, and C. R. Whittaker (eds), Trade in the Ancient
Economy. Cambridge, 131–44.
Pleket, H. W. (1998). ‘Models and Inscriptions: Export of Textiles in the Roman
Empire’, Epigraphica Anatolica 12: 25–37.
Pleket, H. W. (2003). ‘Economy and Urbanization: Was there an Impact of Empire in
Asia Minor?’, in E. Schwertheim and E. Winter (eds), Stadt und Stadtentwicklung in
Kleinasien. Bonn, 85–95.
Poblome, J. (2004). ‘Comparing Ordinary Craft Production: Textile and Pottery
Production in Roman Asia Minor’, JESHO 47: 491–506.
Poland, Fr. (1909). Geschichte des griechischen Vereinswesens. Leipzig.
 Hierapolis and its Professional Associations 297
Pugliese Carratelli, G. (1963–4). ‘Chresmoi di Apollo Kareios e Apollo Klarios a
Hierapolis in Frigia’, ASAA 25–6: 357–70.
Ramsay, W. M. (1890). The Historical Geography of Asia Minor. London.
Ramsay, W. M. (1895–7). The Cities and Bishoprics of Phrygia: Being an Essay of the
Local History of Phrygia from the Earliest Times to the Turkish Conquest. Oxford.
Rebillard, E. (2003). Religion et sépulture: L’Église, les vivants et les morts dans
l’antiquité tardive. Paris.
Ritti, T. (1989–90). ‘Hierapolis di Frigia: Santuari e dediche votive’, Scienze dell’anti-
chità 3–4: 861–74.
Ritti, T. (1992–3). ‘Nuovi dati su una nota epigrafe sepolcrale con stefanotico da
Hierapolis di Frigia’, Scienze dell’antichità 6–7: 41–68.
Ritti, T. (1995). ‘Associazioni di mestiere a Hierapolis di Frigia’, in B. M. Giannattasio
(ed.), Viaggi e commerci nell'antichità (Atti VII giornata archeologica, Genova, 25
November 1994). Genova, 65–84.
Ritti, T. (1997). ‘Personaggi di età flavia a Hierapolis’, Miscellanea Greca e Romana 21:
339–57.
Ritti, T. (2003). ‘La neocoria di Hierapolis di Frigia’, in Epigraphica (Atti delle giornate
di studio di Roma e di Atene in memoria di M. Guarducci). Rome, 177–215.
Ritti, T. (2004). ‘Iura sepulcrorum a Hierapolis di Frigia nel quadro dell’epigrafia
sepolcrale microasiatica: Iscrizione edite e inedite’, in Libitina e dintorni: Libitina
e i luci sepolcrali, les leges libitinariae campane, iura sepulcrorum: Vecchie e nuove
iscrizioni (Atti dell’ XI rencontre franco-italienne sur l'épigraphie). Rome, 455–625.
Ritti, T. (2006). An Epigraphic Guide to Hierapolis (Pamukkale). Istanbul.
Ritti, T. (2007). Museo archeologico di Denizli-Hierapolis: Catalogo delle iscrizioni
greche e latine. Naples.
Ritti, T., Miranda, E., and Guizzi, F. (2007). ‘La ricerca epigrafica: Risultati dell’ultimo
quadrennio e prospettive future’, in Fr. d’ Andria and M. Piera Gaggia (eds),
Hierapolis di Frigia I: L’attivita delle campagne di scavo e restauro 2000–2003.
Istanbul, 583–618.
Robert, L. (1937). Études anatoliennes: Recherches sur les inscriptions grecques de l’Asie
Mineure. Paris.
Robert, L. (1960). Hellenica: Recueil d’épigraphie, de numismatique, et d’antiquités
grecques. xi–xii. Paris.
Robert, L. (1963). Noms indigènes dans l’Asie Mineure gréco-romaine. Paris.
Robertis, F. M. de (1970). ‘Autonomia statutaria e personificazione giuridica nel
regime associativo romano’, in Études J. Macqueron. Aix-en-Provence, 591–4.
Ruffing, K. (2004). ‘Die Selbstdarstellung von Händlern und Handwerkern in den
griechischen Inschriften’, MBAH 23: 85–101.
Ruffing, K. (2007). ‘Textilien als Wirtschaftsgut in der römischen Kaiserzeit’, in
S. Günther, K. Ruffing, and O. Stoll (eds), Pragmata: Beiträge zur Wirtschaftsgeschichte
der Antike im Gedenken an H. Winkel. Wiesbaden, 41–53.
Ruffing, K. (2008). Die berufliche Spezialisierung in Handel und Handwerk: Untersu-
chungen zu ihrer Entwicklung und zu ihren Bedingungen in der römischen Kaiserzeit
im östlichen Mittelmeerraum auf der Grundlage griechischer Inschriften und Papyri.
Rahden.
298 Ilias Arnaoutoglou
Ruffing, K. (2009). ‘Thermalquellen und Kult—das Beispiel Hierapolis’, in E. Olshausen
and V. Sauer (eds), Die Landschaft und die Religion (Stuttgarter Kolloquium zur
historischen Geographie des Altertums, 9, 2005). Stuttgart, 287–300.
Sanidas, G. M. (2011). ‘Les activités textiles dans les villes grecques aux époques hellénis-
tique et romaine: Questions d’espace et d’économie’, in C. Alfaro, J.-P. Brun, Ph.
Borgard, and R. Pierobon Benoit (eds), Tissus et teintures dans la cité antique (Actas del
III Symposium international sobre textiles y tintes del Mediterraneo en el mundo
antiguo, Naples, 13–15 November 2008). Valencia and Naples, 31–40.
Simsek, C. (2013). Laodikeia [Laodicea ad Lycum]. Istanbul.
Sommer, St (2006). Rom und die Vereinigungen im südwestlichen Kleinasien (133 v.
Chr.–284 n. Chr.). Hennef.
Sturgeon, M. C. (2004). ‘Dedications of Roman Theaters’, in A. P. Chapin (ed.), Charis:
Essays in Honor of Sara A. Immerwahr. Princeton, 411–30.
Thonemann, P. (2011). The Maeander Valley: A Historical Geography from Antiquity
to Byzantium. Cambridge.
Waltzing, J.-P. (1895–1900). Étude historique sur les corporations professionelles chez
les Romains. Louvain.
Yannakopoulos, N. (2008). O thesmos tes gerousias ton ellenikon poleon kata tous
romaikous chronous: Organosse kai leitourgies. Thessalonike.
Ziebarth, E. (1896). Das griechische Vereinswesen. Leipzig.
Ziebarth, E. (1903). ‘Beiträge zum griechischen Recht: I. Die Stiftung nach griechischen
Recht’, Zeitschrift für vergleichende Rechtswissenschaft 16: 249–315.
Zimmermann, C. (2002). Handwerkevereine im griechischen Osten des Imperium
Romanum. Darmstadt.
Zuiderhoek, A. (2011). ‘Oligarchs and Benefactors: Elite Democracy and Euergetism in
the Greek East of the Roman Empire’, in O. M. van Nijf and R. Alston (eds), Political
Culture in the Greek City after the Classical Age. Leuven, 185–95.
 IV
Space
 13

Working Together: Clusters of Artisans

in the Roman City
Penelope Goodman

In the fields of economics and geography, spatial clusters made up of similar,

interconnected firms are widely recognized. They were originally discussed by
Alfred Marshall in the nineteenth century, and the last two decades have seen
the development of an extensive literature concerned with questions such as
what constitutes a cluster, how and why they form, and what advantages they
offer.1 This literature already has a historical element, because clusters often
develop slowly and function over long periods.2 Historical enquiry has been
important, for example, in understanding industrial clusters in northern Italy,
while some researchers are now examining historically attested clusters in the
context of their own time.3 As yet, though, these enquiries have not extended
into the ancient world. Individual clusters from the Roman era have been
identified, but an overview of the phenomenon has not been attempted, nor
have the known examples been set in dialogue with the literature on their
modern equivalents.
This chapter addresses that omission. In doing so, it has two main aims:
to broaden the range of cultures in which economic clustering has been
explicitly recognized, and to extend our understanding of Roman clusters
through comparison with more recent examples. Some comment is first
needed on the legitimacy of this exercise. Is it meaningful to compare
Roman economic clusters with those of the modern world, given the very
different social, technological, and economic conditions in which they must
have evolved? In terms of the polarizing debates of the 1950s and 1960s,

1
Marshall (1890: 267–77; 1919: 283–8, 599–608). Key recent publications include Porter
(1990, 1998); Pyke, Beccatini, and Sengenberger (1990); Krugman (1991); Enright (1998);
Beccatini et al. (2003); Asheim, Cooke, and Martin (2006); Cumbers and MacKinnon (2006).
2
Krugman (1991: 59–63); van der Linde (2003: 139–40).
3
Wilson and Popp (2003b); Akoorie (2011); Lesger (2011). On industrial clusters in northern
Italy, see Capecchi (1990); Beccatini (2003).
302 Penelope Goodman

committed formalists would claim that the same basic economic factors were
active in both contexts, while substantivists would counter that ancient society
was too different from the post-industrial world for the comparison to be
useful.4 But the extreme stances of both positions have largely been left behind,
and it is now possible to acknowledge that ancient individuals could pursue
profit maximization while also recognizing that all economic activity—ancient
and modern—is socially embedded.5
Indeed, economic clustering is already recognized across a wide range of
cultures, both contemporary and historical. Contemporary examples have been
identified in neo-liberal Western economies, developing countries, and social
corporatist states such as China or Taiwan.6 Research into historical examples
has focused largely on post-industrial Europe, but Akoorie argues that clustering
was already common in medieval industry, where it was connected with the
social networks provided by the guilds.7 Each individual cluster of course
responds uniquely to its particular historical, social, and geographical context.8
But researchers who have taken a broad perspective on the phenomenon have
been able to identify comparable factors at work across a wide range of individual
cases.9 Thus, for example, while post-industrial clusters are no longer associated
with medieval-style guilds, social networks remain consistently important to
their success.10 Economic clustering appears to be a widespread characteristic
of complex human societies, akin to institutions such as households or cities
that can similarly be recognized across a range of different cultures. We should
not be too surprised if we can also identify it in the Roman Empire.
Meanwhile, research into modern economic clusters can cast valuable light
on their Roman equivalents. Clearly, we should not expect individual Roman
clusters to have developed or operated identically with those from any other
society: or indeed other Roman examples, given the cultural diversity of the
empire. But we may at least ask whether the broad-brush economic, social, and
political factors at work in better-documented modern clusters are plausible in
a Roman context. Such comparisons may also enhance our understanding of
the effects of elite control or regulation on the spatial distribution of economic
activity in Roman cities. Debate surrounding this issue has long been shaped by
a belief that the locations of some forms of economic activity were influenced
by elite concerns about literal or moral pollution.11 But these ideas are starting

4
Morley (2004: 43–5).
5
Rathbone (1991); Silver (1995: 172–5); Morley (2004: 48–50).
6
Western economies: Enright (1998); van der Linde (2003: 135–6); developing countries:
Schmitz and Nadvi (1999); van der Linde (2003: 135–6); social corporatist states: Kim (2005);
Feldman and Francis (2006: 128–9).
7 8
Akoorie (2011). Krugman (1991: 59–63); van der Linde (2003: 145–8).
9
e.g. Enright (1998); Porter (1998).
10
Beccatini (1990); Wilson and Popp (2003a: 11–15); Kim (2005).
11
Wallace-Hadrill (1995); Béal (2002); Robinson (2005); Laurence (2007: 82–101).
 Clusters of Artisans in the Roman City 303

to be questioned in the context of work that instead explores the contribution of

free-market factors in determining the same patterns.12 A systematic investi-
gation into a form of economic activity with a distinctly spatial aspect—the
cluster—offers the opportunity to contribute further to this debate.

CONTEMPORARY CLUSTERS

I will begin by reviewing the main factors behind the development and
operation of contemporary clusters, as identified in the academic literature.
This will include both industrial and retail clusters, on the grounds that these
two sectors were not sharply distinguished in the Roman world. Goods were
often made in and sold from the same premises,13 while descriptive terms for
artisans, such as unguentarius, could equally well refer to manufacturers or
retailers.14 Roman economic clusters may thus have formed in response to a
combination of the factors now identified separately for modern industrial and
retail clusters. To clarify the practicalities of cluster operation, I will use two
real examples: Birmingham’s Jewellery Quarter and London’s Bond Street.
Both are located in built-up urban environments, and occupy no more than 1
square kilometre, so that they operate on a scale that is not reliant on modern
transport. Indeed, both developed in the eighteenth century, well before this
was available. But this is not to say that clusters cannot also form on a larger
scale, or in a non-urban environment. Many such clusters do exist today, and
they are also recognized in the context of the Roman pottery industry in
particular.15 I have chosen to focus on urban examples in order to explore the
specific issue of elite control or regulation in this context.

Industrial Clusters

Industrial clusters generally develop in response to local attractors such as cheap

land, suitable raw materials, skilled workers, or consumers.16 In Birmingham
(Fig. 13.1), a major prompt for the development of jewellery-making in the
late eighteenth century was a decline in the local buckle-making industry.17
Such conditions encourage entrepreneurialism, as unemployed workers seek
new sources of income.18 In Birmingham, it was relatively easy for former

12
Ellis (2004); McGinn (2004); DeFelice (2007); Monteix (2010).
13 14
Mac Mahon (2005); Pirson (2007). See also Ruffing, Chapter 5, this volume.
15
Modern examples: van der Linde (2003: 135–6); ancient examples: Peacock (1982: 99–113);
Peña (2007: 32–3); Goodman (2013).
16 17
Porter (1998: 253–6). Mason (1998: 31–2); Carnevali (2003: 192).
18
Feldman and Francis (2006: 117–18).
304 Penelope Goodman

Fig. 13.1. Progressive clustering in the Birmingham jewellery industry between 1767
and 1913. Reproduced from Mason (1998: 10–11, fig. 5a–d).

buckle-makers to set up new businesses, since the equipment required was

modest and there were no guilds to control entry into the trades.19 Meanwhile,
the Colmore estate on the north-western edge of the town was being sold off for
property development.20 This was attractive to small-scale artisans, because
much of the land was packaged up into small plots available for a low rent,

19 20
Smith (1989: 97); Carnevali (2003: 192–3). Mason (1998: 34–6).
 Clusters of Artisans in the Roman City 305

while an absence of restrictive leases allowed workshops to be installed in attics

and backyards.21 Furthermore, the development of Birmingham’s canal and
railway networks between the 1780s and 1840s offered good transport facilities
for raw materials and finished products.22
Once an industrial cluster has taken hold, it becomes an economic attraction
in itself. This is because it allows individual firms to achieve agglomeration
economies: efficiencies arising from cooperation and collaboration, similar to
economies of scale in large single firms.23 Examples of agglomeration economies
include drawing on the same supplies of raw materials.24 Suppliers’ travel costs
are reduced by the geographical concentration of their customers, so they can
offer better deals. Similarly, firms within a cluster can interact efficiently with one
another by specializing in one stage of production, subcontracting or outsourcing
work, or sharing capital such as machinery or furnaces.25 In the Jewellery
Quarter, a range of specialized firms had developed by the mid-nineteenth
century, and items of jewellery regularly passed through several workshops
before completion.26 Subcontracting also began early on, with larger firms giving
out weekly orders to smaller firms and home workers.27
In a cluster, transaction costs are reduced, partly because firms are close
together, but also because their personnel usually know one another, and
much can be done on trust.28 Firms also draw from the same pool of labour
and expertise, meaning that workers can transfer between firms without their
skills being lost, and that information and innovations can spread quickly.29
Thus in late eighteenth- and early nineteenth-century Birmingham, the
jewellery industry helped to generate a rate of patent applications second
only to London.30 Finally, goods can be promoted collectively or passed to
salespeople on favourable trading terms, since their travel costs are again
reduced.31 Examples from the Jewellery Quarter include trade publications,
shows, and an official website.32 The importance of spatial proximity to
achieving agglomeration economies is clear from research into 102 Jewellery
Quarter firms that moved 1–2 miles out of the cluster between 1946 and
1971.33 These experienced a 77 per cent failure rate, apparently because their

21
On the small size of plots and low rent levels, see Smith (1989: 97); Mason (1998: 9). On the
absence of restrictive leases, see Smith (1989: 103); Carnevali (2003: 194).
22
De Propris and Lazzeretti (2007: 1304).
23
Krugman (1991: 36–38); Enright (1998).
24
Krugman (1991: 49–52); Enright (1998: 324–5); Porter (1998: 230–2).
25
Marshall (1919: 233–4); Lazerson (1990); Enright (1998: 326).
26
De Propris and Lazzeretti (2007: 1309–13, 1321–5).
27 28
Smith (1989: 96–9); Mason (1998: 10). Dei Ottati (2003); Carnevali (2003: 195).
29
Porter (1998: 232–3, 236–9); Krugman (1991: 38–49, 52–4).
30
De Propris and Lazzeretti (2007: 1316–17).
31
Porter (1998: 233–4); Enright (1998: 324–5).
32
Mason (1998: 127–31); Jewellery Quarter Marketing Initiative (2011).
33
Smith (1989: 101–2).
306 Penelope Goodman

reduced access to agglomeration economies was not sufficiently compensated

for by a reduction in rents.
A cluster may develop its own institutions.34 One crucial example for the
Jewellery Quarter was the Birmingham Assay Office.35 Before this was founded,
Birmingham jewellery had to be hallmarked in Chester, which incurred expense
and sometimes damage during the journey. Birmingham’s Assay Office was
founded in 1773, initially in the town centre, but it moved into the Jewellery
Quarter itself in 1877.36 In 1890, a School of Jewellery was also founded,
ensuring an ongoing supply of skilled labour.37 At the same time, the jewellers
began to form professional associations: in particular, the Birmingham Jewellers
and Silversmiths Association in 1887.38 This sought to promote the industry,
develop overseas markets, train new workers, engage with parliament, deal with
rogue traders, and detect and punish wrongdoing. A Trade Enquiry and Debt
Collection scheme also developed in 1898, with the aim of protecting its
members against fraudulent transactions and helping to collect unpaid debts.39
Bodies of this type can enhance the profitability of an industry by allowing
conscious decisions about how the cluster is governed—in particular, striking
a mutually beneficial balance between cooperation and competition.40 They
also enhance social connections within the cluster, facilitating business trans-
actions and the sharing of expertise. But far less formal social institutions can
serve the same purpose.41 In the Jewellery Quarter, family networks helped to
bind workers together, as did shared social spaces like St Paul’s Church or the
Jeweller’s Arms pub.42 Meanwhile, at any stage, the actions of powerful
individuals from within the industry can help to develop a cluster. Jewellery
Quarter examples include Matthew Boulton, a leading industrialist who
secured the opening of the Birmingham Assay Office, and Jacob Jacobs and
Charles Green, jewellery manufacturers and local councillors who drove the
establishment of the professional associations.43
The Jewellery Quarter today enjoys support from Birmingham City Council.44
Indeed, policy-makers worldwide regularly encourage the formation or
development of clusters, in the hope of fostering economic growth and social
regeneration.45 But Van der Linde, and Feldman and Francis, have found that
such initiatives are not usually effective on their own.46 Feldman and Francis

34 35 36
Porter (1998: 234–5). Mason (1998: 26–8). Mason (1998: 70–1).
37 38
Mason (1998: 82–4). Mason (1998: 79–81); Carnevali (2003: 196–206).
39 40
Carnevali (2003: 204–5). Enright (1998: 326–7); Porter (1998: 274--6).
41
Porter (1998: 241–3); Wilson and Popp (2003a: 11–15).
42
Family networks: Mason (1998: 73–5, 133–4). Shared social spaces: Haddleton (1987:
25–6); Mason (1998: 35, 54); De Propris and Lazzeretti (2007: 1313–16).
43
On Boulton, see Carnevali (2003: 193). On Jacobs and Green, see Mason (1998: 82–4);
Carnevali (2003: 196–201).
44
Smith (1989); Jewellery Quarter Marketing Initiative (2011).
45
Brusco and Pezzini (1990); Enright (1996); Porter (1998: 261–71).
46
Van der Linde (2003: 147–8); Feldman and Francis (2006).
 Clusters of Artisans in the Roman City 307

maintain that cluster success depends chiefly on entrepreneurs, who can best
determine what they need to help their businesses grow.47 Meanwhile, Duranton
argues that policy-makers’ interventions are often ineffective or actively harmful
because of a poor understanding of the problems that they are trying to correct, a
mistaken belief that larger clusters are always more efficient, or a preference for
pleasing key supporters over achieving real economic growth.48 Indeed, the
Jewellery Quarter has seen its own aborted regeneration initiative. After decline
caused by the Second World War, the Birmingham Jewellers and Silversmiths
Association, together with the city council, proposed the regeneration of the
quarter by replacing its small Victorian workshops with large factory blocks.49 In
the event, only one was built, and firms proved reluctant to move into it.50 The
quarter’s ongoing success relates more to the development of direct retail to the
public, the updating of the School of Jewellery, and its promotion as a tourist
attraction.51

Retail Clusters

Bond Street in central London has been an up-market shopping street since the
eighteenth century.52 This is no surprise, since it is easily accessible from a
number of wealthy residential areas, and close to other attractions such as
parks, theatres, and the political centres of Westminster and Whitehall. During
the late nineteenth and early twentieth centuries, the street was increasingly
populated by art dealers and antique shops, and also became the home of the
Fine Art Society and Sotheby’s the auctioneers.53 The smaller shops acquired
goods and expertise from these institutions while retailing to the general public,
until many of them were forced to close in the recession of the late 1980s.54
A small cluster persists around Sotheby’s and the Fine Art Society, but Bond
Street is now dominated by the designer fashion industry (Fig. 13.2).55
The fashion outlets of Bond Street lack a clear industry hub such as Sotheby’s
and the Fine Art Society provided for the art and antique shops. But they can
still benefit from clustering together. This is because customers who buy high-
value items like designer clothing (or art and antiques) generally prefer to shop
around and compare the goods on offer before making their purchases. For this
reason, such items are often described as ‘shopping goods’ or ‘comparison goods’:
and research into this form of retail shows that customers will spend more if they

47
Feldman and Francis (2006: 129–31). See also Enright (1996) and Porter (1998: 240–1).
48 49
Duranton (2011). Smith (1989: 104–5); Mason (1998: 138–40).
50 51
Mason (1998: 150–4). Smith (1989); Mason (1998: 156–7, 160–70).
52
Glennie (1998: 937); Rappaport (2000: 10).
53
Scott (1970: 29); Hermann (1980: 135–7, 153–7).
54
Hermann (1980: 298); Wrigley and Lowe (2002: 192).
55
Wrigley and Lowe (2002: 192–6).
308 Penelope Goodman
FASHION
0 50 100m
WATCHES AND JEWELLERY
OX
ART AND ANTIQUES FO
RD
STR
EE
T

BLENHEIM STREET

DERING STREET

NEW BOND STREET

BRO
OK S
TRE
ET

BRO
OK S
TRE
ET

LANCASHIRE
COURT
GRO
SVE
N OR S
TRE
ET

EET
STR
DOX
MAD

NEW BOND STREET

SOTHEBY’S

E ET
STR
THE FINE ART SOCIETY DUIT
CON

BRUTON STREET

CLIFFORD STREET

GRAFTON STREET
NEW BOND STREET

BURLINGTON GARDENS

ROYAL ARCADE

STAFFORD STREET

PICCADILLY PICCAD
ILLY

Fig. 13.2. Retail clustering on Bond Street, London. Drawn by Alex Santos.

can make their comparisons easily, without incurring extensive travel costs.56
Thus, although the individual shops on Bond Street compete with one another
for custom, each still attracts more customers than it would alone—another form
of agglomeration economy. Clustering allows their customers to make compari-
sons easily, while competition occurs on price and quality, rather than location.57
This forms a sharp contrast with low-value ‘convenience goods’—that is, daily
purchases such as bread and milk.58 Shops selling these goods compete on
location rather than price, since customers are generally unwilling to travel

56
Scott (1970: 26–7); Jones and Simmons (1990: 119).
57
Jones and Simmons (1990: 41).
58
Scott (1970: 96); Jones and Simmons (1990: 38–45).
 Clusters of Artisans in the Roman City 309

P1

P2

P3

P4 P7

P6
P5

Fig. 13.3. The distribution of bakeries in Amsterdam, 1742. Reproduced from Lesger
(2011: 32, fig. 2).

long distances in order to save a few pence on a low-value purchase. As a result,

they are usually fairly evenly distributed across a city. In eighteenth-century
Amsterdam, Lesger has observed clear differences in the distribution of conveni-
ence and comparison retailers.59 Contemporary tax registers show that bakeries
were distributed evenly through most parts of the city, while retailers of high-
value goods such as books and fine fabrics were located in the city centre:
sometimes in distinct clusters (Fig. 13.3). Lesger emphasized that, although
eighteenth-century Amsterdam possessed extensive civic regulations, these
included no restrictions on the distribution of shops.60 Thus, the retail patterns
observed were shaped primarily by market forces rather than legislation.
Once a retail cluster has developed, it too becomes self-replicating. Bond Street
now has an international reputation as a centre of fashion, making it essential
for any firm wishing to position itself as a designer retailer to open a shop there.61
Its businesses also promote themselves collectively, like those in the Jewellery
Quarter—for example, through an official website.62 Retail clusters may also
develop into social hubs. This is certainly true for Bond Street, which has become
a place to ‘be seen’ for fashionable celebrities and professionals, complete with bars
and restaurants where they can congregate. These constitute another attractor to
the area, and enhance social connections between both customers and workers.

59 60 61
Lesger (2011). Lesger (2011: 44–6). Fernie et al. (1997).
62
Bond Street Association (2011).
310 Penelope Goodman

CLUSTERING IN ROMAN URBANISM

We can now look at Roman economic clusters in the light of their modern
equivalents. The focus here is again on urban examples, so that the potential
impact of elite regulation or control in this context can be explored. Competition
for space is also usually more intensive in the centre of a city than on its periphery
or in the countryside. This means that economic clusters in urban centres should
be the result of stronger imperatives towards cluster formation, because compe-
tition from other activities is also stronger.

Clusters Attested in the Archaeological Record

Clear examples of urban economic clustering are preserved at Pompeii in Italy,

Timgad in North Africa, and Silchester in Britain. These sites are unusual in
two respects. First, occupation at all three ceased during or at the end of
antiquity, meaning that their Roman-period remains were not obscured by
later buildings. Secondly, they have undergone large-scale clearance excava-
tions that uncovered all or most of the built-up urban area. This is important
for the identification of clustering, since we need to be able to compare the
distribution of workshops across a city in order to determine whether a
particular type occurs with above-average frequency in one area. But rapid
clearance excavations are often achieved at the expense of collecting proper
dating evidence and recording the work in detail. This can make it difficult to
be certain whether all of the workshops within an apparent cluster were in use
simultaneously. Clearance excavations are also much better at identifying
workshops with fixed installations than those used by artisans working with
lighter equipment: for example, shoe-makers, basket-weavers, or indeed jew-
ellers. Such industries are therefore likely to be under-represented in the
archaeological record, while the clusters that we can see represent industries
that required significant capital investment for each new workshop. Once
formed, such clusters must have been relatively inert, but other industries
may have clustered and dispersed more flexibly.

Pompeii

In the 1960s, Walter Moeller identified workshop types in Pompeii that he

believed were involved in textile production.63 Most occurred throughout the
city, but one group formed a distinct cluster, particularly in regio VII, insulae

63
Moeller (1966, 1976).
 Clusters of Artisans in the Roman City 311

Fig. 13.4. Officinae lanifricariae at Pompeii. Map: Miko Flohr.

Fig. 13.5. An example of an officina lanifricaria, Pompeii VII 2, 17. Photo: Miko Flohr.

9–12, just east of the city’s forum (Fig. 13.4). These workshops contained long
masonry tables, small vats, and lead-lined pans with furnaces underneath
(Fig. 13.5). On the basis of a graffito on the street front close to one reading
lanifricari dormis (‘wool-scrubber, you sleep’), Moeller dubbed them officinae
lanifricariae.64 He suggested that the tables were used for beating the dirt out
of raw wool fleeces, the vats for soaking the wool, and the pans for heating and
scouring it. Moeller argued that the lanifricarii must have been organized into

64
CIL 4.1190; Moeller (1966; 1976: 12–13).
312 Penelope Goodman

a collegium, but that there was no evidence for such a group because it
operated as part of the collegium of the fullers.65
Moeller’s interpretation of these workshops was heavily criticized by Willem
Jongman.66 Jongman questioned the relevance of the lanifricari dormis graffito,
which actually comes from the wall of an inn two doors away from one of the
workshops, and suggested that the establishments may instead have served hot
food.67 Certainly, large quantities of bones were discovered in at least some of the
workshops, indicating some form of animal processing.68 He further pointed out
that Moeller’s identification of I 4, 26 as an officina lanifricaria, based on an 1875
excavation report, was not supported by any surviving remains.69 But even
Jongman agreed that the remainder of workshops formed an ‘archaeological
type’—that is, they all shared the same distinct characteristics.70 This is enough
to identify the group as an economic cluster, since the workshops clearly all
performed the same function. Meanwhile, Pompeii’s destruction by Vesuvius
means that these workshops represent the last usage of the plots that they occupy
before AD 79. This does not prove that they were all active at that time, but it does
mean that they all existed simultaneously, and strongly suggests that most were
active during the mid-first century.
Without being sure what went on in these workshops, it is difficult to say
why they were clustered in this area of Pompeii. Moeller argued that the
people operating them bought their raw materials from the nearby macellum
or Eumachia building.71 This is perfectly likely, even if the materials were not
wool fleeces, while either building could equally have been used to sell their
finished products. Both had back entrances leading directly into the quarter,
and this would have allowed access even after the eastern edge of the forum
had been sealed off by a line of monumental public buildings.72 But this
monumentalization project itself must have affected the streets east of the
forum. Dobbins dates its final phase to the post-62 earthquake restoration
period, but adds that the eastern streets may already have been blocked in the
Augustan era.73 After this, pedestrians could reach the macellum and Euma-
chia building from the east, but steps or narrow passages would have barred
animals and carts, while of course the buildings’ back entrances may not
always have been open. The district thus became one of the least accessible
in the city. Kaiser has shown that, whether the streets here were approached

65 66 67
Moeller (1976: 74–5). Jongman (1991). Jongman (1991: 166–9).
68
Flohr (2013: 59) argues for animal bone finds in at least two, perhaps three, of these
workshops. Monteix (2013: 81–4) believes that contemporary excavation journals confirm the
existence of only one.
69
Jongman (1991: 168). The same could be said about VII 3, 24 and VII 4, 39–40. Cf. Moeller
(1976: 31–2). However, even if all three are discounted, the cluster in insulae VII 9–12 remains
unaffected.
70 71
Jongman (1991: 167). Cf. Monteix (2013). Moeller (1976: 68–71).
72 73
Dobbins (1994). Dobbins (1994: 689–91).
 Clusters of Artisans in the Roman City 313

from the city gates or the forum, an above-average number of junctions

needed to be negotiated by comparison with other Pompeian streets in
order to reach them.74
Various activities considered immoral or illegal by the elite, including gam-
bling and prostitution, were also concentrated in this area.75 Wallace-Hadrill and
Laurence see this as a deliberate choice on the part of the urban elite, though they
place different emphasis on the question of whether it was achieved through
direct regulation or other methods such as social coercion and property owner-
ship.76 But McGinn has questioned whether this aspect of the quarter’s character
was really so sharply different from the rest of Pompeii.77 In any case, the reasons
behind it are difficult to assess without knowing when the streets leading east
from the forum were blocked, since some uses of space here may be relics of an
earlier period before this happened. We can say only that, once it did happen, it
should have had an economic impact. Ellis, Laurence, and Kaiser have all shown
that plots with entrances on accessible main streets were favoured for most other
private land use in Pompeii: commerce, other forms of production, and elite
housing.78 Once blocked, the back streets east of the forum would have become
unattractive for these activities. Rents in the area may have fallen, and have
represented a particular bargain for those whose activities did not depend on
access to busy streets and passing traffic. This might plausibly include workshops
that traded via the macellum or Eumachia buildings, but did not sell directly to
the general public.
The distinction between convenience goods and comparison goods may
cast further light on the operation of these workshops, and particularly Jong-
man’s suggestion that they produced hot food. A relatively low-value purchase
such as a meal for immediate consumption is a convenience good rather than
a comparison good, making customers unlikely to travel long distances to buy
it. We should expect to find hot food shops spread relatively evenly through-
out the city, and indeed do see this for other convenience goods in Pompeii—
for example, bars that are known to have sold hot food and bakeries with
attached shops.79 In this context, Jongman’s theory appears weak. But the
focus of the cluster in a quarter with low accessibility would make comparison
shopping equally unlikely. Its position matches better with what we would
expect for an industrial cluster with little emphasis on retail. This is supported
by Flohr’s investigation of these workshops, which found that they were

74 75
Kaiser (2011: 120). Wallace-Hadrill (1995: 51–5); Laurence (2007: 87–8, 95–6).
76
Wallace-Hadrill (1995); Laurence (2007: 83, 99).
77
McGinn (2004: 78–84, 240–1).
78
Ellis (2004: 378–80); Laurence (2007: 116); Kaiser (2011: 125–9).
79
Ellis (2004); Monteix (2010)—though Monteix notes that bakeries with shops gravitate
towards the centre of Pompeii. This is itself explicable as an attempt to capture the greatest
market share. Cf. Jones and Simmons (1990: 44).
314 Penelope Goodman

around half as likely to include a recognizable shop front as the bakeries,

fulleries, dyeries, or tanneries of Pompeii.80
Many of the workshops were also neighbours, occupying units directly
adjacent to or opposite one another.81 Much as in Birmingham’s Jewellery
Quarter, their occupants could hardly have avoided knowing one another, and
this would have put them in a good position to cooperate and achieve agglom-
eration economies if they had chosen to do so. Indeed, while Moeller may have
been overstretching the evidence when he identified them as lanifricarii operat-
ing under the oversight of a fullers’ guild, the extensive evidence for trade
associations in Pompeii means that it would be surprising if these artisans had
no association of their own. A social structure of this type would have brought
them together, offering opportunities to strike formal or informal business
agreements. There is also considerable variation in the size of the installations,
from the single-room workshop at VII 9, 41 to the complex at VII 12, 22–3,
which was over ten times the size.82 Flohr found that, out of fourteen identifiable
officinae lanifricariae, almost half had four rooms or fewer, while only three had
fifteen or more.83 This suggests an industry consisting mainly of small firms, but
with a few larger enterprises. If so, this may have encouraged outsourcing from
larger to smaller firms of the kind attested in the Jewellery Quarter: though, of
course, there is no direct evidence for this.

Timgad

Working from late-nineteenth- and early twentieth-century excavation re-

ports, Andrew Wilson identified at least twenty-two workshops in Timgad
equipped with similar rectangular vats, circular tubs, and wells (Fig. 13.6).84
Wilson believes that these workshops were probably fulleries, with the circular
tubs used for treading cloth, but he adds they could also have been used
for cold-water dyeing.85 The poor quality of the excavation reports meant
Wilson was unable to locate all the workshops precisely within the urban plan.
However, eighteen workshops could be traced to one or more rooms within an
insula block, while the rest could be located to the level of the block as a whole.
This is quite enough to be sure that the workshops formed a distinct cluster in
the north-east corner of the city (Fig. 13.7), where seventeen of the twenty-
two were located.86 Moreover, Wilson suggests that the sheer number of
workshops points towards more than local service fulling. He envisions the

80
Flohr (2007: 133–4).
81
Adjacent: VII 11, 2–3 and 4–5; VII 12, 22–3 and VII 12, 24. Opposite: VII 11, 2–3 and VII
10, 13.
82 83
Moeller (1966: 496; 1976: 31–5). Flohr (2007: 132–3).
84 85
Wilson (2000; 2002: 237–41). Wilson (2000: 273–5).
86
Wilson (2000: 278).
 Clusters of Artisans in the Roman City 315

I J

Unpaved
A street

G H
B

W
F
C
D E
N

0 1 2 3 4 5m

Key;

Structural walls
Stone slabs or blocks
Concrete surface

A Terracotta fulling tub; internal diameter 0.68 m; thickness 0.06 m; depth 0.68 m.
B Terracotta fulling tub; internal diameter 0.80 m; thickness 0.07 m; depth 0.54 m.
C Terracotta fulling tub; internal diameter 0.77 m; thickness 0.07 m; depth 0.62 m.
D Terracotta fulling tub; internal diameter 0.65 m; thickness 0.07 m; depth 0.48 m.
E Terracotta fulling tub; internal diameter 0.50 m; thickness 0.05 m; depth 0.44 m.
F Terracotta fulling tub; internal diameter 0.82 m; thickness 0.07 m; depth 0.60 m.
G Sandstone fulling tub; internal diameter 0.78 m; thickness 0.10 m; depth 0.50 m.
H Terracotta fulling tub; internal diameter 0.85 m; thickness 0.07 m; depth 0.57 m.
I Terracotta fulling tub; internal diameter 0.84 m; thickness 0.06 m; depth 0.65 m.
J Terracotta fulling tub; internal diameter 0.84 m; thickness 0.06 m; depth 0.60 m.
W Well

Fig. 13.6. An example of a probable fullers’ workshop at Timgad. Reproduced from

Wilson (2000: 274, fig. 12.01).

large-scale finishing of cloth, probably linked with the presence of at least two
textile markets in the city.87
No dating evidence is available for these workshops. They simply represent
the last recognizable structures on the plots that they occupied before the
city was abandoned in late antiquity. Nonetheless, the similar characteristics of
the workshops suggest that they were set up around the same time, following
a common template. Indeed, this may represent a simple version of the
dissemination of innovations widely recognized in modern industrial clusters.
Wilson further points out that the circular tubs were generally found un-
broken, making it unlikely that some were dismantled or allowed to fall into
disrepair while others remained in operation.88 As at Pompeii, the individual
workshops vary considerably in scale. Some contain up to ten circular tubs

87 88
Wilson (2000: 281–7). Wilson (2000: 280).
316 Penelope Goodman

Modern buildings N
Rooms in which fulling or dyeing can be precisely located
Areas in which fulling or dyeing occurred, but cannot be located to
individual rooms from the available descriptions

North-east
Baths

Possible
cloth
market

Late
kilns

1060

Bronze foundry
10 8 0

Pottery workshop

Fig. 13.7. Probable fulleries at Timgad. Reproduced from Wilson (2000: 279, fig. 12.08).

and occupy a quarter of an insula, while others have only one or two within a
single room.89 Again, the right conditions for outsourcing are in place, but
there is no direct evidence for it.

89
Wilson (2000: 274–8).
 Clusters of Artisans in the Roman City 317

Between the uncertainty over the function of these workshops and an

absence of written evidence to illuminate their economic or social context, it
is difficult to be certain why they are clustered in one area of the city. But the
cluster’s relationship with its surroundings offers some grounds for specula-
tion. It is noticeable that no workshops are located along the main roads
leading towards the forum at the centre of the city. In fact, all are situated at
least three blocks away from these routes. Two possible scenarios could
explain this. One is that the landowning elite in Timgad used formal legisla-
tion or informal agreements to keep workshops away from the main roads.
The other is that the workshops were simply priced out of this area by other
competing forms of land use. As at Pompeii, this could include shops or
workshops that sold goods directly to passing customers, or wealthy house-
holders who wished to live at prominent locations within the urban network.
It is certainly quite likely that land values, and thus also rental prices, were
comparatively low in the north-eastern quarter, especially away from the main
thoroughfares. By the end of the second century AD, the south-western part of
Timgad had become a prime focus of development. The west gate had been
rebuilt as a monumental arch with a temple and market outside it; houses,
including two atrium-peristyle domus, had been constructed along a demolished
stretch of the city wall; and the city’s Capitolium had been built at the south-
western corner of the city.90 Elite expenditure was clearly focused on this area,
and this is likely to have been associated with higher land values. By contrast, the
north-eastern quarter lacked a piped water supply, and its housing occurs in
small units with little architectural decoration.91 It seems to have been compara-
tively unattractive to developers or tenants, suggesting that rents were lower, and
thus perhaps favouring the use of the area for artisanal activity. If so, this would
bear some similarity with the development of the Jewellery Quarter on the north-
western edge of Birmingham.
The workshops also gravitate around two infrastructural features: a small
north-eastern gateway and the northern ‘intra-pomerial’ road running
along the interior of the city’s walled circuit. This circuit in itself seems to
have become obsolete in the latter half of the second century. Parts of it were
demolished, and new arches were built 200–300 metres out along the major
approaches to the city, probably reflecting an extension of the urban bound-
aries.92 But much of its course was preserved by structures built against it or
over its remains. In the north-eastern quarter, this meant that the gateway and
road remained important channels of movement. It is thus probably signifi-
cant that the north-eastern gateway opens directly onto an insula more than
half taken up by four workshops, while the majority of the rest lie within a

90
West gate quarter: Lezine (1966–7); Tourrenc (1968). City wall development: Lassus
(1966). Capitolium: Ballu (1903: 193–206).
91
Lohmann (1979: 179); Wilson (2000: 286–7).
92
Demolition: Lassus (1966). New east and west arches: Ballu (1903: 111; 1911: 10–13).
318 Penelope Goodman

radius of 100 metres. Similarly, at least seven workshops line the intra-
pomerial road, including three that are otherwise displaced in relation to the
main cluster. It is tempting to suggest that the gateway and road were
important to the activity of these workshops, perhaps for transporting their
raw materials or finished products. It is also worth noting that the north-
eastern gateway was one of only two purpose-built openings cut through
Timgad’s walled circuit after its construction.93 Without proper dating evi-
dence, we cannot be certain, but it is possible that this modification related
directly to the needs of the adjacent cluster.

Silchester

Finally, at least thirty-one distinctive circular furnaces were identified by the

Victorian excavators of Silchester (Fig. 13.8).94 These were built from terra-
cotta tiles, although no more than three courses of any are preserved. All were
around 2.5 to 3 feet wide, with a short flue about a foot wide and evidence of

Fig. 13.8. Circular furnace discovered in Insula XI, Block III, Silchester. Reproduced
from Fox (1895: fig. 2).

93
Ballu (1903: 11–13).
94
Main discussion: Fox (1895). Other references: Fox (1893: 560); Fox and St John Hope
(1896: 217 and 246); St John Hope (1897: 411); St John Hope and Fox (1898: 104; 1899: 239;
1900: 93); Fox and St John Hope (1901: 237 and 240–1); St John Hope (1902: 18, 29; 1905: 335–7;
1906: 154).
 Clusters of Artisans in the Roman City 319

heat damage on the furnace floors and the ground in front. Most were
apparently built outside, but some were inside buildings, and others could
have stood in wooden structures missed by the excavators.95 The furnaces do
not resemble pottery kilns, and no diagnostic waste products were found with
them. Fox suggested that they were dyeing boilers on the basis of comparisons
with similar furnaces at Pompeii, but others have since suggested beer-making
or baking.96 No dating evidence was collected from the furnaces, so we cannot
say when they were active or how they relate to one another chronologically.
But their similar dimensions and materials suggest that most were built and
used around the same time.
The Silchester furnaces occur throughout the city, including close to the
forum and the basilica. But they are particularly prominent in the north-
western quadrant of the city. Fourteen occur in insulae X, XI, XIII, and XIV
near to the west gate, while at least four are in insula XXIII, mainly in its
north-western corner (Fig. 13.9).97 Again, it is difficult to say why this might

N
0 100 200 300 metres

XXV XXIV

XXVI XXIII XXIIa

XXIIb
AMPHITHEATRE

XI X IX I XXI XXVII XXXVI

XXXVII

XIII XIV II V XXVIII XXIX

IV
XXX
XV XVI III VI XXXIV XXXI

XX XIX XVII VII XXXV XXXIII XXXII

XVIIIb XVIIIa VIII

Fig. 13.9. Circular furnaces at Silchester. The plan shows all furnaces mentioned in
Fox and St John Hope’s excavation reports that can be securely associated with a
precise location on the city plan. Drawn by Alex Santos.

95
Clarke and Fulford (2002). Fox (1895: 463) acknowledges this possibility.
96
Fox (1895: 459–67); Boon (1974: 286–9); Wacher (1995: 287).
97
Fox and St John Hope (1901: 237) actually reports remains of at least eight furnaces in
insula XXIII, but only four are shown on their plans. It appears that only fairly well-preserved
remains were drawn onto the excavation plans, so that there were probably more furnaces in the
city than can now be located accurately.
320 Penelope Goodman

be. But fifteen of the furnaces are within 50 metres of the main routes into the
city from the west or north gates. This is rather different from the workshops
at Pompeii and Timgad, which were situated away from busy main roads.
Here, we must assume that the local elite were perfectly happy for furnaces to
cluster along these routes, since their political and economic power would
have enabled them to displace the industry if not. The pattern may also suggest
that direct retail was more important for this industry than for the workshops
previously discussed. Certainly, the rectangular strip-buildings along the
route in from the west gate suggest that this was a busy thoroughfare with
plenty of passing trade, since these represent a way of maximizing access to an
economically attractive street frontage.98 Meanwhile, proximity to the gates
would facilitate the delivery of raw materials, as well as the collection of any
finished products not sold directly from the workshops.

Clusters Attested in Texts

Written evidence also hints at the presence of clusters in some cities where
they are not attested archaeologically. Rome is particularly well documented in
this respect, and yields a number of toponyms likely to reflect artisanal
clustering (Table 13.1).99

Table 13.1. Toponyms from Rome that suggest artisanal clustering

Toponym Product or service Reference

clivus argentarius Banking LTUR 1.280

clivus capsarius Bathing supplies? LTUR 1.281
scalae anulariae Rings LTUR 4.238–9
sigillaria Gifts LTUR 4, 310
vicus caprarius Goats LTUR 5.156
vicus cornicularius Horn LTUR 5.160
vicus frumentarius Corn LTUR 5.166
vicus lorarius Harnesses LTUR 5.175
vicus materiarius Carpentry/timber? CIL 6.975
vicus mundiciei Luxury objects/toilet articles? LTUR 5.181
vicus sandaliarius Shoes LTUR 5.189
vicus t(h)urarius Incense LTUR 5.194–6
vicus unguentarius Perfume LTUR 5.197–8
vicus vitrarius Glass LTUR 5.200

98
Perring (2002: 55–60).
99
References are given where possible to entries in the LTUR (Steinby 1993–2000), which
provide full details of the primary sources. See also MacMullen (1974: 130), Morel (1987), and
Holleran (2012: 51–60).
 Clusters of Artisans in the Roman City 321

Though toponyms alone are not conclusive evidence for the presence of
clusters, an analogy used by Augustine shows that the two could go together:
We laugh, indeed, when we see them [the gods] distributed by the fiction of
human imagination into their separate roles, like those who farm small portions
of the public revenue, or like workmen in the vicus argentarius, where one vessel,
in order that it may go out perfect, is passed between many craftsmen . . .100
Augustine seems to be thinking of a vicus argentarius in Carthage, rather than
Rome,101 but he wrote for a cosmopolitan audience, suggesting that the scenario
would be widely recognized. He also reveals that the artisans in this particular
vicus102 specialized and collaborated, much like the workers in Birmingham’s
Jewellery Quarter. Back in Rome, references to the vicus sandaliarius also show
that clusters might form and disperse while toponyms remained unchanged.
Though the name suggests a cluster of shoe-makers, Aulus Gellius and Galen in
the second century describe it as lined with booksellers.103 Rome also shows that
clusters could exist without generating toponyms. In late republican and early
imperial literature, the Via Sacra appears as a focus for jewellers, and the vicus
tuscus for fine clothes traders, while work by Monteix on the workplaces recorded
on artisans’ tombstones reflects these same concentrations.104 Some texts also use
the formulae in . . . or inter. . . to locate activities or places ‘among the’ practitioners
of a particular trade.105 Arguably, these are more persuasive evidence for economic
clustering than straightforward toponyms, since they show that at the time of
writing these districts really were recognizably dominated by a particular trade.
Not all of these examples can be located within Rome, but several of those
that can occur around the forum (Fig. 13.10). It is striking that these relate to
high-value items such as jewellery, books, or luxury fabrics: in other words,
comparison goods. They thus match up well with eighteenth-century Amster-
dam and modern Bond Street, where similar clusters were also located close to
the accessible hearts of the cities. Indeed, in a world without telephone
directories or the Internet, the impetus towards retail clustering may have
been even stronger, since it would help customers to find the goods they
wanted. Meanwhile, Gellius’ and Galen’s comments about the vicus sandaliar-
ius recall Bond Street’s role as a social hub. Both report encountering intel-
lectual disputes there: one about the meaning of certain words in Sallust, and

100 101
August., De civ. D. 7.4. Rougé (1966: 323–4); Wilson (2002: 259).
102
The etymologists define an urban vicus as a ‘district’ or ‘quarter’ (Varro, Ling. 5.145, 160;
Isid., Etym. 15.2.22), but the word was often also applied to its main street (e.g. Livy 1.48.6–7).
103
Gell., NA 18.4.1; Gal., Libr. Propr. Kühn 19.8.
104
Morel (1987: 145); Holleran (2012: 55–7); Monteix (2012).
105
in figlinis (potteries: LTUR 2.252–3), inter aerarios (bronze-workers: CIL 6.9186), inter
falcarios (scythe-makers: Cic. in Cat. 1.8 and pro Sulla 52), inter figulos (potteries: LTUR 2.253),
inter lignarios (carpenters: Livy 35.41.10), inter olivarios (olive-sellers: CIL 6.37043 (= 1².809)),
inter vitores (basket-makers: LTUR 5.207).
322 Penelope Goodman

0 50 100m

CLIV
US
ARG
NTAE
R IU
S
VIC
US
SA
NDA
LIAR
CAPITOLINE IUS
?

HILL

SA
CR
A
VIA
SCALAE ANULARIAE?
US
SC
TU
US
VIC

PALATINE
HILL

Fig. 13.10. Streets characterized by retail clusters in the Forum area, Rome. Drawn by
Alex Santos.

the other about the authorship of a work attributed to Galen.106 The vicus thus
seems to have functioned as a place to discuss as well as buy books. Rome also
boasted numerous atria, basilicae, fora, horrea, macella, and portici named
after particular goods or trades.107 These are purpose-built structures, and as
such represent the outcomes of conscious decisions to devote space to a
particular economic activity. Nonetheless, they would have allowed artisans
or traders to achieve agglomeration economies or attract comparison shoppers
in the same way as clusters of many individual shops or workshops.
Elsewhere, papyri, inscriptions, and literary references suggest the presence
of similar clusters in streets, districts, porticoes, or marketplaces across the
empire. Stöckle collected twenty-three such examples from Egyptian papyri,
now supplemented by Krüger for Oxyrhynchus, while MacMullen identified

106
Gell., NA 18.4; Gal. Libr. Propr. Kühn 19.8–9.
107
MacMullen (1974: 132); Morel (1987: 137–40, 148–54).
 Clusters of Artisans in the Roman City 323

another fourteen in Italy, Greece, Asia Minor, and Judaea.108 This evidence is
clearly affected by the embeddedness of the epigraphic habit and the survival
of papyrus, and thus bypasses the north-western provinces. Most examples are
also only brief mentions, which do not prove the existence of economic clusters.
But some use the formula ἐν τοῖς . . . , the Greek equivalent of in . . . or inter . . . ,
which does point towards genuine concentrations of artisans.109 The textual
corpus also covers a wider range of activities than the archaeological examples,
including trade in perishable goods or service provision, which would be all but
undetectable through excavation.110 It also ranges over a considerable period,
from Cicero’s scythe-makers in the late first century BC to Augustine’s vicus
argentarius in the fifth century AD.

Broadening the Picture

The examples presented above confirm that economic clustering was a real
phenomenon in the Roman world. It is probably under-represented in the
surviving evidence, since we can expect to see it only in places that have
undergone large-scale clearance excavations or left an unusually rich textual
record. Few sites meet these criteria, but where they do, clusters can be
detected: albeit in limited detail. The known cases range widely across the
empire and through time. They also vary in character, involving many differ-
ent trades and industries, and ranging from back-street industrial clusters, as
at Pompeii and Timgad, to those geared more towards direct retail, as in
Rome. The local circumstances that gave rise to them must have varied, as
must the details of how they operated. But such variation is already allowed for
within the literature on contemporary clustering, and does not prevent the
ancient examples from being recognized as part of the same phenomenon.
Meanwhile, by going beyond the urban examples introduced above, we
can speculate further about the possible mechanisms under which they oper-
ated. Non-urban clusters provide more concrete evidence for economic co-
operation, and this is especially true of the terra sigillata industry, thanks to
the information provided by name-stamps and close dating techniques.
Here, decorative schemes and stamp designs tended to evolve quickly, espe-
cially during periods of peak production (for example, between 15 BC and AD
30 at Arezzo)—a phenomenon that is comparable with the rapid spread of

108
Stöckle (1911: 149–52); MacMullen (1974: 133); Krüger (1990: 82–8). See also Droß-
Krüpe, Chapter 14, this volume.
109
Arsinoë: BGU 1087 5, 7.8 (bakers) and 4, 13 (garland-makers). Thebaïs: P. Grenf. 1, 21
(potters) and P. Grenf. 1, 42 (gooseherds).
110
e.g. wool-traders in Jerusalem (Joseph. BJ 5.8.1.331) or barbers in Velitrae, Italy (CIL
15.7172).
324 Penelope Goodman

innovations in modern clusters.111 The shared use of fixed capital such as kilns
is also demonstrated by surviving firing lists, while Dannell has argued for the
use of outsourcing at la Graufesenque.112 Specialization is also attested in terra
sigillata production centres, with some individuals, for example, concentrating
on the production of moulds that were then used in more than one work-
shop.113 Such conditions did not apply in all sigillata production centres, but
they do demonstrate that these methods of working, all of which are familiar
from modern clusters, could also develop in a Roman context. In fact, Ruffing
argues that specialization in particular may have been more common across all
types of Roman industry than is usually recognized, and regularly occurs in
conjunction with spatial clustering.114
Roman social structures could also have provided support for economic
clusters, as is attested in other periods and cultures. Trade associations existed
throughout the empire, in spite of restrictions sometimes imposed by the
emperors.115 Interpretations of their character have ranged from religiously
oriented social clubs to organized economic interest groups: and of course we
should not assume a universal model.116 But, even if their functions were
primarily social, such groups could support spatial clustering by enhancing
personal connections within the trade. Indeed, Hawkins argues that they
lowered transaction costs between small firms by creating networks of trust,
thus reducing the incentives to form larger, but less flexible, integrated firms,
and leaving the very clusters of small workshops under investigation here.117
Although none of the archaeologically attested clusters discussed is known to
have had a related trade association, a link does emerge in the context of
toponyms. In Lydia, the Greek word plateia (broad street) was used to describe
some trade associations, strongly suggesting that they formed a spatial cluster
as well as a social network.118 Family networks could similarly strengthen
social connections, and plenty of trades were certainly family-dominated. For
example, pottery stamps and inscriptions reveal that the production of mor-
taria in a group of kilns at Aoste in Gaul was led by the local Atisii for over a
hundred years.119 Apprenticeships also provided access to professional social
networks, and could have encouraged spatial concentration—as suggested by
the case of an Egyptian boy apprenticed to a coppersmith in the same street as
his father.120 Again, there is no direct evidence for the influence of family

111 112 113

Kenrick (2004: 256). Polak (1998); Dannell (2002). Mees (1994).
114
Ruffing (2008; Chapter 5, this volume).
115 116
Cotter (1996); van Nijf (1997); Aubert (1999); Diosono (2007). Perry (2006).
117
Hawkins (2012: 186–93).
118
TAM 5.1 79–81, 146; Arnaoutoglou, Chapter 12, this volume.
119
Rémy and Jospin (1998).
120
PSI 871 in Johnson (1936: 391–2). On apprenticeship in general, see Freu, Chapter 8, this
volume.
 Clusters of Artisans in the Roman City 325

networks or apprenticeship schemes in the urban clusters discussed, but such

institutions were part of their wider social context.

E L I T E RE G U L A T I O N AN D C O N TR O L

We can now return to the relationship between the spatial distribution of

Roman economic activity and conscious control or regulation by the urban
elite. Local elites were certainly in a good position to control the economic
topography of Roman cities. The wealth inequalities of the ancient world
would lead us to expect that most shops and workshops were owned by the
wealthy elite and staffed by their dependants or rented to artisans (directly or
through business managers), and this is supported by the evidence of rental
notices and property development patterns in Pompeii.121 Where workshops
included fixed equipment, Egyptian papyrus contracts show that it was typi-
cally provided and maintained by the owner, and some Italian evidence points
towards the same practices.122 The urban elite could thus determine the
locations and functions of shops or workshops, either via direct legislation
or as property-owners.
The case for legislation is weak. Wallace-Hadrill argued that the locations of
brothels and gambling dens in Pompeii were restricted by rulings enforced by
the aediles.123 But he could present no direct evidence for legal restrictions on
the spatial distribution, as opposed to the conduct, of economic activity, while
his concept of moral zoning has since been criticized.124 The closest fit among
the ancient evidence is the Urso charter, which forbids the construction of
‘tile-kilns with a capacity of more than three hundred tiles or tile-like objects
in the oppidum [i.e. urban centre] of the colony’.125 But, if this restriction was
breached, the kiln was simply forfeited to the civic authorities—not demol-
ished or put out of action. Large tile kilns could exist in the centre of Urso,
then, so long as they were not privately owned, while smaller kilns and other
workshops were apparently unrestricted.126 This leaves us with no credible
evidence for the use of civic legislation to control the economic topography of
Roman cities.
As large-scale property-owners or benefactors, the elite clearly did exercise
control over urban space. But this does not mean that they used it consistently to

121
Pirson (1997, 1999); Robinson (2005).
122
For Egypt, see, e.g. BGU 1067 (mill), BGU 1117 (bakeshop), BGU 1127 (goldsmith’s),
P. Teb. 342 (pottery). See further Cockle (1981). For Italy, see CIL 10.2226 (wool workshops),
CIL 4.1136 (baths), CIL 4.3340.142 (fullery).
123 124
Wallace-Hadrill (1995). McGinn (2004); DeFelice (2007).
125
Lex Coloniae Genetivae Iuliae 76.24–8; Crawford (1996: 393–454).
126
Goodman (2007: 106–8).
326 Penelope Goodman

prioritize moral and aesthetic values over the maximization of profits, as some
have implied. Dio Chrysostom’s speeches about the construction of a colonnade
in Prusa, Bithynia, which entailed the displacement of some workshops, are a
case in point.127 Béal sees these texts as revealing an elite belief that artisanal
activity compromised the dignity of a city, and should not be the basis of its
wealth.128 But Dio’s plans were opposed by other members of the local elite, who
specifically defended the workshops.129 No doubt the argument was a proxy for
wider rivalries, but the case does show that elite consent to the removal of
workshops from urban centres was not automatic. Meanwhile, Dio built new
workshops in two other locations.130 This would have given him control over
how the workshops were used: perhaps rather like the third-century benefactor,
Marcus Nikephoros, who awarded places between the columns of a portico along
the ‘Marble Street’ in Ephesus.131 These spaces were shared between at least
seven different trade associations, but, where a single individual controlled a
comparable block of commercial space, it is perfectly plausible that they might
offer units first to one particular trade group, especially if they served as its
patron. Indeed, Nikephoros donated at least four places to the taurinadoi
(probably shoemakers)—just over 20 per cent of the eighteen surviving alloca-
tions.132 This may not be representative of their original share, but within a
colonnade around 180 metres long, even four taurinadoi would seem like
something of a cluster.
Private property-ownership might have a comparable effect. The Insula
Arriana Polliana and the Praedia Julia Felix in Pompeii are examples of
concentrated property holdings: complete blocks owned by single individuals,
within which smaller units were available for rent.133 Something similar may
have existed in the south-western corner of Timgad, where a strip of land was
developed for housing after the demolition of the city wall.134 Lassus pointed
out that the decision not to build any new roadways through this development
is characteristic of a profit-driven private entrepreneur—probably the Sertius
who owned a large domus at the southern end of the development and
financed a public market alongside its northern end. Property-owners in this
position could rent on the open market like the Pompeian landlords, who
would presumably let to whomever offered the highest price. But again they
might plausibly privilege a particular group of artisans by offering places to
them privately and/or on favourable terms, thus helping to create a cluster.
Such behaviour would have to be socially motivated, since it would reduce the
owner’s rental income, but enhanced status and influence as a patron might

127 128
Dio Chrys., Or. 40.6–10, 45.12–16, 47. Béal (2002: 5–7).
129
Dio Chrys., Or. 40.8–9, 47.11; Bekker-Nielsen (2008: 125–7, 130–1).
130 131
Dio Chrys., Or. 40.9, 46.9. Knibbe (1985); van Nijf (1997: 83–5).
132
IK 16.2080, 2081.
133
Insula Arriana Polliana: CIL 4.138; Pirson (1997). Praedia Julia Felix: CIL 4.1136.
134
Lassus (1966); Wilson (2002: 261).
 Clusters of Artisans in the Roman City 327

compensate for this.135 Equally, those who controlled continuous blocks of

property could have chosen not to rent them out to the practitioners of a given
trade, thus pushing them out of one area of the city and encouraging them to
cluster in another. Again, the motive would have to be social, and here moral and
aesthetic preferences are the only available explanation. Thus, for example,
Sertius could have encouraged the concentration of fullers’ workshops in the
north-eastern quarter of Timgad because he preferred not to rent out suitable
units in the property that he controlled in the south-west.
But work on contemporary clustering suggests that the consistent applica-
tion of such moral or aesthetic preferences is more likely to have impeded the
creation of clusters. Economists and geographers emphasize the action of the free
market, and particularly individual entrepreneurs, in their accounts of modern
cluster development.136 If Roman property-owners consistently prioritized
moral considerations when equipping and renting shops and workshops, this
should have distorted whatever profit motivations were being expressed by
artisans or traders. In modern terms, it would dampen the input of entrepre-
neurs, acting as a barrier to the formation and development of clusters. Yet
clusters did form. This should suggest that some property-owners were willing to
listen to the views of artisans concerning the most profitable uses of particular
economic units. In doing so, the property-owners might be motivated by a
combination of ‘pure’ economic and social factors: a desire to maximize both
their profits as landlords and their social standing as patrons. Indeed, the social
role of the elite as trade patrons may actively have encouraged clustering, though
it is unlikely that this was a conscious aim, as it is for modern policymakers. The
formation of economic clusters in Roman cities, then, can be explained with
reference to both social and economic motivations on the part of the urban elite.
The two are not always easily separable, and it needs remembering that social
factors are important in modern clustering too—for example, through the
influence of social networks. But the findings of research into modern clustering
weigh against explaining the distribution of Roman urban workshops in terms of
the moral preferences of the urban elite. Where economic clustering can be
identified, it is more likely that this particular social factor was not at work.

ACKNOWLEDGEMENTS

I would like to thank Andrew Wilson and Miko Flohr for inviting me to
participate in the ESF workshop, my fellow participants for their responses,
Andy Evans and Dan Olner for comments on a written draft, and Charlotte
Trémouilhe for proof-reading.

135
Pirson (2007: 470).
136
Enright (1996); van der Linde (2003: 147–8); Feldman and Francis (2006).
328 Penelope Goodman

REFERENCES

Akoorie, M. (2011). ‘A Challenge to Marshallian Orthodoxy on Industrial Clustering’,

Journal of Management History 17/4: 451–70.
Asheim, B., Cooke, P., and Martin, R. (2006). Clusters and Regional Development:
Critical Reflections and Explanations. London and New York.
Aubert, J. (1999). ‘La Gestion des collegia: Aspects juridiques, économiques et sociaux’,
CCG 10: 49–69.
Ballu, A. (1903). Les ruines de Timgad (antique Thamugadi): Nouvelles découvertes. Paris.
Ballu, A. (1911). Les ruines de Timgad (antique Thamugadi): Sept années de découvertes
1903–1910. Paris.
Béal, J.-C. (2002). ‘L’artisanat et la ville: Relecture de quelques textes’, in J.-C. Béal and
J.-C. Goyon (eds), Les artisans dans la ville antique. Lyons and Paris, 5–14.
Beccatini, G. (1990). ‘The Marshallian Industrial District as a Socio-Economic Notion’, in
F. Pyke, G. Beccatini, and W. Sengenberger (eds), Industrial Districts and Inter-Firm
Cooperation in Italy. Geneva, 37–51.
Beccatini, G. (2003). ‘Industrial Districts in the Development of Tuscany’, in Beccatini
et al. (2003), 11–28.
Beccatini, G., Bellandi, M., Dei Ottati, G., and Sforzi, F. (2003). From Industrial
Districts to Local Development: An Itinerary of Research. Cheltenham and North-
ampton, MA.
Bekker-Nielsen, T. (2008). Urban Life and Local Politics in Roman Bithynia: The Small
World of Dion Chrysostomos. Aarhus.
Bond Street Association (2011). ‘Bond Street Luxury Shopping London’ <http://www.
bondstreet.co.uk/> (accessed 15 October 2011).
Boon, G. C. (1974). Silchester: The Roman town of Calleva. London.
Brusco, S., and Pezzini, M. (1990). ‘Small-Scale Enterprise in the Ideology of the Italian
Left’, in Pyke, Beccatini, and Sengenberger, (1990), 142–59.
Capecchi, C. (1990). ‘A History of Flexible Specialisation and Industrial Districts in
Emilia-Romagna’, in Pyke, Beccatini, and Sengenberger (1990), 20–36.
Carnevali, F. (2003). ‘“Malefactors and Honourable Men”: The Making of Commercial
Honesty in Nineteenth-Century Birmingham’, in Wilson and Popp (2003b), 192–207.
Clarke, A., and Fulford, M. (2002). ‘The Excavation of Insula IX, Silchester: The First
Five Years of the “Town Life” Project, 1997–2001’, Britannia 33: 129–66.
Cockle, H. (1981). ‘Pottery Manufacture in Roman Egypt: A New Papyrus’, JRS 71: 87–97.
Cotter, W. (1996). ‘The collegia and Roman Law: State Restrictions on Voluntary
Associations, 64 BCE–200 CE’, in J. S. Kloppenborg and S. G. Wilson (eds), Voluntary
Associations in the Graeco-Roman World. London and New York, 74–89.
Crawford, M. H. (1996). Roman Statutes, i. London.
Cumbers, A., and MacKinnon, D. (2006). Clusters in Urban and Regional Development.
London and New York.
Dannell, G. B. (2002). ‘Law and Practice: Further Thoughts on the Organization of the
Potteries at la Graufesenque’, in M. Genin and A. Vernhet (eds), Céramiques de la
Graufesenque et autres productions d’époque romaine: Nouvelles recherches. Hommages
à Bettina Hoffmann. Montagnac, 211–42.
 Clusters of Artisans in the Roman City 329
Dei Ottati, G. (2003). ‘Trust, Interlinking Transactions and Credit in the Industrial
District’, in Beccatini et al. (2003), 108–30.
DeFelice, J. (2007). ‘Inns and Taverns’, in J. J. Dobbins and P. W. Foss (eds), The World
of Pompeii. London and New York, 474–86.
De Propris, L., and Lazzeretti, L. (2007). ‘The Birmingham Jewellery Quarter:
A Marshallian Industrial District’, in European Planning Studies, 15/10: 1295–1325.
Diosono, F. (2007). Collegia: Le associazioni professionali nel mondo romano. Rome.
Dobbins, J. (1994). ‘Problems of Chronology, Decoration, and Urban Design in the
Forum at Pompeii’, AJA 98/4: 629–94.
Duranton, G. (2011). ‘California Dreamin’: The Feeble Case for Cluster Policies’,
Review of Economic Analysis 3: 3–45.
Ellis, S. (2004). ‘The Distribution of Bars at Pompeii: Archaeological, Spatial and
Viewshed Analyses’, JRA 17: 371–84.
Enright M. J. (1996). ‘Regional Clusters and Economic Development: A Research
Agenda’, in U. Staber, N. V. Schaefer, and B. Sharma (eds), Business Networks:
Prospects for Regional Development. New York, 190–214.
Enright, M. J. (1998). ‘Regional Clusters and Firm Strategy’, in A.D. Chandler,
P. Hagström, and Ö. Sölvell (eds), The Dynamic Firm: The Role of Technology, Strategy,
Organization and Regions. Oxford, 315–42.
Feldman, M., and Francis, J. L. (2006). ‘Entrepreneurs as Agents in the Formation of
Industrial Clusters’, in B. Asheim, P. Cooke, and R. Martin (eds), Clusters and
Regional Development: Critical Reflections and Explanations. London and New York,
115–136.
Fernie, J., Moore, C., Lawrie, A., and Hallsworth, A. (1997). ‘The Internationalization
of the High Fashion Brand: The Case of Central London’, Journal of Product and
Brand Management 6/3: 151–62.
Flohr, M. (2007). ‘Nec quicquam ingenuum habere potest officina? Spatial Contexts of
Urban Production at Pompeii, AD 79’, BABesch 82: 129–48.
Flohr, M. (2013). ‘The Textile Economy of Pompeii’, JRA 26/1: 53-78.
Fox, G. E. (1895). ‘Excavations on the Site of the Roman City at Silchester, Hants, in
1894’, Archaeologia 54: 439–72.
Fox, G. E., and St John Hope, W. H. (1893). ‘Excavations on the Site of the Roman City
at Silchester, Hants, in 1892’, Archaeologia 53: 539–73.
Fox, G. E., and St John Hope, W. H. (1901). ‘Excavations on the Site of the Roman City
at Silchester, Hants, in 1900’, Archaeologia 57: 229–56.
Glennie, P. (1998). ‘Consumption, Consumerism and Urban Form: Historical Perspec-
tives’, Urban Studies 35: 927–51.
Goodman, P. J. (2007). The Roman City and its Periphery: From Rome to Gaul. London
and New York.
Goodman, P. J. (2013). ‘The Production Centres: Settlement Hierarchies and Spatial
Distribution’, in M. Fulford and E. Durham (eds), Seeing Red: New Economic and
Social Perspectives on terra sigillata. London, 121–36.
Haddleton, M. E. (1987). The Jewellery Quarter: History and Guide. Birmingham.
Hawkins, C. (2012). ‘Manufacturing’, in W. Scheidel (ed.), The Cambridge Companion
to the Roman Economy. Cambridge, 175–94.
Hermann, F. (1980). Sotheby’s: Portrait of an Auction House. London.
330 Penelope Goodman
Heichelheim, F. (1938). ‘Roman Syria’, in T. Frank (ed.), An Economic Survey of
Ancient Rome, iv. Baltimore, 121–257.
Holleran, C. (2012). Shopping in Ancient Rome: The Retail Trade in the Late Republic
and the Principate. Oxford.
Jewellery Quarter Marketing Initiative (2011). ‘Official Website of Birmingham’s
Jewellery Quarter’ <http://www.jewelleryquarter.net/> (accessed 15 October 2011).
Johnson, A. C. (1936). Roman Egypt to the Reign of Diocletian (= T. Frank (ed.), An
Economic Survey of Ancient Rome, ii). Baltimore.
Jongman, W. (1991). The Economy and Society of Pompeii. Amsterdam.
Jones, K., and Simmons, J. (1990). The Retail Environment. London.
Kaiser, A. (2011). ‘What was a via? An Integrated Archaeological and Textual Approach’,
in E. Poehler, M. Flohr, and K. Cole (eds), Pompeii: Art, Industry, and Infrastructure.
Oxford, 115–30.
Kenrick, P. (2004). ‘Signatures on Italian sigillata: A New Perspective’, in J. Poblome,
P. Talloen, R. Brulet, and M. Waelkens (eds), Early Italian Sigillata: The Chrono-
logical Framework and Trade Patterns. Leuven, 253–62.
Kim, J. Y. (2005). ‘The Formation of Clustering of Direct Foreign Investment and its
Role of Inter-Firm Networks in China: Case Study of Qingdao Development Zones’,
in C. Karlsson, B. Johansson, and R. Stough (eds) Industrial Clusters and Inter-Firm
Networks. Cheltenham and Northampton, MA.
Knibbe, D. (1985). ‘Der Asiarch M. Fulvius Publicianus Nikephoros, die ephesischen
Handwerkszünfte und die Stoa des Servilius’, JÖAI 56: 71–7.
Krüger, J. (1990). Oxyrhynchos in der Kaiserzeit: Studien zur Topographie und Litera-
turrezeption. Frankfurt.
Krugman, P. (1991). Geography and Trade. Belgium and Cambridge, MA.
Lassus, J. (1966). ‘Une opération immobilière à Timgad’, in R. Chevallier (ed.),
Mélanges d’archéologie et d’histoire offerts à André Piganiol. Paris, 1221–31.
Laurence, R. (2007). Roman Pompeii: Space and Society. London and New York.
Lazerson, M. H. (1990). ‘Subcontracting in the Modena Knitwear Industry’, in Pyke,
Beccatini, and Sengenberger (1990), 108–33.
Lesger, C. (2011). ‘Patterns of Retail Location and Urban Form in Amsterdam in the
Mid-Eighteenth Century’, Urban History 38/1: 24–47.
Lezine, A. (1966–7). ‘Note sur l’arc dit de Trajan à Timgad’, Bulletin d'Archéologie
Algérienne 2: 123–7.
Linde, C. van der (2003). ‘The Demography of Clusters: Findings from the Cluster
Meta-Study’, in J. Bröcker, D. Dohse, and R. Soltwedel (eds), Innovation Clusters
and Interregional Competition. Berlin, 130–49.
Lohmann, H. (1979). ‘Beobachtungen zum Stadtplan von Timgad’, in Diskussionen
zur Archäologischen Bauforschung, 3. Berlin, 167–87.
Mac Mahon, A. (2005). ‘The Shops and Workshops of Roman Britain’, in A. Mac
Mahon and J. Price (eds), Roman Working Lives and Urban Living. Oxford, 48–69.
McCallum, M. (2011). ‘Pottery Production in Pompeii: An Overview’, in E. Poehler,
M. Flohr, and K. Cole (eds), Pompeii: Art, Industry, and Infrastructure. Oxford,
103–14.
McGinn, T. (2004). The Economy of Prostitution in the Roman World: A Study of
Social History and the Brothel. Ann Arbor.
 Clusters of Artisans in the Roman City 331
MacMullen, R. (1974). Roman Social Relations, 50 BC to AD 284. New Haven and
London.
Marshall, A. (1890). Principles of Economics. London.
Marshall, A. (1919). Industry and Trade: A Study of Industrial Technique and Business
Organization, and of their Influences on the Conditions of Various Classes and
Nations. London.
Mason, S. (1998). Jewellery Making in Birmingham 1750–1995. Chichester.
Mees, A. (1994). ‘Potiers et moulistes: Observations sur la chronologie, les structures et
la commercialisation des ateliers de terre sigillée décorée’, in SFECAG: Actes du
Congrès de Millau. Marseilles, 19–41.
Moeller, W. O. (1966). ‘The lanifricarius and the officinae lanifricariae at Pompeii’,
Technology and Culture 7: 493–96.
Moeller, W. O. (1976). The Wool Trade of Ancient Pompeii. Leiden.
Monteix, N. (2010). ‘La localisation des métiers dans l’espace urbain: quelques
exemples pompéiens’, in P. Chardron-Picault (ed.), Aspects de l’artisanat en milieu
urbain: Gaule et Occident romain. Dijon, 147–60.
Monteix, N. (2012). ‘ “Caius Lucretius . . . marchand de couleurs de la rue du fabricant
de courroies”: Réflexions critiques sur les concentrations de métiers à Rome’, in
G. Sanidas and A. Esposito (eds), Quartiers artisanaux en Grèce ancienne: Une
perspective méditerranéenne. Villeneuve d’Ascq, 333–52.
Monteix, N. (2013). ‘The Apple of Discord: Fleece-Washing in Pompeii’s Textile
Economy’, JRA 26/1: 79–87.
Morel, J.-P. (1987). ‘La topographie de l’artisanat et du commerce dans la Rome
antique’, in École française de Rome (ed.), L’Urbs: Espace urbain et histoire.
Rome, 127–55.
Morley, N. (2004). Theories, Models and Concepts in Ancient History. London and
New York.
Nijf, O. van (1997). The Civic World of Professional Associations in the Roman East.
Amsterdam.
Peacock, D. P. S. (1982). Pottery in the Roman World: An Ethnoarchaeological
Approach. London.
Peña, J. T. (2007). Roman Pottery in the Archaeological Record. Cambridge and New
York.
Perring, D. (2002). The Roman House in Britain. London.
Perry, J. (2006). The Roman Collegia: The Modern Evolution of an Ancient Concept.
Leiden.
Pirson, F. (1997). ‘Rented Accommodation at Pompeii: The Insula Arriana Polliana’,
in R. Lawrence and A. Wallace-Hadrill (eds), Domestic Space in the Roman World:
Pompeii and Beyond. Portsmouth, RI, 165–81.
Pirson, F. (1999). Mietwohnungen in Pompeji und Herculaneum. Munich.
Pirson, F. (2007). ‘Shops and Industries’, in J. J. Dobbins and P. W. Foss (eds), The
World of Pompeii. London and New York, 457–73.
Polak, M. (1998). ‘Old Wine in New Bottles: Reflections on the Organization of the
Production of terra sigillata at la Graufesenque’, in J. Bird (ed.), Form and Fabric:
Studies in Rome’s Material Past in Honour of B. R. Hartley. Oxford, 115–21.
Porter, M. (1990). The Competitive Advantage of Nations. Basingstoke.
332 Penelope Goodman
Porter, M. (1998). ‘Clusters and Competition’, in M. Porter (ed.), On Competition.
Boston, 213–303.
Pyke, F., Beccatini, G., and Sengenberger, W. (1990). Industrial Districts and Inter-Firm
Cooperation in Italy. Geneva.
Rappaport, E. (2000). Shopping for Pleasure: Women in the Making of London’s West
End. Princeton and Oxford.
Rathbone, D. (1991). Economic Rationalism and Rural Society in Third-Century AD
Egypt: The Heroninos Archive and the Appianus Estate. Cambridge.
Rémy, B., and Jospin, J.-P. (1998). ‘Recherches sur la société d’une agglomération de la
cité de Vienne: Aoste (Isère)’, Revue archéologique de Narbonnaise 31: 73–89.
Robinson, D. (2005). ‘Re-Thinking the Social Organisation of Trade and Industry in
First Century AD Pompeii’, in A. Mac Mahon and J. Price (eds), Roman Working
Lives and Urban Living. Oxford, 88–105.
Rougé, J. (1966). Expositio totius mundi et gentium: Introduction, texte critique,
traduction, notes et commentaire. Paris.
Ruffing, K. (2008). Die berufliche Spezialisierung in Handel und Handwerk: Untersu-
chungen zu ihrer Entwicklung und zu ihren Bedingungen in der römischen Kaiserzeit
im östlichen Mittelmeerraum auf der Grundlage griechischer Inschriften und Papyri.
Rahden.
Schmitz, H., and Nadvi, K. (1999). ‘Clustering and Industrialization: Introduction’,
World Development 27/9: 1503–14.
Scott, P. (1970). Geography and Retailing. London.
Silver, M. (1995). Economic Structures of Antiquity. London.
Smith, B. (1989). ‘The Birmingham Jewellery Quarter: A Civic Problem that has
Become an Opportunity in the 1980s’, in B. Tilson (ed.), Made in Birmingham:
Design and Industry 1889–1989. Studley, 96–112.
St John Hope, W. H. (1897). ‘Excavations on the Site of the Roman City at Silchester,
Hants, in 1896’, Archaeologia 55: 409–30.
St John Hope, W. H. (1902). ‘Excavations on the Site of the Roman City at Silchester,
Hants, in 1901’, Archaeologia 58: 17–36.
St John Hope, W. H. (1905). ‘Excavations on the Site of the Roman City at Silchester,
Hants in 1903 and 1904’, Archaeologia 59: 333–70.
St John Hope, W. H. (1906). ‘Excavations on the Site of the Roman City at Silchester,
Hants, in 1905’, Archaeologia 60: 149–68.
St John Hope, W. H., and Fox, G. E. (1896). ‘Excavations on the Site of the Roman City
at Silchester, Hants, in 1895’, Archaeologia 55: 215–56.
St John Hope, W. H., and Fox, G. E. (1898). ‘Excavations on the Site of the Roman City at
Silchester, Hants, in 1897’, Archaeologia 56: 103–26.
St John Hope, W. H., and Fox, G. E. (1899). ‘Excavations on the Site of the Roman City
at Silchester, Hants, in 1898’, Archaeologia 56: 229–50.
St John Hope, W. H., and Fox, G. E. (1900). ‘Excavations on the Site of the Roman City at
Silchester, Hants, in 1899’, Archaeologia 57: 87–112.
Steinby, E. M. (1993–2000). Lexicon Topographicum Urbis Romae. Rome.
Stöckle, A. (1911). Spätrömische und byzantinische Zünfte. Leipzig.
Tourrenc, S. (1968). ‘La Dedicace du Temple du Genie de la colonie à Timgad’, AntAfr
2: 197–220.
 Clusters of Artisans in the Roman City 333
Wacher, J. (1995). The Towns of Roman Britain. London.
Wallace-Hadrill, A. (1995). ‘Public Honour and Private Shame: The Urban Texture of
Pompeii’, in T. J. Cornell and K. Lomas (eds), Urban Society in Roman Italy. London,
39–62.
Wilson, A. (2000). ‘Timgad and Textile Production’, in D. J. Mattingly and J. Salmon
(eds), Economies beyond Agriculture in the Classical World. London, 271–96.
Wilson, A. (2002). ‘Urban Production in the Roman World: The View from North
Africa’, PBSR 70: 231–73.
Wilson, J. F., and Popp, A. (2003a). ‘Districts, Networks and Clusters in England: An
Introduction’, in Wilson and Popp (2003b), 1–18.
Wilson, J. F. and Popp, A. (eds) (2003b). Industrial Clusters and Regional Business
Networks in England, 1750–1970. Aldershot.
Wrigley, N., and Lowe, M. (2002). Reading Retail: A Geographical Perspective on
Retailing and Consumption Spaces. London.
 14

Spatial Concentration and Dispersal

of Roman Textile Crafts
Kerstin Droß-Krüpe

Certain toponyms of ancient Rome such as vicus materiarius, vicus frumentarius,

or scalae anulariae suggest the presence of distinct quarters of urban craftsmen.1
These toponyms—or perhaps better hodonyms and prodonyms—raise the
question of whether they point to the location of specialized traders’ and workers’
quarters in ancient towns.2 A close look at Egyptian documentary papyri enables
us to investigate the distribution of crafts within Roman cities. Using textile
crafts—dyeing and fulling in particular—as a case study, this chapter will discuss
the spatial distribution of textile workshops within settlements in the province of
Egypt in the Roman imperial period. First, a close look will be taken at the
sources that shed light on the location of these crafts. As a second step, modern
clustering theory will be used to explain the presence or absence of textile craft
quarters.

URBAN TOPOGRAPHIES OF TEXTILE CRAFTS

Three large trade tax registers dating back to AD 276 survive from Ptolemais
Euergetis, the capital of the Arsinoite nome in Egypt.3 These accounts of
various trade taxes paid by individuals, grouped by occupations, allow decent
insight not only into the amount of tax paid but also into the geographical
distribution of certain crafts within the city: for the overwhelming majority of
the craftsmen listed, the tax records specify the town quarter in which they

1
Vicus materiarius and vicus frumentarius: CIL 6.975 (= ILS 6073). Scalae anulariae: Suet.
Aug. 72.1. Cf. Goodman, Chapter 13, this volume.
2
As stated, e.g. by Kolb (1995: 496). On hodonyms and prodonyms, see Koß (1995: 458–63).
3
BGU I.9, BGU IV.1087, and BGU XII.2280.
 Roman Textile Crafts 335

lived. In interpreting these documents, one has among the other peculiarities
of these registers to bear in mind their fragmentary character. This means,
among other things, that the data preserved indicate a minimum number of
persons exercising the stated profession: they cannot be used to reconstruct
the actual number of people performing a certain craft. Nevertheless, assum-
ing that workspace and living space overlapped, they reveal a vivid impression
of the topography of crafts in Ptolemais Euergetis.
The tax registers show an extensive number of craftsmen of different occu-
pations. There are γρυτοπῶλαι (presumably traders of junk or trinkets),
μυροπῶλαι (traders of ointments), βάφοι (dyers), στιβεῖς (fullers), ἀρτυματᾶτες
(traders of spices), ζυτοπῶλαι (traders of beer), ἀρτοκόποι (bread-makers/
bakers), κορσασᾶτες (barbers), and κασσιτερᾶτες (pewterers). In total, the
three documents list 121 different people practising at least 11 different occu-
pations. The trade tax these people had to pay varied according to what they did.
Dyers, for example, had to pay 24 drachmas a month, whereas traders of junk
were supposed to pay 8 or 12 drachmas; traders of ointment even had to pay 60
drachmas.
The dyers form one of the largest professional groups in each of the three
fragments: up to twelve are attested; their distribution over the registers is shown
in Table 14.1. The number of dyers, and the amount of tax they pay, suggest that
they were an important group in the local economy. Some people appear in more
than one tax register, while each fragment also gives names of dyers not
mentioned in the other two. Taking a closer look at the dyers, a profession
appearing in all three tax registers of that very year, shows a surprisingly fixed
number of dyers within the nome capital. Only in one case (BGU I.9) is the list of
dyers complete, consisting of twelve people, several of whom appear in one or
both of the other tax registers. In BGU XII.2280 seven dyers are mentioned by
name. The total recorded amount of payments by this group of craftsmen
indicates that an additional four monthly payments had to be made, thus
suggesting that the number of members of the dyeing craft was eleven. The
third register (BGU IV.1087) lists seven dyers, including one name not appearing
in the other lists. Only one dyer, Εὐδέμων, appears on all three lists. All in all,
within the short time span of a few months, fifteen different dyers are identifiable.
Together, these three lists from the same year provide important information
about the society of this nome capital and its craftsmen, and they allow us to
assess the spatial position of the textile craftsmen. As we know the town quarter
of almost all craftsmen, we can discuss how they were distributed across the
town: the fifteen dyers mentioned lived in at least thirteen different town
quarters. As Table 14.2 shows, the same is true for the other professions listed.
Craftsmen performing the same craft were spread over the whole town
(Tables 14.2, 14.3). This is all the more surprising for the craft of dyeing, since
the process of dyeing with its chemical procedures and the related material were
bound to produce considerable smells: such factors have often led scholars to
336 Kerstin Droß-Krüpe
Table 14.1. Dyers in Ptolemais Euergetis according to papyrological evidence
Name of dyer BGU I.9 BGU IV.1087 BGU XIII.2280

Διοκωρος ουπου ̣πα παρὰ Ἀλύπιν x — x

Παῦλος ἐν τῇ Μύρι x — x
Εὐδέμων ἐν τῇ Συριακῇ x x x
Σαβῖνος ἐν τῷ Καπίτωνος x — x
Κόννυμος ἐν τῷ Νυμ[ ̣]σ̣ίου x — —
Σαραπιὰς ἐν τῇ Παληᾷ Παραπωλης x x —
Μέλας ἐν τῷ Σεβήρου x — —
Ἡρακεδης ἐν τῇ Ἀθηνᾷ x x —
Κουτᾶς ἐν τῇ Μύρι x — x
Μωρίων ἐν τῷ Καπίτωνος x x —
Κύριλος ἐν τῷ Λαγίῳ x x —
Εὐπωρίων ἐν τῷ Φρέμι x x —
Εὐπωρίων — x —
Μέλας ἐν τῷ Παλαστίῳ — — x
Κούκωμος ἐν τῷ Γυμασίῳ — — x

Table 14.2. Spatial distribution of craftsmen in Ptolemais Euergetis

Profession Craftsmen Town quarters

γρυτοπῶλαι 3 3
μυροπῶλαι 4 4
βάφοι 15 13 or 14
στιβεῖς 7 7
ἀρτυματᾶτες 4 3 or 4
ζυτοπῶλαι 7 5 to 7
ἀρτοκόποι 9 7 to 9
κορσασᾶτες 4 2 to 4
κασσιτερᾶτες 4 4

assume that those kinds of workshops were found on the peripheries of

settlements. The situation with fullers is similar: the seven fullers listed in
BGU IV.1087 all lived in different parts of the town, although their use of
urine and other chemical substances was bound to produce unpleasant odours.4
Analysing also the other professions listed in these three tax registers shows
that spatial concentration in the sense of explicitly allocated craft quarters
cannot be detected at Ptolemais Euergetis. Dyers and fullers are not an
exception, but follow the usual local pattern of craftsmen in Ptolemais Eu-
ergetis, even though their work—unlike other crafts—produced ‘unsavoury
fragrances’ for the neighbours.
In total, ninety-six craftsmen from forty-five town quarters can be identi-
fied. Most of the town quarters host one (22) or two craftsmen (10); the

4
A different view is presented by Flohr (2003: 448); cf. Droß-Krüpe (2011: 45 with n. 117)
and Flohr and Wilson (2011).
 Roman Textile Crafts 337
Table 14.3. Spatial distribution of dyers in Ptolemais Euergetis
Town quarter/geographical specification Number of dyers

παρὰ Ἀλύπιν 1
ἐν τῇ Μύρι 2
ἐν τῇ Συριακῇ 1
ἐν τῷ Καπίτωνος 1
ἐν τῷ Νυμ[ ̣]σ ̣ίου 1
ἐν τῇ Παληᾷ Παραπωλης 1
ἐν τῷ Σεβήρου 1
ἐν τῇ Ἀθηνᾷ 1
ἐν τῷ Καπίτωνος 1
ἐν τῷ Λαγίῳ 1
ἐν τῷ Φρέμι 1
ἐν τῷ Παλαστίῳ 1
ἐν τῷ Γυμασίῳ 1
Quarter lost 1

25

20
Number of quarters

15

10

0
1 2 3 4 5 6 7 8 9
Number of craftsmen

Fig. 14.1. Number of town quarters in Ptolemais Euergetis hosting 1 to 9 craftsmen.

Image: Kerstin Droß-Krüpe.

highest number of craftsmen living in the same part of the town is nine
(Fig. 14.1). Thus, specific working quarters cannot be detected, and it seems
that craftsmen did not live separated from townspeople who had different
occupations or belonged to different social strata. The question arises whether
this situation was particular to Ptolemais Euergetis, or whether dyers and
338 Kerstin Droß-Krüpe

fullers lived and worked right among their fellow craftsmen in other towns
and villages as well. From the late second century AD we learn from a
papyrus—probably from Antinoopolis—that Aurelia Claudia Leontarion (alias
Amonilla) leased properties for commercial purposes to different craftsmen.5
These properties were located in two different streets: the ἐργαστήια ἐπὶ τῆς
πλατεία ς̣ ̣ comprised not only of a ὀθονιοπώλιον and an χρυσ ̣ ̣οχοε ̣ῖ ̣ον, but also
of several ἱματιοπώλης, a κιστοπλόκιορ, a ὑδραγώγιον, and an ἐνδικοπλύτιον.
Those ἐπὶ τῆς ῥύμης ἐργαστηρίων hosted a παλαιοράφιον and a πλοκόπιον.
Again, very different workshops could be found in close proximity to each
other.
A remarkable document from Oxyrhynchos dating to the late third
century provides an insight into workshops in another nome capital. P.Oxy.
XLVI.3300, written in AD 271–2, depicts the ownership of properties that had
to be registered with the town council. Listed are the names of owners and
occupiers of properties, and sometimes their occupations are also added.
It shows a random pattern of very diverse professions working and living
side-by-side, as shown in Fig. 14.2. The dyer Aniketos lives and works directly
next to Aphynchios, who is an embroiderer, and near the fisherman Pasois,
and Harpokr[—], a builder. Obviously this city did not require a spatial
grouping of individual craftspeople.
Panopolis, another nome capital, provides a similar but much longer and much
more detailed document: a real estate register, first published in 1975 and—after
undergoing substantial corrections—later presented as SB XXIV.16000.6 This is
an exceptional document that sheds light on the occupational and economic
structure of this town, which played a significant economic role in the Panopolites
nome. Dating to the early fourth century AD, the document lists properties
arranged by their geographical position within the town, mentioning not only
the name of the owner of each property but in a third of all cases also his
profession. Again, people performing the same occupation appear distributed
randomly over the urban area, and there is no evidence for people in the same
occupation living in adjacent houses. In total, thirty-three different professions are
verifiable. Textile crafts are also well represented; listed are λινόϋφος (four times),
λινεψός (twice), λινέμπορος (twice), γέρδιος (twice), γναφεύς (three times), and
ἱματιοπώλης (once). They are spread over the whole town of Panopolis.
That textile professions and textile crafts in Egyptian towns were not
spatially concentrated is also clear from the situation at Kellis. This village in
the Dakleh Oasis was occupied from the late Ptolemaic Period until the end of
the fourth century AD and was known for its textile production. Implements
used for spinning and weaving alike were found in almost every household,
including spindles, spindle whorls, loom weights, shuttles, combs, and unspun

5 6
SB XIV.11978. Borkowski (1975: 160). SB VIII. 9902 with BL 6, 160.
 Roman Textile Crafts 339

N
Fisherman

W E Fisherman

Builder
Embroid-
erer
S
Dyer
Vegetable-
seller

Linen-
weaver

Carpenter

Bakery

Fig. 14.2. Craftsmen in a neighbouring residential area according to P.Oxy. 46/3300.

Image: Kerstin Droß-Krüpe.

yarns.7 In one complex, inhabited from the mid-third until the end of the
fourth century AD, the excavators found many loom weights fallen in position,
together with the remains of loom beams. The discovery of large quantities of
loom weights all over the settlement indicates that many villagers were involved
in the production of cloth. While it is possible that the textiles they produced
were meant to cater for family needs, the evidence also suggests that the craft of
weaving was highly professionalized and did not only serve to satisfy internal
domestic demand of households: many textiles were produced to be sent to
other settlements or to be sold. Papyrological evidence further corroborates this
picture: two papyri from the middle of the fourth century AD refer to (small-
scale?) textile production at Kellis. One indicates that textiles were sent to the
Nile valley.8 Several Coptic papyri, originating from the housing complex
mentioned above, show a weaving and tailoring business conducted under
the direction of a woman, Tehat, with the help of a male relative.9 Yet there is
no evidence that textile crafts at Kellis were spatially concentrated.

7 8
Carroll (1988: 23–5) and Bowen (2001: 18–28); see also P.Kellis I.71. P.Kellis I.51.
9
P. Kellis Copt. XVII. 44, 46, 48. Cf. Gardner, Alcock, and Funk (1999) and Nevett (2011:
19–24).
340 Kerstin Droß-Krüpe

But what was the situation elsewhere in the Empire? From the city of
Ephesos inscriptions show a mixture of different crafts: in the third century
AD, M. Fulvius Publicianus Nicephorus refurbished a stoa linking the theatre
and the stadium of the city. Epigraphic evidence shows that he gave away the
intercolumnia of this stoa to at least eighteen different groups of craftsmen
(or professional associations) that were offering their products in close prox-
imity to each other.10
Well-known cities such as Ostia and Pompeii can serve as further com-
parative examples. When it comes to detecting workshops of textile craftsmen
especially, those crafts needing specific technical equipment can be identified
with the help of archaeology. In Ostia, six fulling workshops (fullonicae) have
been identified.11 One fullonica was situated on the other side of the Tiber,
while the other five are lying within the city walls. These five fullonicae are
spread over the whole town, covering four of the regiones into which modern
scholars divide the town.12 Bakeries show a similar pattern. We can thus argue
that, also in Ostia, identifiable workshops were not clustered in one or several
town quarters.
A similar picture emerges in Pompeii. The six dyeing workshops and the
twelve fulleries identified can be found in several districts, including the town
centre (Figs 14.3–4).13 Other crafts do not seem to be spatially concentrated
either.14 Regardless of the degree of nuisance for the neighbouring people and
buildings, it is evident that not only the workshops of dyers but also those of
bakers and fullers were distributed throughout the whole area of the town.
Only one occupation could perhaps form an exception: the so-called officinae
lanifricariae are remarkably concentrated within regio VII.15 As the traditional
interpretation of these workshops as wool-washing facilities was based on a
graffito found adjacent to one of these workshops, the function of these
workshops is still puzzling researchers.16 It is however interesting to see that
Dionysius, a freedman of Lucius Popidius Secundus, who probably worked in
an officina lanifricaria—his name appears in several graffiti near the doorway

10
IK 12.444–5, 549; IK 16.2076–82. Cf. Ruffing (2008: 378).
11
These include fullonica I XIII 3 along the cardo, the two fullonicae in the caseggiato della
fullonica (II XI 1 and 2), the fullonica behind the Temple of the Fabri Navales (III II 2), and the
fullonica in the Via degli Augustali (V VII 3). The sixth workshop, across the Tiber, has not been
published yet. Cf. De Ruyt (2002).
12
Cf. De Ruyt (2002: 51): ‘On ne constate pas pour elles [i.e. fullonicae] de concentration en
quartier. Au contraire, leur implantation est très dispersée et évite la zone des quartiers centraux
autour de la place publique.’ The same observation occurred to her when analysing the spatial
distribution of bakeries.
13
fullonicae: I 10, 6; I 4, 7; I 6, 7; V 1, 2; VI 3, 6; VI 8, 20–21.2; VI 14, 21–22; VI 15, 3; VI 16,
3–4; VI 16, 6; VII 2, 41, and IX 6, a.1; cf. Flohr (2007, 2008, 2011); dyeing workshops: I 8, 19; V 1,
4; V 1, 5; VII 2, 11; VII 14, 5.17–18, and IX 3, 1–2.
14 15
Cf. Flohr (2007: 148). Moeller (1966); Goodman, Chapter 13, this volume.
16
CIL 4.1190. See now Flohr (2013: 57–60); Monteix (2013).
 Roman Textile Crafts 341

Fig. 14.3. Fulleries in Pompeii. Map: Miko Flohr.

Fig. 14.4. Dyeing workshops in Pompeii. Map: Miko Flohr.

of one of these workshops—designated himself as fullo.17 Given the debate

about the function of lanifricariae, further research on their productive facil-
ities would be essential to understand their spatial pattern.
Despite the situation with the so-called officinae lanifricariae at Pompeii, it
is still true that textile workshops do not usually show a clear spatial concen-
tration and are not locally clustered in particular parts of cities and towns.

17
CIL 4.2966.
342 Kerstin Droß-Krüpe

They are distributed across the entire urban area. In this respect, the picture
from the nome capitals of Egypt is similar to the well-excavated Italian cities.18
While the evidence suggests the same spatial peculiarities occurred in
different parts of the Roman world, it also indicates a fundamental difference
with some later pre-industrial economies—in medieval Europe, the clustering
of crafts in special quarters or streets is widely attested. This all raises several
issues: on the one hand, the question arises of what factors lead to a localized
concentration of certain crafts; on the other hand, the question remains of
how to understand ancient toponyms suggesting artisanal clustering. Further-
more, one has to make sense of the differences in spatial distribution between
antiquity and the Middle Ages.

CLUSTERING IN THEORY AND P RACTICE

The phenomena of clustering, (industrial) concentration, and regional special-

ization have been the subject of growing interest in economic research. In
particular, the main advantages of forming clusters attracted attention. As
early as 1890, Alfred Marshall published the first volume of his Principles of
Economics in which he developed his theory about industrial districts—the
foundation of modern cluster research.19 Marshall dealt with the phenomenon
of the localization of industry, focusing on the concentration of specialized
industries of the secondary sector (manufacturing) in particular localities,
which he called ‘industrial districts’.20 According to Marshall, the main advan-
tages of a clustered industry—both regionally and within one settlement—were
(1) an environment that promotes innovation and (2) the development of a local
labour market for highly specialized workers.21 One hundred years later, Michael
Porter’s influential The Competitive Advantages of Nations led to the break-
through of modern cluster theory.22 In this book, Porter argued that the
geographical concentration of industries and crafts increases productivity, as it
gives entrepreneurs access to specialized skills. Moreover, it fosters innovation:
competition between adjacent firms or workshops raises the incentive to innov-
ate. Though Porter focused on regional and national clusters, his considerations
may be applied to intra-urban situations as well. Modern cluster research
has shown that the tendency of economic activity to cluster in particular
locations is driven mainly by efficiency advantages (such as lowered transaction
costs), flexibility advantages (such as high mobility of labour), and innovation

18
This observation once more challenges the old paradigm that Egypt was a place where
circumstances were different from those in all other regions of the Roman Empire.
19 20 21
Marshall (1890). Marshall (1890: 268–84). Marshall (1890: 328–33).
22
Porter (1990).
 Roman Textile Crafts 343

advantages.23 When it comes to the establishment of clusters, external forces

such as the historical and cultural background of a region or city, its geographical
and infrastructural characteristics, and existing institutions and organizations
are, however, crucial as well. Infrastructural prerequisites, such as access to
waterways and other transport connections, play an especially decisive role.24
Favourable local circumstances, such as the vicinity to a junction of several major
roads or an advantageous ground water level, can encourage clustering.
Yet in Roman cities, none of these factors seems to have had any influence
in the location choice of textile crafts. In cities, access to fresh water, a condicio
sine qua non for fulleries and dye-works alike, was generally unproblematic—
even in Egypt: P.Lond. III.1177, an extensive part of a set of accounts of the
urban water supply in Ptolemais Euergetis dating to the reign of Trajan, shows
that the water supply of the nome capital of the Arsinoites was extremely well
organized and that several quarters of the town were connected to the water
supply system.25 A brewery even had a direct connection to the piped system.
While no other workshops are explicitly mentioned in the document, this
papyrus clearly attests the existence of a widely accessible and well-organized
water infrastructure. The implication is that spatial concentration of water-
intensive businesses was unnecessary: water was available everywhere across
town without great difficulty, and therefore did not prompt a concentration of
fulling or dyeing workshops in a particular part of town.26
While infrastructural conditions did not foster clustering, neither did textile
workers to have benefited from economies of scale. On the contrary, compared
to the cluster benefits already discussed, it seems to have been more advanta-
geous and more convenient for craftsmen to have an even distribution of
weavers, fullers, and dyers (and thus of qualified craftsmen) over the area of a
town or city. To some extent, this might be explained from the fact that the
innovative nature of the techniques used in these crafts was rather poor—the
most important innovations being the invention of the spinning-wheel and
treadle loom during the High Middle Ages. Though several apprenticeship
contracts for weavers from Roman times are known, next to nothing is known
about the education of fullers and dyers at that time.27 It seems that highly

23
Porter (1990: 33); cf. Goodman, Chapter 13, this volume.
24 25
Sölvell (2009: 24–5). Habermann (2000).
26
If we extend our geographical horizon, we may also include an inscription from Antioch on
the Orontes, dating to the reign of the emperor Vespasian. This inscription seems especially
interesting with regard to the integration of fulleries into an existing water supply network. In it,
the construction of a ‘fullers’ canal’ 2.5 kilometres long and 3.5 square metres in cross-section is
mentioned, leading the water from the river Orontes into the city. Cf. Feissel (1985); SEG
35.1483.
27
Hoernes (2011) was able to detect sixty-two references to apprentices for different crafts
and arts from Roman times. Up to now only two references from the Hellenistic period are
known. From the middle of the third century BC one tax register mentions apprentices of fullers
(SB X.10447 from the Herakleopolites). From the Delphian manumission reports from the
344 Kerstin Droß-Krüpe

specialized workers were not needed, or needed only to a small extent—at least
when it comes to fulling and dyeing. In summary, for these crafts no efficiency
advantages emerged, nor were there flexibility advantages or innovation
advantages.
Yet not only was it true that clustering would not have brought advantages
for textile craftsmen; it is also possible that they would have benefited from
being more-or-less evenly distributed across a settlement. This does not mean
that there was an active spread policy or that ancient textile workers reasoned
along the lines of modern economists—but, even if their workshops did not
follow any planned pattern, it is wrong to think that they were unaware of
good locations and competitive pressure.28 The crucial thing is that, even
though dyed fabrics can, to a certain extent, be regarded as better-quality
products, most textiles and garments predominantly serve basic human needs.
Therefore, they should be considered convenience goods.29 Producers and
retailers of convenience goods face strong competition compared to those
dealing with luxury goods or niche products. This makes it extremely import-
ant for them to keep their customers loyal and satisfied.30 As both demand for
and supply of textiles is omnipresent, textile craftsmen are under constant
pressure to secure the loyalty of their customers: the convenience goods they
produce are not very specific. Moreover, textiles are also ‘search goods’: their
quality can be checked before they are purchased and used.31 If these crafts-
men were clustered, very similar products and services would be on display
directly next to each other. A conglomerate of textile workshops would make
comparing qualities and prices much easier for the customers. Quality and
price advantages would carry more weight, especially because the loyalty of
customers of convenience goods is comparatively small anyway. Conversely, if
textile workshops were spread over the settlement, comparability would be
reduced because search costs—and thus transaction costs—for the consumers
would increase. In the case of fulling and dyeing, an equal distribution of
workshops throughout the city would increase customer loyalty. In such a
situation, it would be more likely that customers would return to a fuller or a
dyer that was either situated in their neighbourhood or with whom they
maintained intense social relations—provided that price differences were not
too dramatic.

second century BC we learn that a certain Sosas obtained his liberty and was required to learn the
craft of fulling with Artemidoros. After finishing his apprenticeship, he was supposed to work for
his former master Dromokleidas (SGDI II.1904). Later sources are lacking.
28
e.g. Tab.Vindol. II.343: Octavius reports that a contubernalis of his friend Frontius had
come to him and asked for hides, but did not turn up again, because he already gathered them
elsewhere. Cf. Grønlund Evers (2011: 15–18). For considerations about dealing with competition
in ancient times, see Kudlien (1994).
29 30 31
Holton (1958). Cf. Bufe (1981: esp. 21–43). Nelson (1970: 318 ff.).
 Roman Textile Crafts 345

T O P O N YM S

To sum up, local city administrations or governmental institutions did not

regulate or promote the spatial concentration of individual business sectors,
and infrastructural factors, such as access to fresh water, nor did they foster
clustering. External factors or obligations, therefore, did not lead to spatial
concentrations of textile artisans. Moreover, producers of convenience goods
like everyday textiles had few advantages to expect from clustering: from their
point of view, it seems to be much more reasonable to foster an equal spread of
production and retail across the whole settlement. The logical consequence,
however, is that clustering might have been advantageous for producers and
retailers of shopping or luxury goods. This brings us back to the toponyms
discussed at the start of this chapter.
Ancient toponyms connected to crafts are actually rather scarce—and
almost all of them come from the city of Rome.32 As can be seen from
Table 14.4, craft toponyms cover a wide timeframe. Strikingly, all vici whose
name refers to an occupation are mentioned only once; the late antique
descriptions of the Roman regions list several toponyms, but not one is
known from earlier sources. One suspects that the earlier designations were
no longer used during the fourth century.
Some of the crafts mentioned in Table 14.4 indeed deal with luxury items or
shopping goods. These people, who include perfume-sellers, ring-makers, and
glassworkers, may have found it attractive to form clusters because their
products were subjected to different market rules.33 The precious raw mater-
ials they process or the products they trade find their way to their workshops
by means of only a few tradesmen or retailers: they are not offered by a great
quantity of providers; clustering may improve the position of these craftsmen
by strengthening their bargaining power vis-à-vis the traders. Furthermore, a
spatial concentration of these craftsmen would automatically lead to a spatial
concentration of the demand for their products. Again, this does not mean any
active cluster policy existed; a formation of clusters of luxury products may
have followed practical reasons or the trial-and-error method. The clustering
of materiarii and frumentarii may be explained by a peculiarity that connects
these two occupations: both are related to goods that not only had to be brought
in from rather far away, but also were relatively expensive to transport, given
their small margins of profit. Thus they were to a much bigger extent than textile
workers depending on imported products, transported over land or sea from
outside. If we take this into consideration, these craftspeople might actively have
chosen to build their workshops next to the warehouse districts.

32
Some older publications list the vicus mundiciei among the craft toponyms, but, according
to Richardson (1992: 426), the name is related to the well-attested gens Mundicia.
33
Rogers (1965); Porter (1990).
346 Kerstin Droß-Krüpe
Table 14.4. Toponyms from Rome referring to crafts
Toponym Source Date Location

forum vinarium CIL 6.9181–2 = Unknown Unknown

ILS 7502
portus vinarius CIL 6.9189 = ILS Unknown Unknown
7929; CIL 6.9090;
CIL 6.37807 = ILS
9429
vicus lorari CIL 6.9796 Unknown Unknown
inter lignarios Livy 35.41.10 192 BC Outside the Porta Trigemium
inter figulos Varro, LL 5.154 first century BC Unknown
inter falcarios Cic. Catil. 1.8.4; 63/62 BC Unknown
Sull. 52
scalae anulariae Suet. Aug. 72,1 Augustan Near Forum Romanum?
vicus sandaliarius CIL 6.761 AD 12 Regio IV, probaby north-east of
the Templum Pacis
basilica argentaria Not. Reg. fourth century Forum Nervae
AD; building
identified as
Trajanic
vicus materiarius CIL 6.975 = ILS AD 136 Regio XIII, warehouse district
6073 between the Aventine and the
(Capitoline base) Tiber
vicus frumentarius CIL 6.975 = ILS AD 136 Regio XIII, warehouse district
6073 between the Aventine and the
(Capitoline base) Tiber
vicus vitrarius Cur. Urb. Rom. fourth century AD Regio I
vicus unguentarius Not. Reg. fourth century AD Regio VIII
aream carruces Not. Reg. fourth century AD Regio I, between Porta Appia
and the temple of Mars?
porticus Not. Reg. fourth century AD Regio VIII
margaritaria
campus lanatarius Not. Reg. fourth century AD Regio XII
porticus fabaria Not. Reg. fourth century AD Regio XIII, between the
warehouses of the lower
Aventin?
forum pistorum Not. Reg. fourth century AD Regio XIII, near the horrea at the
southern Aventine?

Some other towns offer similar toponyms, but the evidence is scarce. Quite
often leatherworks are mentioned as ἡ σκυτικὴ in Apamea, ἡ πλατεία τῶν
σκυτοτόμων in Saittai, and ἡ πλατεία σκυτέων in Hermoupolis Magna.34 From
some eastern regions we also know of ἄμφοδα boring the name of crafts or

34
Apamea: IGR 4.788 ff.; Saittai: TAM 5.79 ff. (AD 153), TAM 5.146 (AD 166–7), and TAM
5.81 (AD 173–4); Hermoupolis Magna: P. Brem. 23 (AD 116).
 Roman Textile Crafts 347

trade: τὸ ἄμφοδον λινυφείων in Ptolemais Euergeteis, Soknopaiou Nesos,

and Theadelpheia, τὸ ἄμφοδον ποιμενικόν and τὸ ἄμφοδον χηνοβοσκῶν in
Oxyrhynchos, and, finally, τὸ ἄμφοδον σειτικόν in Skythopolis.35 If we recall
the tax registers from Ptolemais Euergetis, it can be shown that craft toponyms
are hardly reliable as witnesses of spatial distribution. As indicated, these
registers mention several bakers (ἀρτοκόποι), while, at the same time, they
refer to a street as ἐν τοῖς ἀρτοκόποις.36 Though there are indeed two bakers
living in this street (Πλούταρχος and Σαβεῖνος), six others practise their
profession in different parts of the nome capital. Thus, a toponym deriving
from a craft or trade does not necessarily refer to a spatial concentration of
these particular economic activites.

AN TIQUITY AND MEDIEVA L EUROPE

One final point concerns the difference between antiquity and the Middle Ages:
the picture from the Roman world contrasts sharply with that from medieval
Europe. Especially for later medieval Europe, the concentration of crafts in specific
parts of the town is well known. Smithies, tanneries, glassmaking workshops, and
fulleries—workshops either bearing a high fire risk or causing strong olfactory
nuisance—were often located in the periphery of towns in somewhat less popu-
lated areas. The existence of manufacturing quarters in medieval London has been
identified by archaeological excavations.37 The establishment of such craft quarters
was often controlled by town officials. For example, bakers, potters, and bell-
founders were banned from the city of Basel by a decision of the town council after
the city fire of 1427. In Siegen, in 1561, tanners and butchers were forced to live and
work in one particular street because of the bad smell of their workshops, while
blacksmiths and locksmiths were brought together in another street because of the
noise their crafts produced.38 Some textile crafts were spatially concentrated as
well. For example, wool-weaving workshops in medieval Göttingen were concen-
trated in one town quarter—even though it was a fairly innocuous craft.39
The origin of these differences between antiquity and the Middle Ages may lie
in the different roles—and therefore different power—of both town officials and
professional associations. Where medieval guilds were in control, they shaped
labour, production, and trade within their town and their craft.40 Guilds con-
trolled and restricted economic interaction and market activity by creating

35
SEG 8.43 (first century AD). Ruffing (2008: 379) remains sceptical about the diagnostic value of
these toponyms and considers it ‘fraglich, inwieweit die Toponyme auf eine tatsächliche Konzen-
tration der jeweiligen Gewerbe in dem betreffenden Viertel oder an der betreffenden Straße deuten’.
36 37
BGU IV.1087. Schofield and Vince (2003: 144).
38 39
Cramer (1981: 75 with nn. 283, 284). Steenweg (1990: 300).
40
Cf. Kluge (2007) and Epstein and Prak (2008).
348 Kerstin Droß-Krüpe

oligopolies or monopolies. Guild coercion could also lead to a forced amal-

gamation of all craftsmen performing the same craft: concentrating craftsmen in
certain streets or town quarters made it easier to check whether members
complied with the rules and regulations established by the guilds. Thus, one
advantage of spatial clustering was that it reduced enforcement costs for guilds.
For town councils, clustering made it easier to reach all people performing one
craft. Apart from that, concentrating crafts causing unpleasant smells, environ-
mental pollution, or noise in marginal town quarters was a way of banning them
from the sight of the upper classes: in many medieval towns not only safety
aspects but also pride of place led to the establishment of craft quarters.

CONCLUSIO N

To sum up: for ancient textile production hardly any concentration of workshops
in special parts of towns can be identified.41 Concentrating their workshops
in certain streets or squares was no reasonable strategy for textile craftsmen. In
contrast to later times, no state-controlled restrictions can be found that would
have led to an organized settlement of any specialized craft in determined zones of
a settlement. Indications for clustered craftsmen deriving from craft toponyms
mainly come from Rome itself. Maybe the size of this city encouraged the use of
toponyms of all kinds to ensure guidance; craft toponyms may date from a certain
agglomeration of craftsmen in a street or town quarter, but evidence shows that
they neither mean that all these craftsmen were to be found in these areas nor
indicate that the location of craft shops remained unchanged over the years. Thus,
researchers have to resist the temptation to see any craft toponym that occurs as an
indicator of clustering, and should investigate each individual case carefully.

ACKNOWLEDGEMENTS

Special thanks are due to Markus Diedrich (Marburg) for his comments on the
English version of this chapter. I would like to thank Ramona Grieb (Frankfurt)
for her useful comments on the theoretical approach and my colleagues at
Philipps-Universität Marburg and Kassel University for discussing the available
sources and their interpretation with me.

41
One notable exception is Timgad, where workshops associated with fulling or dyeing were
concentrated in the north-eastern part of the town. Cf. Lohmann (1979), Wilson (2000), and
Droß-Krüpe (2011: 126–7, 134). See also Goodman, Chapter 13, this volume.
 Roman Textile Crafts 349

REFERENCES

Beltrán de Heredia Bercero, J. (2000). ‘Los restos arqueológicos de una fullonica y de

una tinctoria en la colonia romana de Barcino (Barcelona)’, Complutum 11:
253–9.
Beltrán de Heredia Bercero, J., and Jordi y Tresserras, J. (2000). ‘Nuevas aportaciones
para el estudio de las fullonicae y tinctoriae en el mundo romano: Resultados de las
instalaciones de la colonia de Barcino (Barcelona, España)’, in D. Cardon and
M. Feugère (eds), Archéologie des textiles des origins au Ve siècle: Actes du colloque
de Lattes. Montagnac, 241–6.
Borkowski, Z. (1975). Une description topographique des immeubles à Panopolis.
Warschau.
Bowen, G. E. (2001). ‘Texts and Textiles: A Study of the Textile Industry at Ancient
Kellis’, Artefact 24: 18–28.
Bradley, M. (2002). ‘ “It all comes out in the wash”: Looking Harder at the Roman
Fullonica’, JRA 15: 20–44.
Bufe, R. H. (1981). Güterbeschaffung des täglichen Bedarfs: Ein Betrag zur Ressourcen-
allokation privater Haushalte unter dem Einfluß der Einkaufsstättengestaltung
(Betriebswirtschaftliche Schriften, 108). Berlin.
Carroll, D. L. (1988). Looms and Textiles of the Copts: First Millennium Egyptian
Textiles in the Carl Austin Reitz Collection of the California Academy of Science.
Seattle and London.
Cramer, J. (1981). Gerberhaus und Gerberviertel in der mittelalterlichen Stadt (Studien
zur Bauforschung, 12). Bonn.
De Ruyt, C. (1995). ‘Ricerche archeologiche nel tempio dei Fabri Navales a Ostia’,
Archeologia Laziale 12: 401–6.
De Ruyt, C. (2001). ‘Les Foulons, artisans des textiles et blanchisseurs’, in J.-P. Descoedres
(ed.), Ostia : Port et porte de la Rome antique. Geneva, 186–91.
De Ruyt, C. (2002). ‘Boulangers et foulons d’Ostie à l’époque impérial: Quelque
réflexions sur l’implantation de leurs ateliers et sur leurs fonctions précises dans la
ville portuaire’, in J.-C. Béal and J.-C. Guoyon (eds), Les Artisans dans la ville
antique (Collection archéologie et histoire de l’antiquité Université Lumière Lyon,
2, 6). Paris, 49–53.
Droß-Krüpe, K. (2011). Wolle–Weber–Wirtschaft: Die Textilproduktion der römischen
Kaiserzeit im Spiegel der papyrologischen Überlieferung (Philippika, 46). Wiesbaden.
Epstein, R. S., and Prak, M. (eds) (2008). Guilds, Innovation, and the European
Economy, 1400–1800. Cambridge.
Feissel, D. (1985). ‘Deux listes de quartiers d’Antioche astreints au creusement d’un
canal (73–74 après J.-C.)’, Syria 62: 77–103.
Flohr, M. (2003). ‘Fullones and Roman Society: A Reconsideration’, JRA 16/2: 447–50.
Flohr, M. (2007). ‘Nec quicquam ingenuum habere potest officina? Spatial Contexts of
Urban Production at Pompeii, AD 79’, BABesch 82: 129–48.
Flohr, M. (2008). ‘Cleaning the Laundries II: Report of the 2007 Campaign’, Fasti Online
Documents & Research 111 <www.fastionline.org/docs/FOLDER-it-2008-111.pdf>
(accessed 1 March 2015).
350 Kerstin Droß-Krüpe
Flohr, M. (2011). ‘Cleaning the Laundries III: Report of the 2008 Campaign’, Fasti
Online Documents & Research 214 <http://www.fastionline.org/docs/FOLDER-it-
2011-214.pdf> (accessed 1 March 2015).
Flohr, M. (2013). ‘The Textile Economy of Pompeii’, JRA 26/1: 53–78.
Flohr, M., and Wilson, A. I. (2011). ‘The Economy of Ordure’, in A. O. Koloski-Ostrow,
G. C. M. Jansen, and E. M. Moormann (eds), Roman Toilets: Their Archaeology and
Cultural History (BABesch, Supplement 19). Leiden, 147–56.
Gardner, I., Alcock, A., and Funk, W.-P. (1999). Coptic Documentary Texts from Kellis
I. Oxford.
Grønlund Evers, K. (2011). The Vindolanda Tablets and the Ancient Economy (BAR
British Series, 544). Oxford.
Habermann, W. (2000). Zur Wasserversorgung einer Metropole im kaiserzeitlichen
Ägypten. Neuedition von P. Lond. III 1177. Text–Übersetzung–Kommentar (Vestigia,
53). München.
Hoernes, M. (2011). ‘ “. . . er wird dem Lehrling das Handwerk gänzlich beibringen, wie
er es selbst versteht”: Zur römischen Lehrlingsausbildung in Handwerksberufen’,
historia.scribere 3: 37–79.
Holton, R. (1958). ‘The Distinction between Convenience Goods, Shopping Goods
and Speciality Goods’, Journal of Marketing 23/1: 53–6.
Kluge, A. (2007). Die Zünfte. Stuttgart.
Kolb, F. (1995). Rom. die Geschichte der Stadt in der Antike. Munich.
Koß, G. (1995). ‘Die Bedeutung der Eigennamen: Wortbedeutung/Namensbedeutung’,
in E. Eichler, G. Hilty, H. Löffler, H. Steger, and L. Zgusta (eds), Name Studies: An
International Handbook of Onomastics/Les Noms propres: Manuel international
d’onomastique/Namenforschung: Ein internationales Handbuch zur Onomastik
(Handbücher zur Sprach- und Kommunikationswissenschaft, 11). Berlin and New
York, 458–63.
Kudlien, F. (1994). ‘Die Rolle der Konkurrenz im antiken Geschäftsleben’, MBAH 13/
1: 1–39.
Lawrence, R. (1992). A New Topographical Dictionary of Ancient Rome. London.
Lohmann, H. (1979). ‘Beobachtungen zum Stadtplan von Timgad’, in Wohnungsbau
im Altertum (Diskussionen zur archäologischen Bauforschung, 3). Berlin, 167–87.
MacMullen, R. (1974). Roman Social Relations 50 BC to AD 284. New Haven and London.
Marshall, A. (1890). Principles of Economics. London.
Moeller, W. O. (1966). ‘The “Lanifricarius” and the “Officinae Lanifricariae” at
Pompeii’, Technology and Culture 7/4 (Autumn), 493–6.
Monteix, N. (2013). ‘The Apple Of Discord: Fleece-Washing in Pompeii’s Textile
Economy: A Response to M. Flohr’, JRA 26/1: 79–87.
Nelson, P. (1970). ‘Information and Consumer Behavior’, Journal of Political Economy
78: 311–29.
Nevett, L. (2011). ‘Family and Household, Ancient History and Archaeology: A Case
Study from Roman Egypt’, in B. Rawson (ed.), A Companion to Families in the
Greek and Roman World. Oxford, 15–31.
Pietrogrande, A. L. (1976). Le Fulloniche (Scavi di Ostia, 8). Rome.
Porter, M. E. (1990). The Competitive Advantages of Nations. New York.
Richardson, L., Jr (1992). A New Topographical Dictionary of Rome. Baltimore.
 Roman Textile Crafts 351
Rogers, A. (1965). ‘A Stochastic Analysis of the Spatial Clustering of Retail Establish-
ments’, American Statistical Association Journal 60: 1094–1103.
Ruffing, K. (2008). Die berufliche Spezialisierung in Handel und Handwerk: Unter-
suchungen zu ihrer Entwicklung und zu ihren Bedingungen in der römischen
Kaiserzeit im östlichen Mittelmeerraum auf der Grundlage griechischer Inschriften
und Papyri (Pharos, 24). Rahden.
Schofield, J., and Vince, A. (2003). Medieval Towns: The Archaeology of British Towns.
2nd edn. Leicester.
Sölvell, Ö. (2009). Clusters: Balancing Evolutionary and Constructive Forces, 2nd edn.
Stockholm.
Steenweg, H. (1990). Göttingen um 1400. Göttingen.
Taylor, C. (2005). ‘The Disposal of Human Waste. A Comparison between Ancient
Rome and Medieval London’, Past Imperfect 11: 53–72.
Thüry, G. E. (1995). Die Wurzeln unserer Umweltkrise und die griechisch-römische
Antike. Salzburg.
Wilson, A. I. (2000). ‘Timgad and Textile Production’, in D. J. Mattingly and J. Salmon
(eds), Economies beyond Agriculture. London, 271–96.
 15

Industry and Commerce in the City

of Aquincum
Orsolya Láng

In over 120 years of archaeological excavations in the city of Aquincum,

several shops, workshops, and even town quarters dedicated to industrial
and commercial activities have been identified. This evidence was usually
connected to the heyday of the town in the second and third centuries AD
and draws a suggestive picture of traditional trade routes in the area and of the
inhabitants of Aquincum. However, thus far the discussion has mainly been
based on the results of early excavations, whose methodology does not meet
modern criteria. This chapter will start with a brief overview of the history of
the civilian settlement of Aquincum and a brief description of the economic
aspects of the urban topography of Aquincum as it was reconstructed in
earlier scholarship; it will then present a case study focusing on recent control
excavations carried out in the north-eastern zone of the civilian settlement.
These shed new light on the topographic position and importance of this town
quarter, its functions, and the phases in its development. Of particular interest
is the probably quite malodorous glue-, hide-, and horn-making workshop
that is the first fully excavated workshop of its kind in the town. It appears that
such smelly, dirty, and even flammable activities were not prohibited from
the urban centre, but indeed welcomed. This chapter will discuss some of the
reasons for this attitude.

THE HISTORY OF THE CIVILIAN SETTLEMENT

O F A QUIN C UM

The civilian settlement of Aquincum is part of the tripartite settlement

complex of Aquincum that was established on the left bank of the Danube,
in the area of modern Budapest (Fig. 15.1). The first military forts were built
 Industry and Commerce in the City of Aquincum 353

Fig. 15.1. The topographical layout of the settlement complex of Aquincum. Image:
Orsolya Láng.

between the middle and the last third of the first century AD.1 During the reign
of Domitian a permanent legionary fortress or castra was established at one of
the main crossing points of the river.2 A military town (canabae) grew around
the castra, and by the late first century a forerunner of the later civilian
settlement was developed some 5 kilometres to the north. This town started
out most probably as a roadside settlement dominated by dwellings in the

1 2
H. Kérdő (2003: 81). Németh (2003: 87).
354 Orsolya Láng

form of sunken-floored housing.3 Among the settlers there were veterans and
their families, merchants and native Celts.4
When in AD 106 the province of Pannonia was divided into two parts,
Aquincum became the capital of the province of Pannonia Inferior.5 This new
role may have provided a strong impetus for the development of the town.
Structures with adobe walls appeared, and within a short time buildings with
stone foundations were erected. The aqueducts, which primarily served the
fort, but also supplied water to the civilian settlement, may have been com-
pleted during the reign of the emperor Trajan. This may also have been the
moment in which the city’s first street network was laid out. It was adapted to
the existing main roads and the aqueduct.
A milestone in the life of the town was the year AD 124, when the settlement
was raised to the rank of municipium and acquired an independent town
council. Around this time, Aquincum began to develop a genuinely urban
character: in connection with the aqueducts, a sewer system was completed, the
construction of the town walls began, and, to the north, outside the walls, the
civilian amphitheatre may have been built. In this period, urban development
within the already established system of insulae was still rather spotty: plots
were not yet entirely built up. Yet the buildings were made primarily of stone,
and the majority of the public buildings known today were completed, includ-
ing the so-called Great Public Baths and a small open-air shrine to Fortuna
Augusta. A podium temple devoted to the Capitoline Triad already stood in the
middle of the forum. The main north–south road, the cardo, was still several
metres wider than what is now visible; it was bordered by a row of shops
or workshops on the western side and an arcade with rows of columns on
the east. Strip buildings appeared on the north side of the decumanus. Their
shorter sides opened onto the road with shops and workshops, while living
quarters were towards the back. A small pottery workshop was built south of
the Great Public Baths. To the east the so-called Double Baths also had a
second-century phase.
Yet the golden age of the city came with the economic boom following the
Marcomannic Wars and, especially, with its promotion to the rank of colonia
in AD 194, which led to an increase in population (Figs 15.2. and 15.3). The
physical appearance of the prosperous city changed: the urban landscape
became much more densely built up, with structures using every bit of space
available, separated only by narrow alleyways. Existing buildings were
expanded where this was possible, and the city ventured out beyond the city
walls. Public buildings, such as the Great Public Baths, took on their most

3
For the earlier research, see Nagy (1971b: 59–81). The earliest history of the town, and
particularly of its north-eastern zone, is discussed in Láng (2012).
4
Nagy (1973: 132).
5
For the history of the civilian settlement, see Zsidi (2004: 188–99).
 Industry and Commerce in the City of Aquincum 355

0 100 m

Fig. 15.2. Plan of the civilian settlement in the third century AD. Map: K. Kolozsvári.

Fig. 15.3. Aerial photo of the civilian settlement. Photo: G. Rákóczi.

356 Orsolya Láng

elaborate forms, and the podium temple of the forum received a new function:
it became a shrine to the cult of the emperor. An elite district developed with
luxurious residences—such as the House of the Dirce Mosaic—and the mer-
ging of several buildings created the most magnificent complex of the city—
the so-called Great Mansion, with its peristyle courtyard, its richly decorated
dining room, and its own bath wing. Additionally, strip buildings were
constructed along the southern edge of the town. At this time a broader
class of merchants joined the population, which thus far had consisted mainly
of native inhabitants and retired soldiers and their families from the western
provinces. These merchants included many people of eastern origins: the
epigraphic and archaeological record suggests that people from North Africa
and the Middle East appeared in Aquincum.6 Many soldiers from these areas
also ended up here.7
This golden age may have lasted until about the middle of the third century,
when life in the city appears to have deteriorated owing to frequent barbarian
raids, the unstable political situation in the empire, and economic decline.
Except for a few larger construction projects—such as the macellum—only
minor repairs were performed on the buildings, and, judging from the material
finds, the civilian settlement went through a difficult period. Unfortunately,
this is all that can be said about this period on the basis of the archaeological
finds. This is related partially to fluvial erosion, and partially to quarrying in the
Middle Ages, though the primitive techniques of the first excavations have also
played a role. In the eastern half of the civilian settlement, early excavations
destroyed the later strata mostly without a trace, and without proper recording.
As a consequence, we know essentially nothing about the city from the end of
the third century onwards—with the possible exception of the zone around the
cardo, where traffic must have continued. After the year 307 we no longer know
the names of the city officials. From the eastern half of the urban area, remains
from the fourth century are almost entirely lacking. Instead, scattered burials
were discovered here: during this period the population may have moved away
from the banks of the Danube, which were increasingly dangerous owing to the
uncertain military situation.

THE P ROBLEM OF IDENTIFYING W ORKSHOPS

Though the civilian settlement of Aquincum, and particularly its eastern part,
has been the subject of constant archaeological investigations since the 1880s,

6 7
Recently: Póczy (2002: 184–5); Topál (2003: 278). Póczy (1996: 144–7).
 Industry and Commerce in the City of Aquincum 357

surprisingly little is known about its economic life.8 This might be due partly
to the fact that during the most extensive excavations of the late nineteenth
and early twentieth centuries only the latest construction phases of buildings
were brought to light, and partly to the problem that attention was paid chiefly
to tracing their ground plans. Even during the large-scale conservation pro-
gramme that took place between 1959 and 1973, only a few deeper trenches
were opened. These were mainly in the zone of the forum, and in buildings
along the cardo.9 Nowadays, as the eastern half of the town is functioning as an
archaeological park, the possibility of such control excavations is even more
limited. Besides the problematic situation of the excavated areas, it should not
be forgotten that the western part of the settlement is basically unknown: only
scattered data are available on its street system and its buildings.10
In spite of these problems, some workshops and shops were identified in the
course of the nineteenth and twentieth centuries. These identifications, how-
ever, were mainly based on ground plans, persisting misinterpretations, and
speculation. The evidence is very thin. For example, Tibor Nagy identified a
soaking pit of a late-first-century tannery in the north-eastern section of the
town on the basis of sediments observed on the walls of the structure.11 The
so-called House of the Merchant of the second and third centuries was
identified based on the sole fact that there are large rooms in the building.12
The House of the Butcher—possibly dating to the third century—owed its
name and its identification to the large amounts of animal bones found there,
and to its proximity to the macellum. However, there is no information on the
character of the bones, and it is unclear whether they were household or
workshop refuse.13 A ‘wine or oil pressing workshop’—dated to the second
and third centuries—was located in one of the strip buildings in the north-
eastern zone of the settlement, and got its name from a large press slab and
three barrels filled with limelike material that were found in one of the rooms.
A fullonica was located in Building IX in the south-eastern zone of the
settlement, and its identification was based on a cross-shaped, stone-paved
construction that could instead have been a heating channel.14 Stories about
these workshops still live on, and they cannot be proven wrong, either because

8
For the history of research of the civilian settlement, see Zsidi (2006: 11–18).
9
For a summary of the programme until 1967, see Póczy (1970: 177–94).
10
Only small-scale, mainly rescue excavations were carried out here: Zsidi (2003a: 143;
2003b: 151); Láng (2009b: 18–19). However, a geophysical survey carried out with the Österrei-
chisches Archäologisches Institut in 2011 will, it is hoped, shed some new light on the structure
of this part of the town.
11
Nagy (1964a: 16). However, according to a re-evaluation of the excavation documentation,
the pit turned out to be a drainage ditch, and the same type of sediments were observed during a
control excavation in the neighbourhood: these are mainly precipitations of iron and other
minerals as a result of the steeping-sinking ground water: Láng (2012).
12 13
Recently: Zsidi (2006: 97). Zsidi (2006: 95–6).
14
Heating channel: Kuzsinzsky (1899: 27); fullonica: Zsidi (2006: 89) with earlier bibliography.
358 Orsolya Láng

the layers, finds, and structures related to the room or building in question are
already destroyed or because there is no possibility for a control excavation.
However, there are a few examples of workshops with more promising
evidence as well. Beneath the macellum, a small pottery workshop dating to
the beginning or middle of the third century was discovered. Finds confirming
its identification included a small kiln, moulds for oil lamps, fragments of
firing waste and even a clay rod to hold the vessels in the kiln, and a thick,
burnt, layer of debris.15 Further, while the fullonica probably never existed, the
presence in its environment of houses with large backyards, wells, and mill-
stones does point towards a quarter with a strongly economic character in the
southern zone of the town, and it has been thought of as an artisanal quarter.16
Besides the archaeological data, epigraphic evidence also attests the presence
of a number of commercial or industrial activities in the city. Commemorative
inscriptions from the necropoleis around the city attest the presence of different
collegia from the period of the municipium onwards. The most frequently
mentioned was the collegium fabrum et centonariorum, which started off as
a combined collegium but was later divided into two separate ones.17 Its best-
known praefectus was Caius Iulius Viatorinus, who famously donated a water
organ to the collegium in AD 228.18 Inscriptions mentioning a collegium nego-
tiantium, a collegium dendrophorum, and a nummularius attest other occupa-
tions that were performed in the civilian settlement of Aquincum.19

THE E CONOMIC HISTORY OF AQUINCUM

It is clear that our understanding of the economic history of Roman Aquin-

cum is compromised by the quality of the evidence. Besides the difficulties
already mentioned of identifying specific economic processes, another prob-
lem that plays a role is that of periodization and dating. In what follows, only
the workshops and shops that are identifiable in the archaeological record will
be discussed, and there will be an emphasis on their chronological develop-
ment. This may help us to discuss whether there are broader historical
tendencies in the economic life of the civilian town of Aquincum (Fig. 15.4).
In the very first phase of the settlement, in the late first century, the roadside
sunken-floored houses of the vicus-like settlement and other structures were

15
Recently, on the dating of this workshop: Láng (2003: 167, 172–3).
16
For a summary on this quarter, see Zsidi (2006: 77–89).
17
Sometime around the middle of the second century: Nagy (1973: 136); Zsidi (2002a: 112);
Liu (2008: 66).
18
Nagy (1934).
19
Collegium negotiantium: CIL 3.10430; dendrophorum: see Nagy (1973: 137); Póczy and
Zsidi (2003: 198); nummularius: CIL 3.3500 (of unknown provenance); Ürögdi (1964: 239–45).
 Industry and Commerce in the City of Aquincum 359

Municipium

Colonia

4th century

Fig. 15.4. Ground-plan of the civilian settlement with the identifiable workshops and
shops of different periods. Map: K. Kolozsvári.
360 Orsolya Láng

concentrated along the decumanus running down to the Danube. Thus, the first
workshops must also have been located here. Metal-working could have been
one of the most important activities; a smelting oven was identified during a deep
probe below the northern part of the so called Basilica (Building I).20
The early decades of the second century, when Aquincum became a proper
town, still saw the existence of industrial activity in the area of the former vicus: here,
north–south-oriented strip buildings were constructed with large rooms opening
on the decumanus.21 Their northern parts were perhaps used as dwellings, while the
southern parts facing the road could have been used for economic purposes.22
This is also the time when the emphasis was shifted to the main north–
south road (cardo) coming from the legionary fortress: the road was lined with
a wooden portico on the east side of the cardo; the first phase of the row of
shops along the western side is also datable to this period, or slightly earlier.23
The small pottery workshop along the main north–south road was also
established in this period. From stylistic analyses of gravestones, it is thought
that two stone-carving workshops were active in second-century Aquincum,
but the locations of these workshops are uncertain.24
Most of the ruins that are presently visible date from the most flourishing
period of the town’s life—the colonia period of the third century—and infor-
mation on the town’s economic life in this period is more abundant. The shops
along the west side of the cardo were reconstructed at this time (Fig. 15.5).25
Guidebooks to the site usually specify the function of these shops:26 allegedly,
the retail of local pottery, statuettes, oil-lamps, and Samian Ware is attested,
and one shop is even thought to have accommodated a money-changer.27 Yet,
because of the lack of finds and proper excavation records, the purposes for
which the shops were used are actually completely unknown.28 Moreover, the
situation of these rooms seems to have been more complex: the western part of
the rooms remains hidden underneath the modern road. Rescue excavations
connected to a reconstruction of the road in the 1970s revealed ample remains
of this part of the complex, including the remains of doors and window at the
rear below the arches of the aqueduct and a staircase leading to an upper
floor.29 The rooms, 6 metres wide, measured 15–20 metres from east to west.
Their relatively large size, the discovery of underground storage rooms during
the rescue excavation, combined with the fact that large amounts of fresco
fragments were discovered in one of these and the additional evidence for

20 21
Nagy (1971b: 64). Láng (2012: 215).
22
This is mainly based on the periodization of Building XXIX. See Láng (2012: 216).
23 24
Németh and Hajnóczi (1976: 423). Nagy (1973: 158).
25
Nagy (1964b: 302).
26
Recently, with earlier bibliography: Póczy (2003: 148–9); Zsidi (2002a: 167; 2006: 74–6).
27
Juhász (1936: 33–48); Zsidi (2002a: 67).
28
The rooms were excavated in the nineteenth century, and only two of them were checked
by later excavations. None of these later excavations produced finds enabling the function of the
rooms to be identified: Nagy (1964b: 302); Pető (1976b: 423).
29
Póczy (1984: 21).
 Industry and Commerce in the City of Aquincum 361

Fig. 15.5. Row of workshops or/and shops along the western side of the cardo of the
settlement. Photo: P. Komjáthy.

upper storeys, suggest that we are dealing with a large-scale building that
included workshops, shops, and houses. However, further research would be
needed to determine the exact details of the complex.
Elsewhere along the cardo, the small second-century pottery workshop was,
by the middle of the third century, replaced by a tholos-type macellum built
after North African models.30 Further, the craftsmen’s houses in the southern
zone of the town are also dated to this period, while the strip houses of the
north-eastern zone received their present form. Change in the industrial
activities can also be observed here: the metal-working gave way to a glue-
manufacturing workshop (Building XXIX). This is the period when most of
these buildings must have had some industrial or commercial functions.
Outside the city centre, raw materials as well as semi-finished and finished
products indicated the presence of a stone-carving workshop next to the
amphitheatre just outside the town wall.31
Evidence for manufacturing and retail is very limited for the period after
AD 250. The row of workshops/shops along the cardo was rebuilt and retained
its commercial function.32 The macellum was heavily rebuilt around AD 270,
but its role afterwards is still uncertain; the west part of the building, which

30
Láng (2003: 165–204; 2007a: 817–30).
31
Póczy and Zsidi (2003: 201); Nagy (1973: 159).
32
Németh and Hajnóczi (1976: 423).
362 Orsolya Láng

Workshops or tabernae Building I Building XXIX

(East Wing) N
Building Public Baths
Domestic space Building I XXVIII

XXX
XXVI

XXIX
XXVII
Cardo

XXXI
XXVIII
Tabernae (?)
Decumanus

Fig. 15.6. Plan of the industrial-commercial quarter in the north-east zone of the
civilian settlement. Image: K. Kolozsvári and Orsolya Láng.

faced the road, was divided into smaller units, possibly shops. Further rebuild-
ing seems to have taken place in the fourth century.33 In the north-eastern
zone, the glue-manufacturing workshop was still functioning, and a small, late
bronze-casting workshop was excavated in one of the strip buildings.34
To conclude, there are not many securely identifiable workshops and shops in
Aquincum. Still, some general remarks can be made. Unsurprisingly, it seems
that the location of workshops and shops depended on the position of the most
important roads in the city.35 In the early phase of the settlement the decumanus
was the dominant route, which meant that the first workshops of Aquincum
concentrated along this road. However, from the first decades of the second
century onwards, the cardo, which connected the city to the legionary fortress,
became more important. Workshops and shops were built along this road to
take advantage of the busy traffic. At the same time, the still important decu-
manus became the focus of a row of strip buildings with commercial functions.
The colonia period brought the spread of workshops and shops all over the
town, though still with a concentration along the cardo and the decumanus. The
situation in the last period is less well known, but some commercial buildings
must have retained their function, while some new ones were also established.36

33
Láng (2003: 173–4).
34
The workshop was located in the eastern wing of Building I. Small bronze statuettes, a
casting ladle, and pan were discovered on the floor of one of the rooms: Póczy and Hajnóczi
(1973: 34) and excavation documentation (BTM RA inv. no. 19–73)
35
Again, we must bear in mind that here only the eastern part of the town is dealt with, so the
conclusions pertain only to this area.
36
A similar shift of emphasis of areas can be observed in case of cemeteries and large pottery
workshops outside the town: Zsidi (2002b: 135–9).
 Industry and Commerce in the City of Aquincum 363

A CASE S TUDY: BUILDING X X I X IN TH E

NORTH-EASTERN ZONE

New excavations carried out in Building XXIX in the north-eastern zone of the
settlement and the re-evaluation of some old excavation records and materials
of some of the strip buildings here seem to confirm the presence of a
commercial quarter in this part of the town.37

History of Research

The first excavations were carried out in 1890 and 1891 by Bálint Kuzsinszky,
who discovered an atrium-like room with a floor covered by stone slabs
together with column bases (Fig. 15.7).38 Some of the rooms opening from
the atrium were equipped with a hypocaust system (rooms 6, 9, and 10). An
altar dedicated to Diana and Silvanus Silvestris by M. Aurelius Pompeius was
probably discovered during Kuzsinszky’s work, together with a small marble
bust of Minerva and another female portrait in marble (Fig. 15.8).39
The excavation was extended to the southern part of the building by János
Szilágyi shortly after the Second World War.40 He identified some of the
rooms as a pottery- or lime-manufacturing workshop, based on the discovery
of three pits lined with barrel staves and filled with lime. At the same time, a
carved limestone slab, probably a press, found in the eastern part of the
building led him to the conclusion that the building also contained an oil-
or wine-pressing workshop. The coins discovered here (minted by Commo-
dus, Septimius Severus, and Severus Alexander) permitted Szilágyi to date this
phase to a period after the Marcomannic wars.41 The most recent excavations
were carried out between 2004 and 2007 by the present author.42 They served
to define the construction phases of the building, but it also proved possible to
identify some of the activities that took place in this building complex in
certain periods in the material record. Using small finds and stratigraphic data,
it was possible to identify six major construction phases in the building dating

37
Láng (2012).
38
Kuzsinszky (1891: 134–40). Kuzsinszky excavated rooms 6–11, and rooms B and D. For a
more detailed history of research, see Láng (2009a: 273–4).
39
Altar of Diana and Silvanus. The description of the excavation is not clear. According to the
museum register, the findspot of the altar was in Building XXIX, inv. no.: 64.10.96. Bust of
Minerva: inv. no.: 64.11.177. Recently: Zsidi (1993: 186). Female portrait: inv. no.: 64.11.85.
Recently: Zsidi (2006: 57).
40
Szilágyi (1950: 312–17). Szilágyi excavated rooms 16–18 and 58–64.
41
Szilágyi (1950: 317).
42
Láng (2005: 68–80; 2007b: 117–28; 2008a: 71–80; 2008b: 271–84).
364 Orsolya Láng

B 6
9

10
11

60 58
64
62

63

61

0 10 m

Fig. 15.7. Plan of Building XXIX before the control excavations. Plan: K. Kolozsvári.

from the end of the first to the end of the third century AD (Fig. 15.9).43 Each
phase was dated based on the finds that came from it, although phases from
other previously excavated buildings in the civilian settlement were also taken
into consideration.44

43
Láng (2012).
44
Most recently with earlier bibliography: Láng (2009a: 276–7; 2012).
 Industry and Commerce in the City of Aquincum 365

Fig. 15.8. Finds from B. Kuzsinszky’s excavation: an altar stone and two marble
statues. Image reproduced from BudRég 3 (1891: 138, fig. 10).

Period 2 Period 3 Period 4

Period 5 Period 6 Period 7

0 10 m

Fig. 15.9. Construction phases in Building XXIX. Image: K. Kolozsvári.

366 Orsolya Láng

Industrial Activities

Traces of industrial activities could be observed in the southern part of the

building from the second phase onwards. Some kind of metal-working was
practised here in the second and third phases of the building; work installa-
tions were concentrated on the side of the decumanus—probably not by
accident. Large amounts of metal slag were discovered from the layers in the
southern part of the house; vessels filled with slag were also found (room 62); a
row of three ovens was built along the southern end wall of the building. The
workshop was probably oriented towards the busy decumanus. Similar
arrangements have been found elsewhere in the empire, as at Sapperton
(Britain), where the ovens of a metal workshop were also placed in the street-
front rooms of one of the buildings.45 A row of small melting ovens built of stone
was found close to the street in strip buildings in Bad Wimpfen (Germany).46
However, from the fourth phase onwards the picture changed: a large
amount of horns, horn-cores, metapodia, crushed knuckle bones of animals,
and the archaeological features related to this phase (plastered working
benches, press slab, fireplace, vats lined with barrel staves) indicate a kind of
industrial activity different from metal-working. Archaeozoological research
has determined that these bones were mainly of cattle (60–80 per cent),
though sheep, goats, and even dogs were also present.47 Most of the bones
from the excavated rooms were workshop waste, but food residue, particularly
pig, was also discovered.48 Interestingly, while workshop waste mainly comes
from the southern, commercial area, food residue dominates in the northern,
domestic zone (Fig. 15.10).
Elsewhere in the Roman world, similar situations have primarily been
associated with tanning or accompanying activities, such as glue-boiling and
horn-working.49 In our case the character of the archaeozoological material
makes glue-manufacturing the most likely interpretation, as the evidence
suggests that marrow was being extracted from the crushed bones.50 This by-
product of meat-processing had a high significance in Roman times: it was used
as glue for furniture, military equipment, and a range of other objects and as
fuel in oil lamps.51

45
MacMahon (2005: 62). 46
Filgis (2001: 21–2, figs 4, 6).
47
The proportion changed from room to room: nos 18, 58, 60–4: 60%; nos.10–10a, 11: 80%.
Lyublyanovics (2005: 4; 2007: 1); Daróczi-Szabó (2009: 1).
48
More than half of the bones were found as workshop waste in rooms 17, 18, 61, 62, 64.
Lyublyanovics (2005: 9–14).
49
Regarding the problem of the identification of tanneries, see van Driel-Murray (2011:
69–72). From a methodological point of view, Building XXIX meets the criteria of this workshop
type, because of the presence of special installations and raw material waste: Csippán (2010: 32).
50
Serjeantson (1989: 139–41); Lees and Woodger (1990: 32); Láng (2005: 76–9; 2007b:
124–5).
51
Van Driel-Murray (2011: 78).
 Industry and Commerce in the City of Aquincum 367

Fig. 15.10. Proportion of workshop refuse and food residue in Building XXIX. Image:
K. Kolozsvári.

The features discovered in the building can all be associated with this
activity: benches could have been used for cutting the bones, and the fireplace
was used for boiling the glue.52 Even the press slab, which was previously

52
The case is similar in Augusta Raurica: Deschler-Erb, Schibler, and Plogmann (2002: 170,
fig. 171). Similar working benches are depicted on an eighteenth-century woodcut, along with
368 Orsolya Láng

Fig. 15.11. Press slab from Building XXIX. Photo: O. Láng.

thought to be used as a grape or oil press, can be reinterpreted (Fig. 15.11): it

may have been used for horn-pressing.53 This is suggested by the presence of
so many horns and horn cores: after the appropriate treatment and pressing,
they could be turned into a range of objects, such as furniture inlays, vessels,
and combs.54 Some of the loose finds dated to this period also may be related
to bone-working: a bone-pricker was found in room 64, and a knife handle of
antler came to light in room 62.
Interesting are the three circular vats lined with barrel staves. Elsewhere,
such vats have been seen as obvious indicators of tanneries.55 This is true in
Vitudurum, where excavators found circular wooden-lined vats, at Pompeii,
where the tannery has circular vats of opus signinum technique, and at
Saepinum, where tanning was done in conical vats made of bricks.56 At

tanning pits filled with lime: Petényi and Bartosiewicz (2010: 232, fig. 3). It is less likely that they
were used for slaughtering, as this was normally done outside the town: Choyke (2003: 217) and
van Driel-Murray (2011: 70).
53
The press slab had been considered to be for a wine or oil press, from its discovery until the
control excavations in 2004: Pető (1976a: 116); Zsidi (2002a: 77–8).
54
Van Driel-Murray (2011: 77–8). For an ethnographical analogy from York in the nine-
teenth century, see Wenham (1965: 11, 14–16).
55
Serjeantson (1989: 135); van Driel-Murray (2001: 60; 2011: 72).
56
Vitudurum: Hedinger and Leuzinger (2003: 52–3); Pompeii: Adam (1998: 351–2) and
Leguilloux (2002: 277); Saepinum: Leguilloux (2004: 50). Other tanneries have been identified in
the UK: Alchester: Wilson and Wright (1965: 208–9; 1966: 206); Burnham and Wacher (1990:
 Industry and Commerce in the City of Aquincum 369

Fig. 15.12. One of the vats during its second recovery. Photo: O. Láng.

Carnuntum, eight large, circular late Roman constructions made of clay bricks
and stone have recently been interpreted as tanning pits.57 Pannonian examples
are rare, though Nagy identified ‘soaking pits’ in the vicus of the fortress of
Alberfalva and in the northern zone of the Aquincum canabae.58 A soaking vat
and other installations were also reported from the vicinity of the fortress
in Intercisa.59
While no channel related to the vats was discovered, the Aquincum features
were in fact connected to each other, and they might not be placed incidentally
next to the western end wall of the building, where a channel ran outside the
building. The first excavator of the site noticed some ‘white material’ in the
filling of the vats, but the nature of this material is not clear (Fig. 15.12).60 As
we do not know the exact filling of the vats, we cannot exclude the possibility

95); Calleva: MacMahon (2005: 63); Hope (1907: 448); Boon (1974: 290–1); Durobrivae:
Burnham and Wacher (1990: 75); Londinium: Hall (2005: 135–6); Lakin et al. (2002: 22–3);
Salinae: Wilson and Wright (1969: 210–11); Burnham and Wacher (1990: 46); Viroconium:
Burnham and Wacher (1990: 46); MacMahon (2005: 63).
57
Gugl (2009: 1405–19). Van Driel-Murray (2011: 74) argues against this identification.
58
Alberfalva: Nagy (1971b: 62 n. 5); Aquincum, III. Raktár str. 8: Nagy (1971a: 25).
59
Visy (1977: 28–9).
60
Unfortunately, no remains of this material survives; however, some materials other than
lime can also be considered, like ash. Even though lime was used for tanning in medieval times,
370 Orsolya Láng

that tanning took place in Building XXIX, although, based on the data given,
horn-soaking/rotting seems more plausible.61 The closest analogies to the
Aquincum glue-manufacturing and horn-pressing workshop can be found
in Liberchies, Augusta Raurica, and Venta Silurum.62

Where did the Bones Come from and who were the Customers?

The large amount of animal bones found in Building XXIX could have reached
this workshop after professional cutting and sorting. The location of the
workshop is relevant in this respect. As mentioned earlier, raw material
must have been transported via the decumanus, which led down to the
harbour.63 It is possible that large-scale butcheries were clustered outside the
settlement, in the zone of the harbour, because water was easily available here,
and because there was an important crossing point to the Barbaricum.64 It
cannot be ruled out that most of livestock transports arrived here from the
other bank of the river.65 Indeed, it is suggested by Cassius Dio that livestock
was imported from the Barbaricum, and it seems highly probable that this was
the case here at Aquincum as well.66 Given the lack of detailed archaeometrical
research, it is hard to decide what proportion of the bones found in Building
67
XXIX came from beyond the border. In any case, the roadside location of the
workshop close to the river was also convenient for getting rid of the refuse.68
The decumanus was important not only from the point of view of the raw
materials and waste management, but also from the point of view of the custom-
ers. The glue manufactured here—and possibly the hides as well—was used for a
variety of purposes and was also purchased by civilian customers. However, it

there is no evidence for the same process in the Roman period: Sőregi (1939: 34), Írásné Melis
(1996: 226); van Driel-Murray (2001: 60; pers. comm.).
61
Van Driel-Murray (2011: 78). Even in the eighteenth century, this was the process for
preparing horn for pressing: Csippán (2010: 33).
62
Liberchies: Leguilloux (2004: 51–2) and van Driel-Murray (2011: 70); Augusta Raurica:
Schibler and Furger (1988: 76, 80, 94–5; more tanneries are predicted: see pp. 39, 42, 60–1, 90,
99–100, 109–10, 120, 124–5); Venta Silurum: Brewer (1993: 58).
63
The end of the road is still unknown: it is still traceable under the neighbouring housing
estate, but becomes uncertain in the area of the large Roman pottery workshop, where some
thirty wells were discovered in 1911–12: Kuzsinszky (1932: 71–5). The same problem occurs with
the location of the harbour which is still unknown: Tóth (2006: 212–13).
64
Zsidi (2007: 67).
65
Recently: Láng (2009a: 281). A similar example is known from Alchester (Britain): here
tanneries processed the hides of cattle regularly transported over the river Avon. Burnham and
Wacher (1990: 47).
66
Cassius Dio LXXI, 11.2 and Gabler (1990: 204).
67
Isotopic analyses would be the most useful method in this case, but, as barbarian herds
possibly also drank Danube water, barbarian and Roman herds could not be distinguished.
68
In case of the fortress at Velsen, refuse (including leather off-cuts and animal bones) was
thrown into the water, between the piers of the harbour: van Driel-Murray (1985: 50–1 and fig.4).
 Industry and Commerce in the City of Aquincum 371

might not be an accident that this workshop started to work sometime around the
middle of the second century: examples from the Western provinces show that,
while in the first century the Roman army was self-supplying with leather
products or had them produced centrally, this changed from the first half of the
second century. Certain leather products—such as shoes—began to be produced
by civilian craftsmen, which boosted the business of local workshops.69 Thus, we
cannot exclude the possibility that this workshop not only supplied civilians but
also worked for the legionary fortress of Aquincum.

TOPO GRAPHICAL AN D S OCIO -ECONOMIC

CHARACTERISTICS OF THE NORTH-EAST Z ONE

Building XXIX is in the north-eastern zone of the civilian settlement of Aquin-

cum. The topographical position and characteristic buildings of this part of the
town raise interesting questions—especially from the second half of the second
century onwards. The type of houses that dominates the quarter—the strip
building or Streifenhaus—is often considered as a characteristic building in
vici and other roadside settlements of Gallo-Roman origin as well as in towns
in the north-west provinces.70 These buildings are often thought of as inhabit-
ed by merchants and craftsmen. They have been found facing the main roads
of settlements such as Corstopitum (Corbridge), Londinium, Alchester, Venta
Silurum (Caerwent), Bliesbruch, and Bad Wimpfen.71 This central position in
the urban topography is obviously due to their function: merchants and
craftsmen who used the houses partly as dwellings and partly as workshops
tried to locate their business in the busiest parts of town. In the case of the
Aquincum example, this was along the decumanus.

CO NCLUSION

If we review the research history of the civilian settlement, it is clear that the
identification of workshops and shops of the settlement in nearly all periods is
problematic. This is partly due to the old, mainly inadequate, excavation
techniques, which brought only the latest phases to light, and concentrated on

69
Van Driel-Murray (1985: 55–62, 65–6) and Kocsis (2010: 171).
70
Oelman (1923: 82–8); Lohner (1999: 31); Schalles (2000: 104–6); Hales (2003: 180). Some
would trace its origins back to the days of the oppida (e.g. Bibracte in France): Ellis (2000: 87).
71
Corbridge: Burnham and Wacher (1990: 18, 46, 60); London: Milne and Wardle (1993: 34);
Alchester: Burnham and Wacher (1990: 102); Caerwent: Oelman (1923: 91); Bliesbruch: Schaub,
Petit, and Brunella (1992: 110); Bad Wimpfen: Filgis and Pietsch (1990: 455–8).
372 Orsolya Láng

complete ground plans, while little work was done in later periods to check what
had been found before. This is partly due to the fact that the function of most
establishments was identified according to their plans and to the misinterpret-
ation of artefacts discovered in them. Before we are able to discuss the socio-
economic history of the settlement, the commonplaces of the past century of
scholarship need to be deconstructed. At this point, it is very hard to place the
economic situation of the civilian town of Aquincum into a broader context and
compare it with other provincial settlements. However, on the basis of more
recent archaeological observations, and making extensive use of comparisons
with evidence from elsewhere, a few shops and workshops could be more
securely identified. These include the row of shops and workshops along the
western side of the cardo, the macellum along the same road, a small metal
workshop, and a glazed pottery workshop in the north-eastern zone of the town
where strip buildings were built from the first half of the second century
onwards. A bronze-casting workshop can be identified in the same area.
Recent fieldwork revealed the traces of different industrial activities in
building XXIX in the same zone: while from the beginning to the middle of
the second century there was a small metal workshop in the southern part of
the house, close to the decumanus, the character of the industrial activity
changed from the last third of the second century. A glue-manufacturing and
possibly also a horn-pressing workshop were established here. The latter is the
first workshop in Aquincum to be identified securely, on the basis of its finds
record (the workshop refuse of animal bones), of features related to this
industry (soaking vats, a press slab, a small fireplace, and working benches),
and of analogies with comparable remains elsewhere in the Roman world.
Based on what has been discussed in this chapter, the north-eastern zone of the
civilian settlement of Aquincum, with its strip buildings opening to the decuma-
nus, should be considered one of the key industrial and commercial quarters of
the town. A closer look at its topographical setting and its parallels also revealed
that this zone developed because of the proximity of the busy decumanus leading
to the harbour. Yet, even though the north-eastern quarter of the civilian town has
already revealed some promising examples, there is still a long way to go before
the economic landscape of the settlement is fully understood. Further fieldwork
and revaluations of old documentations are needed critically to reassess extant
hypotheses and to fill in critical gaps in our knowledge.

REFERENCES

Adam, J. A. (1998). Arte di costruire presso i Romani. Milan.

Boon, G. C. (1974). Silchester: The Roman Town of Calleva. London.
Brewer, R. J. (1993). ‘Venta Silurum: A Civitas–Capital’, in B. C. Burnham and J. L. Davies
(eds), Conquest, Co-existence and Change: Recent Work in Roman Wales. Lampeter,
56–65.
 Industry and Commerce in the City of Aquincum 373
Burnham, B. C., and Wacher J. C. (1990). Small Towns of Roman Britain. London.
Choyke, A. (2003). ‘Animals and Roman Lifeways in Aquincum’, in Zsidi (2003c),
210–30.
Csippán, P. (2010). ‘Az állati nyersanyagokat feldolgozó műhelyek azonosítási lehető-
ségei’, in J. Gömöri and A. Kőrösi (eds) Csont és bőr: Az állati eredetű nyersanyagok
feldolgozásának története, régészete és néprajza. Budapest, 31–7.
Daróczi-Szabó, M. (2009). ‘Jelentés a Peristyl–ház 2007-es ásatásának állatcsontanya-
gáról’. Unpublished archaeozoological report. Budapest.
Deschler-Erb, S., Schibler, J., and Hüster Plogmann, H. (2002). ‘Viehzucht, Jagd und
Fischfang’, in L. Flutsch, U. Niffeler, and F. Rossi (eds), Die Schweiz vom Paläolithikum
bis zum frühen Mittelalter 5–SPM V. Die Römerzeit in der Schweiz. Basle, 165–71.
Driel-Murray, C. van (1985). ‘The Production and Supply of Military Leatherwork in
the First and Second Centuries AD: A Review of the Archaeological Evidence’, in
M. C. Bishop (ed.), The Production and Distribution of Roman Military Equipment
(BAR International Series, 275). Oxford, 43–81.
Driel-Murray, C. van (2001). ‘Technology Transfer: The Introduction and Loss
of Tanning Technology during the Roman Period’, in M. Polfer (ed.), L’Artisanat
romain: Évolutions, continuités et ruptures (Italie et provinces occidentales) (Monogra-
phies Instrumentum, 20). Montagnac, 55–68.
Driel-Murray, C. van (2011). ‘Are we Missing Something? The Elusive Tanneries of
the Roman Period’, in R. Thomson and Q. Mould (eds), Leather Tanneries: The
Archaeological Evidence. London, 69–83.
Ellis, S. P. (2000). Roman Housing. London.
Filgis, M. N. (2001). ‘Ausgewählte Baubefunde des Handwerks und Gewerbes in
römische Vicus von Wimpfen’, in S. Altekamp and A. Schäfer (eds), The Impact
of Rome on Settlements in the North Western and Danube Provinces (BAR Inter-
national Series, 921). Oxford, 19–37.
Filgis, M. N., and Pietsch, M. (1990). ‘Das römische Bad Wimpfen am Neckar–
Odenwaldlimes–Ergebnisse neuerrer Ausgrabungen’, in H. Vetters and M. Kandler
(eds), Akten des 14. Internationalen Limeskongresses 1986 in Carnuntum. Vienna,
455–62.
Gugl, C. (2009). ‘Carnuntensis scutaria (Not.Dign.Occ. IX, 20). Archäologische Evidenz
für spätantike Ledererzeugung im Legionslager Carnuntum?’, in A. Morillo,
N. Hanel, and E. Martín (eds), LIMES XX: Estudios sobre la Frontera Romana.
Vol III. Madrid, 1405–17.
Hales, S. (2003). The Roman House and Social Identity. Cambridge.
Hall, J. (2005). ‘The Shopkeepers and Craft-Workers of Roman London’, in A. MacMahon
and J. Price (eds), Roman Working Lives and Urban Living. Oxford, 125–44.
Hedinger, B., and Leuzinger, V. (2003). Tabula rasa—Les Helvètes et l’ artisanat du
bois. Avenches.
H. Kérdő, K. (2003). ‘Das Alenlager und Vicus der Víziváros’, in Zsidi (2003c), 81–4.
Hope, St J. (1907). ‘Excavations on the Site of the Roman City at Silchester, Hants, in
1906’, Archaeologia 60/2: 431–50.
Írásné Melis, K. (1996). ‘Középkori lakóházak és egy XV. sz–i vargaműhely’,
ComArcHung 211–39.
Juhász, Gy. (1936). ‘A lezouxi terrasigillata gyárak aquincumi lerakata’, Arch. Érsito 49:
33–48.
374 Orsolya Láng
Kocsis, L. (2010). ‘Mire használta a római hadsereg az állati eredetű nyersanyagot a
Kr.u. évszázadokban?’, in J. Gömöri and A. Kőrösi (eds), Csont és bőr: Az állati eredetű
nyersanyagok feldolgozásának története, régészete és néprajza. Budapest, 165–74.
Kuzsinszky, B. (1891). ‘Az aquincumi amphiteatrum: Függelékül: Két lakóház. Az 1890
és részben a 1891-i papföldi ásatások’, Budapest Régiségei 3: 81–139.
Kuzsinszky, B. (1899). ‘A papföldi ásatások 1897–98 években’, Budapest Régiségei 6: 19–36.
Kuzsinszky, B. (1932). ‘A gázgyári római fazekastelep Aquincumban’, Budapest
Régiségei 11: 3–423.
Lakin, D., Seeley, F., Bird, J., Reilly, K., and Ainsley, C. (2002). The Roman Tower at
Shadwell, London: A Reappraisal. London.
Láng, O. (2003). ‘Reconsidering the Aquincum Macellum: Analogies and Origins’,
Acta ArchHung 54: 165–204.
Láng, O. (2005). ‘The Peristyle House: Authenticating Excavation in the Northeast
Part of the Aquincum Civil Town’, AqFüz 11: 68–80.
Láng, O. (2007a). ‘Did the Cosinii Build Macella? The Possible Builder of the macellum
in Aquincum’, in M. Mayer, O. G. Baratta, and A. G. Almagro (eds), Acta XII:
Congressus Epigraphiae Graecae et Latinae 2002. Barcelona, 817–30.
Láng, O. (2007b). ‘“Strip House with an Atrium Residence Wing”: Authenticating Exca-
vation in the Northeastern Part of the Aquincum Civil Town II’, AqFüz 13: 117–28.
Láng, O. (2008a). ‘“Strip House with an Atrium Residence Wing”: Authenticating Exca-
vation in the Northeastern Part of the Aquincum Civil Town III’, AqFüz 14: 71–80.
Láng, O. (2008b). ‘Functions and Phases: The “Peristyle-House” in the Civil Town of
Aquincum’, in P. Scherrer (ed.), DOMUS: Das Haus in den Städten der römischen
Donauprovinzen. Akten des 3. Internationalen Symposiums über römische Städte in
Noricum und Pannonien. Vienna, 271–84.
Láng, O. (2009a). ‘ “Unpleasant to Live in, yet it Makes the City Rich”: Functions of
Strip-Buildings in the Aquincum Civil Town, in the Light of New Discoveries’, in
Sz. Bíró (ed.), Ex officina: Studia in honorem Dénes Gabler. Győr, 271–86.
Láng, O. (2009b). ‘Richly Decorated Building in the Western Part of the Civil Town of
Aquincum’, AqFüz 15: 18–29.
Láng, O. (2012). ‘Urbanistic Problems in the Aquincum Civil Town: The So Called NE
Zone’. Unpublished Ph.D. dissertation, Universtity Eötvös Lorand, Faculty of Arts,
Doctoral School of History, Archaeology, Budapest.
Lees, D., and Woodger, A. (1990). The Archaeology and History of Sixty London Wall,
London EC 2. London.
Leguilloux, M. (2002). ‘Techniques et équipments de la tannerie romaine: L’Exemple
de l’officina coriaria de Pompéi’, in F. Audoin-Rouzeau and S. Beyries (eds), Le
Travail du cuir de la préhistorie á nos jours. XXIIe rencontre internationale d’arch-
éologie et d’histoire d’Antibes. Antibes, 267–82.
Leguilloux, M. (2004). Le Cuir et la pelleterie à l’époque romaine. Paris.
Liu, J. (2008). ‘Pompeii and collegia: A New Appraisal of the Evidence’, AHB 22/1–4:
53–69.
Lohner, U. (1999). ‘Zivile Vici im Umfeld von Flavia Solva’, in N. C. Hazel and
C. Schucany (eds), Colonia–Municipium–Vicus: Struktur und Entwicklung städ-
tischer Siedlungen in Noricum, Räetien und Obergermanien (BAR International
Series, 783). Oxford, 29–41.
 Industry and Commerce in the City of Aquincum 375
Lyublyanovics, K. V. (2005). ‘Jelentés az, Aquincum–polgárváros, Peristyl-ház’ lelőhely
állatmaradványainak vizsgálatáról’. Unpublished archaeozoological report, Budapest.
Lyublyanovics, K. V. (2007). ‘Jelentés a Peristyl-házban 2006-ban végzett ásatásállat-
maradványainak vizsgálatáról’. Unpublished archaeozoological report, Budapest.
MacMahon, A. (2005). ‘The Shops and Workshops of Roman Britain’, in A. MacMahon
and J. Price (eds), Roman Working Lives and Urban Living. Oxford, 48–69.
Milne, G., and Wardle, A. (1993). ‘Early Roman Development at Leadenhall Court,
London and Related Research’, Transactions of the London and Middlesex Archaeo-
logical Society 44: 23–170.
Nagy, L. (1934). Az aquincumi orgona. Budapest.
Nagy, T. (1964a). ‘Perióduskutatások az aquincumi polgárváros területén’, Budapest
Régiségei 21: 9–54.
Nagy, T. (1964b). ‘A Budapesti Történeti Múzeum leletmentései és ásatásai 1960–1961-ben’,
Budapest Régiségei 21: 295–336.
Nagy, T. (1971a). ‘n. 34’, Régészeti Füzetek 171: 25–6.
Nagy, T. (1971b). ‘Der Vicus und das Municipium von Aquincum’, Acta ArchHung
23: 59–81.
Nagy, T. (1973). Budapest története I, ed. L.Gerevich. Budapest.
Németh, M., and Hajnóczi, G. (1976). ‘The Works of Rescue and Planned Excavations
Conducted by the Historical Museum of Budapest in the Years 1971–1975.–
Szentendrei út 139’, Budapest Régiségei 24: 423.
Németh, M. (2003). ‘Die Militäranlagen von Óbuda’, in Zsidi (2003c), 85–92.
Oelman, F. (1923). ‘Gallo-Römische Strassensiedlungen und Kleinhausbauten’, Bonner
Jahrbücher 128: 77–97.
Petényi, S., and Bartosiewicz, L. (2010). ‘Tímárkodással kapcsolatos adatok a
Baj-Öregkovács-hegyi koraújkori nemesi udvarház területén’, in J. Gömöri and
A. Kőrösi (eds), Csont és bőr: Az állati eredetű nyersanyagok feldolgozásának
története, régészete és néprajza. Budapest, 229–42.
Pető, M. (1976a). ‘Feltárások a II.–III: Századi aquincumi legiostábor retenturájában’,
Budapest Régiségei 24: 113–19.
Pető, M. (1976b). ‘A Budapesti Történeti Múzeum ásatásai és leletmentései 1971 és
1975 között’, Budapest Régiségei 24: 391–444.
Póczy, K. (1970). ‘Anwendung neuerer Ausgarbungergebnisse bei der Ruinenkonser-
vierung in der Bürgerstadt Aquincum’, ATASH 67: 177–94.
Póczy, K. (1984). ‘Aquincum—Castra, Canabae, Colonia’, Budapest Régiségei, 25:
15–29.
Póczy, K (1996). ‘Der afrikanische Saturnus in Aquincum an der Donau’, in
G. Bauchhenß (ed.), Akten des 3. internationalen Kolloquiums über Probleme des
provinzialrömischen Kunstschaffens. Cologne, 141–7.
Póczy, K. (2002). ‘Aquincum als Provinzhauptstadt zur zeit der Severer’, in K. Kuzmová,
K. Pieta, and J. Rajtár (eds), Zwischen Rom und dem Barbaricum: Festschrift für
T. Kolník zum 70. Geburtstag. Nitra, 183–92.
Póczy, K. (2003). ‘Wasserver- und entsorgung, Gebäude des Stadtzentrums’, in Zsidi
(2003c), 144–9.
Póczy, K., and Hajnóczi, G. (1973). ‘Nr. 63’, Régészeti Füzetek 26: 34–5.
376 Orsolya Láng
Póczy, K., and Zsidi, P. (2003). ‘Lokales Gewerbe und Handel’, in Zsidi (2003c),
185–206.
Schalles, V. H.-J. (2000). ‘Städte im Rheinland: Das Beispiel Xanten’, in L. Wamser (ed.),
Römer zwischen Alpen und Nordmeer: Zivilisatorisches Erbe einer europäischen
Militärmaacht: Katalog zur Landesausstellung des Freisraates Bayern, Rosenheim
2000. Mainz, 104–7.
Schaub, J., Petit, J.-P., and Brunella, P. (1992). ‘Bliesbruch’, Gallia Informations
2: 108–11.
Schibler, J., and Furger, R. A. (1988). ‘Die Tierknochenfunde aus Augusta Raurica:
Grabungen 1955–74’, Forschungen in Augst 9.
Serjeantson, D. (1989). ‘Animal Remains and the Tanning Trade’, in D. Serjeantson
and T. Waldron (eds), Diet and Crafts in Towns: The Evidence of Animal Remains
from the Roman to Post–Medieval Periods (BAR British Series, 199). Oxford,
129–46.
Sőregi, J. (1939). ‘Jelentés a Déri Múzeum 1939: Évi működéséről és állapotáról’,
DDMÉ 5–70.
Szilágyi, J. (1950). ‘Jelentés a Fővárosi Ókortörténeti (Aquincumi) Múzeum kutatásairól
és szerzeményeiről az 1945–1948: Évek folyamán’, Budapest Régiségei 15: 305–21.
Topál, J. (2003). ‘Ägyptische und orientalische Götter und Kulte’, in Zsidi (2003c),
274–83.
Tóth, J.-A. (2006). ‘Budapest III: The Bed of the Danube and the Aquincum Bridge’,
AqFüz 12: 212–14.
Ürögdi, G. (1964). ‘A banküzlet nyomai Aquincumban’, Budapest Régiségei, 21: 239–45.
Visy, Z. (1977). Intercisa. Budapest.
Wenham, L. P. (1965). ‘Hornpot Lane and the Horners of York’, Annual Report of the
Council of the Yorkshire Philosophical Society 2–19.
Wilson, D. R., and Wright, R. P. (1965). ‘Roman Britain in 1964’, JRS 55: 199–228.
Wilson, D. R., and Wright, R. P. (1966). ‘Roman Britain in 1965’, JRS 56: 196–225.
Wilson, D. R., and Wright, R. P. (1969). ‘Roman Britain in 1968’, JRS 59: 198–246.
Zsidi, P. (1993). ‘Zur Verehrung der Minerva in Aquincum’, Budapest Régiségei 30:
185–208.
Zsidi, P. (2002a). Aquincum polgárvárosa. Budapest.
Zsidi, P. (2002b). ‘Transformation of the Town Structure in the Civil Town of
Aquincum during the Severi (AD 193–235)’, Acta ArchHung 53: 132–49.
Zsidi, P. (2003a). ‘Verkehrsverbindungen, Wehranlagen, Straßennetz’, in Zsidi (2003c),
137–43.
Zsidi, P. (2003b). ‘Öffentliche Gebäude und Wohnhäuser außerhalb des Stadtzen-
trums’, in Zsidi (2003c), 150–5.
Zsidi, P. (2003c) (ed.). Forschungen in Aquincum 1969–2002. Budapest.
Zsidi, P. (2004). ‘Aquincum topográfiája’, Specimina Nova 18: 167–226.
Zsidi, P. (2006). Az aquincumi polgárváros: Vezető az aquincumi múzeum romkertjében
(Aquincumi Zsebkönyvek, 1). Budapest.
Zsidi, P. (2007). ‘A Duna szerepe Aquincum topográfiájában’, Budapest Régiségei 41:
45–84.
 16

The Potters of Ancient Sagalassos

Revisited
Jeroen Poblome

In 1987, during the initial exploratory survey campaigns at the archaeological site
of Sagalassos (Pisidia, south-west Turkey), which mostly focused on epigraphic
recording, and urban and monumental architectural reconnaissance studies,
considerable amounts of pottery production waste were found concentrated in
a large area to the east of the theatre (Fig. 16.1). The perceived importance of the
discovery was translated into the designation of this area as the Potters’ Quarter of
Sagalassos. Sigillata, red gloss or red slipped tableware, was considered to be its
main product.1 Nobody in the community of Roman archaeology or ceramology,
let alone in the Sagalassos Archaeological Research Project, had anticipated this
discovery.
This chapter goes back to basics and (re)builds step by step the world of the
potters of ancient Sagalassos. The natural and environmental setting of the
site is presented, as well as aspects of path dependency in the development of
the local potter’s craft. Shifts in scale of production are discussed, along
with the economy of the raw clay materials, the production process and its
organization, policies of investment and ownership, and the workings of familia
and collegia. I switch between general observations and testing on the ground so
as to give as complete a picture as possible.
Sagalassos is located in the west part of the Turkish Taurus Mountains. This
system of mountain ranges separates the Mediterranean coast of Turkey from
the central Anatolian plateau. The complex Mesozoic–Tertiary genesis of the
research area resulted in a particular triangular-shaped formation, the so-called
Isparta Angle. Sagalassos is located on the west limb of this formation, as part of
the Lycean Nappe Complex, fronting the autochthonous carbonate Bey Dağları
Platform. The Lycean nappe is mainly composed of an ophiolitic mélange,
volcanic rocks and allochthonous carbonate platforms. During the Tertiary

1
Mitchell and Waelkens (1988: 60).
378 Jeroen Poblome

Fig. 16.1. The eastern suburbs of Sagalassos featuring the local Potters’ Quarter. Map:
Sagalassos Archaeological Research Project.

nappe emplacement, flysch deposits were formed. Another late Tertiary land-
scape element worth mentioning is the volcanic Lake Gölçük and the associated
lava and tuff deposits, about 5 kilometres to the north-west of the site.2
The ancient town is tucked away in a large bend of the Ağlasun Dağı range
(c.1,800 metres above sea level) (Fig. 16.2), forming the spectacular crest to the
north of the site, with the Akdağ (2,271 m.a.s.l.) dominating the north-east
end of the range. From its position at the top of a V-shaped valley cut into the
mountains, the ancient town (1,490–1,600 m.a.s.l.) overlooked the lower areas
to the east and south. The valley acts as a permanent creek, draining several
springs where permeable limestone formations end on more impermeable
ophiolite and flysch deposits. These springs feed the middle course of the
Ağlasun Çayı, a permanent stream in the valley south of Sagalassos and a
tributary of the ancient river Kestros, part of which formed the east border of
the territory of Roman imperial Sagalassos.3
As far as connectivity is concerned, the site can be reached from the Ağlasun
Valley from its south and east sides. To the north-west of Sagalassos, the
mountain ridge is interrupted by a pass at an altitude of c.1,730 metres, leading
into the Isparta Plain.4 The Ağlasun Çayı and Valley provided a corridor towards
the east, reaching the valley of the ancient river Kestros after some 20 kilometres.

2 3
Degryse et al. (2008a); Muchez et al. (2008). Paulissen et al. (1993: 229–31).
4
Paulissen et al. (1993: 230).
 The Potters of Ancient Sagalassos Revisited 379

Fig. 16.2. View from the Akdağ towards ancient Sagalassos. The town was laid out on
the plateaux in the centre of the image. Photo: Sagalassos Archaeological Research Project.

To the west, this valley connected into the Çanaklı Plain. Most of the research
area is actually a series of interconnected mountain basins. The Burdur Plain
represents the largest tract of flat, fertile lands within the territory of Sagalassos,
about 30 kilometres from the town. Natural corridors from the Bay of Fethiye
and Antalya reach the plain, connecting into the Isparta Plain. The Augustan via
Sebaste followed the east corridor, representing, together with the Kestros Valley,
the major lines of communication from the south coast to the interior.5
Pronounced winter precipitation and summer dryness characterize the
climate in the vicinity of Sagalassos as Mediterranean, with a shorter dry
season and lower temperatures in all seasons compared to coastal zones and
significantly colder winters with a high number of frost–thaw cycles.6
The area of the site formed part of an Oromediterranean vegetation belt, with
deciduous oak forests found below needle-leaved forests. Two palynological
profiles were reconstructed from core-drillings in the central depression of the
Potters’ Quarter of Sagalassos. In their lower parts, a period of Artemisia pollen
dominance was defined, possibly indicating large-scale disturbance of the soil in
the depression in the course of the Hellenistic period. Various light-demanding
herbaceous taxa indicate the absence of forest vegetation in the same period.
Re-afforestation is reconstructed from late Hellenistic times on the slopes north
and east of the site, resulting in a cedar forest, possibly mixed with some pine and
Abies cilicica as a minor constituent throughout the Roman imperial and early

5 6
Mitchell (1993: 70–9). Paulissen et al. (1993: 231–3); Vermoere (2004: 8).
380 Jeroen Poblome

Byzantine periods. In the same periods, patches of deciduous oak woodland were
growing on the hills below Sagalassos, as were walnut trees. Grapes were
cultivated not far from the Potters’ Quarter, possibly on the slopes south of it.
Olive trees, on the other hand, were not grown at or near Sagalassos. In general,
the continuity of the pollen signal hints at rational exploitation schemes of the
forests and cultivated tree species at least until early Byzantine times.7
In the evaluation of how nature–society interactions sustained or limited
the initiation and organization of the local potter’s craft, a variety of quality
raw clay materials was found to be available at and around the site, as were
sufficient quantities of water as well as fuel sources. Sensible management
schemes would have needed to be developed optimally to exploit these natural
resources, and to avoid scheduling conflicts with other primary economic
activities, as well as to keep the balance between the different social parties
involved in artisanal production. Transport of the produce would have needed
to be organized, but would not have been a greater obstacle than for other
resources and goods transported to and from the site. The main drawback that
nature would have forced upon the production cycle was the winter months,
when potting, as much as any other professional activity for that matter, was
simply impossible.8 In considering these mostly processual aspects, however,
we need to avoid assuming that pottery production at Sagalassos was a natural
thing to happen in Roman imperial times.
Apart from integrating an interdisciplinary research strategy into the project’s
archaeological research programme, right from the stage of applying for funds,
no specific theoretical or conceptual disciplinary framework was put forward to
study and ‘explain’ the pots and potters of ancient Sagalassos. The intention was
to keep an intuitive, broad orientation within the discipline and to consider
primarily how patterns in the archaeology of the pots and potters of Roman
Sagalassos could fit with ideas and concepts of social complexity and its region-
ally specific evolution in the long term. This bottom-up approach, grounded in
the detailed study of the artefacts and the archaeological record, was preferred,
building consciously on the conviction that archaeology, as a discipline, harbours
the unique potential to explore societal development by combining the dimen-
sions of materiality and cognition with time and space.9

THE ORIGINS OF SCALE

To be sure, Sagalassos and a fortiori Roman imperial Sagalassos red slip ware did
not always exist. The exact circumstances of the genesis of Sagalassos as a

7
Vermoere et al. (2003); Vermoere (2004: 171–90).
8
For feedback mechanisms affecting pottery production, see Arnold (1985).
9
More inspiration in, e.g. Bintliff (2011) and Lucas (2012).
 The Potters of Ancient Sagalassos Revisited 381

community are still somewhat shrouded in the mist.10 During recent excavation
seasons, a construction fill behind a terracing wall was excavated in the north part
of the Potters’ Quarter, as well as some material retrieved from below the
pavement slabs of the Upper Agora.11 The pottery finds from these contexts
represent the oldest material found at Sagalassos so far and are datable to the
(later) fifth to the (early) third centuries BC. No contemporary potting activities
are attested in the Potters’ Quarter, although the macroscopical fabric charac-
teristics are in line with clay resources at and around the site. The actual slope
gradient of the excavated area in this part of the Potters’ Quarter is about 30 per
cent, so for any organized activity to take place over time the laying-out of
terraces would have been beneficial. Although it remains unclear where and
how civic life was organized in Late Classical and Early Hellenistic Sagalassos, the
use of this steep terrain and the construction of terraces would have required
some sort of communal initiative at organization.
From the fifth century BC onwards, another, larger community had settled
on the nearby plateau of Düzen Tepe, only 1.8 kilometres south from Saga-
lassos.12 We need not dwell on this socially complex community in this
context, apart from mentioning that the remains of a Classical/Hellenistic
(fifth–second centuries BC) potter’s workshop were excavated there.13 The
wares in question were of variable quality and are best considered to reflect
contemporary Pisidian styles and functions.14
Against the background of Sagalassos and Düzen Tepe coexisting at least
partially, and in the context of this chapter, it is important to consider which raw
clay materials were selected from the environment. First, potters of both settle-
ments processed clays that were available locally. In the case of Sagalassos, the
quarrying of ophiolitic clays has been proven by core-drilling and deep-soil
tomography in the central depression of the Potters’ Quarter. A palaeosol devel-
oped on top of the quarried surfaces, the formation of which was C14-dated
to between 370/360 BC and 50/40 BC, providing a terminus ante quem for the
clay-quarrying activities.15 Such activities could also be reconstructed on the east
slope of the quarter, generally pre-dating the installation of an early Roman
imperial potter’s workshop in this zone, while parts of a clay-quarrying pit were
excavated in the south-east area, which was backfilled with material datable to
the first half of the second century AD.16 The nature of the geophysical anomalies

10
Poblome et al. (2013a).
11
Excavations at Site F 2011 and 2012 were supervised by Johan Claeys (unpublished). The
2014 and 2015 excavations on the Upper Agora were coordinated by Peter Talloen. See Talloen
et al. (2015).
12
Vanhaverbeke et al. (2010).
13
Excavations at Tepe Düzen 2008 and 2011 were supervised by respectively Hannelore
Vanhaverbeke and Kim Vyncke. The workshop remains unpublished.
14 15
Poblome et al. (2013a). Six (2004).
16
On the east slope, cf. Degryse et al. (2003). Excavations at Site PQ3 2012 were supervised by
Elizabeth Murphy. Unpublished.
382 Jeroen Poblome

in the south-east part of the quarter is indicative of more clay-quarrying activ-

ities, while it is possible that the specific herbaceous vegetation in Hellenistic
times already mentioned hints at the period during which most of these activities
took place.17 At Düzen Tepe, archaeometric analysis of clays and sherds has
indicated that the clays mostly selected by potters came from an ophiolitic
provenance.18
Secondly, additional raw clay materials were quarried in the environment of
both sites. Although difficulties remain in discriminating provenances of compar-
able clay bodies in great detail in the terrain, clays with similar properties are
available in the wider Ağlasun Valley, located at the foot of both sites.19 Other than
that, archaeometric analysis at the Leuven Centre for Archaeological Sciences has
associated the earliest use of the so-called north-west Çanaklı-clays with Classical/
Hellenistic pottery. Whereas the Ağlasun Valley clays were available within a
1–3 kilometre radius, the distance to the north-west Çanaklı-clays was somewhat
larger, representing, as the crow flies, about 4–5 kilometres from Düzen Tepe and
about 7–8 kilometres from Sagalassos (Fig. 16.3). The greenish grey clays, which
had originally accumulated as part of a sequence of lake deposits in the north-west
section of the Çanaklı Valley, were confirmed as having a consistent and

Fig. 16.3. General view of the north-west part of the Çanaklı Valley. Ancient Sagalassos
was located on the mountain slopes at the back of the image, on the left. Photo: Sagalassos
Archaeological Research Project.

17
Personal communication by Branko Mušič.
18 19
Braekmans et al. (2011); Neyt et al. (2012). Neyt et al. (2012).
 The Potters of Ancient Sagalassos Revisited 383

analytically definable geochemical signature in comparison to other clays from the

same valley or elsewhere in the study region.20 This particular clay resource was
found in clay preparation pits in excavated Roman imperial workshops in the
Potters’ Quarter of Sagalassos and was archaeometrically confirmed as the main
component of contemporary Sagalassos red slip ware.21 Its exploitation had
already been confirmed in Hellenistic times and recently archaeometrical analysis
has indicated that a line of black glazed vessels was made from these clays too.22
The latter ware was found during excavations at both Sagalassos and Düzen Tepe,
albeit only in secondary deposits, which has prevented the establishment of a
stratigraphically anchored chronology. Typological features suggest a fourth–
third-century BC bracket for the ware. Although black glazed pottery was never
common at both sites, it did represent the higher end of the contemporary
tableware market. In this respect, the link between the qualities of the ware and
the selection of a specific clay source is important. When, in contrast to more
conservative Düzen Tepe, the potters of Sagalassos initiated the local production of
a Hellenistic form repertoire around 200 BC, the same link between the quality
clays of north-west Çanaklı and the better tableware was maintained.23 Although
clays were available at and around Sagalassos and Düzen Tepe and tableware was
made with these, they mainly served the production of cooking ware and other
utilitarian vessels. The extra effort to cover the distance towards the north-west
part of the Çanaklı Valley was clearly made to achive better-quality end products,
mainly tableware.
One way or another, (coexisting) Sagalassos and Düzen Tepe were dependent
on largely the same catchment area. Whether this condition resulted in tension
between both communities is not revealed by the current archaeological record,
but it must have had consequences for ownership patterns of land as well as for
marketing options for produce. Similar mechanisms must have played a role in
owning and accessing lands where clays could be quarried. In this context,
control over the clays of north-west Çanaklı should be considered. These were
apparently exploited by both sites though not by other Classical/Hellenistic
communities. Whereas in its original, Classical, phase, Düzen Tepe seems to
have been the larger settlement of the two, possibly commanding more materials
and produce, around 200 BC Sagalassos seems to have taken over the lead in the
area. The launching at Sagalassos of the production of only quality Hellenistic
tableware, implying full, if not unique, access to the clay beds of north-west
Çanaklı, seems to confirm the process of increasing control of land and resources
by Sagalassos. Unfortunately, the details of contemporary landownership pat-
terns cannot be reconstructed. Unlike the wares of Düzen Tepe, however, the
Hellenistic pottery of Sagalassos was distributed to other sites within the wider

20
Neyt (2012: 109–23).
21
Poblome et al. (2001: 159–64); Degryse and Poblome (2008).
22 23
Poblome et al. (2002a); Braekmans et al. (2011). Poblome et al. (2013a).
384 Jeroen Poblome

study area, which possibly indicates a widening radius of action.24 Düzen Tepe
dwindled from the second century BC onwards, whereas Sagalassos continued to
develop. When Sagalassos red slip ware was launched in Augustan times,
Sagalassos was not only in control but also had extensive experience with the
quality raw clay materials of north-west Çanaklı, while its pottery was also
established in a wider zone of distribution.
As a result, when we approach the pots and potters of Roman imperial
Sagalassos, it is important to consider their long-term historical background.
Pottery production can be considered to have been endemic in the area,
possibly from the early days of the formation of the community in the Late
Classical period onwards. This implies that knowledge and skill had reached a
developed stage as factors of production when Sagalassos red slip ware was
launched. This condition could also imply tradition, in the sense that both
potters and customers reacted in particular ways to patterns of change and
continuity in the market. The continued importance of mastoid drinking cups
in the Sagalassos tableware repertoire from mid-Hellenistic times into early
Roman imperial times can serve as an example.25 When, in Augustan times,
Sagalassos red slip ware was initiated with a programme of investment in the
local Potters’ Quarter, more or less the entire Roman world was following the
apparent typological attractions of Italian terra sigillata.26 The introduction of
a comparable set of tableware in the local Potters’ Quarter should have been an
option, especially considering new market conditions created by the recent
influx of thousands of veterans and their families into the region as a result of
Augustus’ policy of creating veteran coloniae in the wider Pisidian region.27 In
its initial stages, however, Sagalassos red slip ware continued to follow Hel-
lenistic morphological traditions, exemplified by the popularity of the mastoid
drinking cups. In other words, existing local processes of ontogenesis and
tradition seem to have played a role at Sagalassos, as they did with other types
of eastern sigillata.28 Possibly, such conditions could also have had an effect on
aspects such as production organization and market policies, when not every-
thing was done à la romaine from day one, if at all.
Apart from aspects of continuity, such as knowledge, skill, and tradition in
making pottery, the long-term trajectory also points at shifts. When comparing
Düzen Tepe and Sagalassos, or Classical with Roman imperial times, it should be
clear that the scale of production as well as the choice of products can be seen to
reflect a community’s ambitions and potential, and to some extent also its
identity. At Düzen Tepe, the partially excavated Classical/Hellenistic potter’s

24
Poblome et al. (2013b).
25
Poblome, Bes, and Lauwers (2007); Poblome et al. (2013b).
26
Poblome et al. (2002b); Poblome and Zelle (2002); Wallace-Hadrill (2008: 407–21).
27
Mitchell (1993: 73–9).
28
For Cypriot Sigillata, see Lund (2002); for Eastern Sigillata B, see Ladstätter (2007).
 The Potters of Ancient Sagalassos Revisited 385

workshop is so far one of a kind and does not seem to have formed part of a
dedicated artisanal neighbourhood. The presence of other workshops should
not be ruled out, but there are no indications, archaeological, geophysical, or
otherwise, for the presence of such installations in the immediate vicinity. At
Sagalassos, a badly damaged mid-Hellenistic potter’s kiln was discovered under-
neath the remains of the Roman imperial odeon. Geophysical analysis provides
hints at the presence of five more kiln structures immediately to the east of the
excavated example.29 The evidence is not yet strong enough to be considered as
direct proof, but it provides at least fairly substantial indications for the identi-
fication of the Hellenistic potters’ quarter of Sagalassos. If its location is estab-
lished by future excavations, the Hellenistic potters’ quarter would be located in
the (intra-mural) eastern part of the contemporary urban community. Taken
together with the incorporation of Hellenistic techniques of production as well
as design trends by the potters of Sagalassos, the presence of an area dedicated
to mostly potting activities indicates differences of scale from Düzen Tepe.
Moreover, these indications imply a relocation of the potters of Sagalassos in
early Roman imperial times. The expansion of the contemporary townscape,
together with the investment made in the local potter’s craft with the launching
of Sagalassos red slip ware, presumably resulted in the abandonment of the
Hellenistic location and the reorganization of part of the eastern necropolis into
an artisanal quarter, which from now on formed part of the eastern suburbs
(Fig. 16.4). Some degree of involvement and planning on behalf of the local
community must be considered in these operations, especially if we take the
afore-mentioned efforts at reforestation of the area into account. In sum, the
long-term pattern implies a build-up in importance of the potter’s craft for
the local communities, from a potter at Classical/Hellenistic Düzen Tepe mostly
providing for his own community, resulting in a marginal income to sustain his
family, to the Potters’ Quarter of Roman imperial Sagalassos, working for
markets beyond its own community, representing an asset in a regional invest-
ment portfolio.

THE ROMAN IMPERIAL P OTTERS’ QUARTER

OF SAGALASSOS

So, how did things work in the Roman imperial Potters’ Quarter of Sagalassos? An
initial attempt at reconstructing the local production organization noted the
absence of potters’ stamps on Sagalassos red slip ware.30 Such is still the case,
preventing the identification of individual potters, their output, or the distribution

29 30
Poblome et al. (2013b). Poblome (1996).
386 Jeroen Poblome

Fig. 16.4. Aerial view of the eastern suburbs. The town centre is located in the top part
of the image. Photo: Sagalassos Archaeological Research Project.

of their wares, which is, to some degree, possible with other specific types of
sigillata, produced mostly in Roman Italy or Gaul.31 A collection of stamped roof
tiles found at Sagalassos was published in the meantime.32 Of these, one Claudius
Alexandros could have been a tile workshop owner, and a member or a freedman
of one of the frequently occurring Tiberii Claudii families, holding Roman
citizenship. Other stamped personal names included Makedonikos and Demeas.
Both could have been tile workshop owners, workmen, or even slaves of an
unknown owner. If the stamp ΠΟΛ is shorthand for πολέως, then a municipal
tile workshop may have existed too. Apart from some smaller categories, so-called
ΕΛΑΙΟΥ stamps form an important category, possibly referring to a toponym of
an olive-oil (and tile-)producing estate. An unpublished study by Philip Mills
noted the variety and complexity of signatures and stamps on Sagalassos tiles in
comparison to those of other sites studied in the eastern or central Mediterranean.
The presence of stamps as well as signatures on a number of pieces suggested that
not only tile-makers were represented. Possibly tenant farmers paid for (part of)
their rent with the production of tiles, with additional identification of landhold-
ings or landlords. Although tiles represent a different medium from tableware, the
attested variety in tile production agents is interesting. This being said, no such

31
Oxé, Comfort, and Kenrick (2000); Hartley and Dickinson (2008–12).
32
Loots et al. (2000).
 The Potters of Ancient Sagalassos Revisited 387

evidence at the personal level is available for the reconstruction of the organiza-
tion of the production of Sagalassos sigillata.33
Since 1996, a targeted interdisciplinary research programme has resulted
in the excavation of some workshops.34 In general terms, the workshops of
the Potters’ Quarter are concentrated in the central zone of the eastern
suburbs, totalling 3.5–4 hectares of the latter suburban area. The excavations
confirmed the initial laying-out of workshop infrastructure from Augustan
times onwards, while the potters moved away from this specialized production
environment in the course of the second half of the sixth century AD. At all
periods, other activities too, such as burying the dead, were organized in the
eastern suburban quarter, which covered 7.5 hectares. Typically, the potters of
this neighbourhood made Sagalassos red slip ware, and no other type of
ceramic product.35 Other products and fabrics were established as part of
the Sagalassos pottery production repertoire, but these fabrics are linked with
clays and possibly places of production in the rural vicinity, adding to the
impact and scale of the pottery-producing enterprise.36 The excavations of
complete workshops (Fig. 16.5) as well as the interpretation of large-scale
geophysical work in the Potters’ Quarter indicated that the traditional

Fig. 16.5. The East Slope Workshops. The original workshop is datable between
Augustan times and the first half of the second century AD, and the late Roman
phase between the fourth and the sixth centuries AD. Plan: Sagalassos Archaeological
Research Project.

33
Cf. Fülle (1997).
34
Poblome et al. (2001); Poblome (2006); Murphy and Poblome (2011).
35 36
Poblome (1999). Degryse and Poblome 2008; Neyt et al. 2012.
388 Jeroen Poblome

infrastructural scale of production units is small. However, if we consider their

shared use of the raw clay material resource in the north-west part of the
Çanaklı Valley for the fabric of Sagalassos red slip ware and the weathered
ophiolitic clays found locally or in the Ağlasun Valley for the slip layer, the
common concept of design of the Sagalassos tableware vessels, as well as the
same markets these workshops were targeting, the concept of production is
perhaps best compared to that of nucleated workshops.37 Even though rows of
potters in a single (manu)factory were absent from Sagalassos, I did once
propose to the world of Roman ceramic studies the idea that the production of
Sagalassos red slip ware constituted a manufactory, based mainly on the
collective impact on the urban framework of Sagalassos and the scale of the
gross output of the quarter.38 Thus, depending on the classification criteria,
the potters of Sagalassos red slip ware were active in either workshops,
nucleated ateliers, or a manufactory. Perhaps this says more about the rigueur
of the classification exercise and methodology than about the varied life
experience of the potters in antiquity.39 The reality of the local archaeological
record is such that there is no easy link between the excavated structural
remains of workshops and the mode and scale of the production organization
that took place within the attested infrastructure.
Production encompasses all types of activities generating goods and services
that are made available to consumers at the appropriate time and place. In
essence, production processes are cyclical, in the sense that input in the
process generates output, affording the continued investment in input. On
the input side, the raw materials that are transformed in the production
process are defined as flowing inputs (clays, water, and fuel). Factors of
production such as labour (time and ability) and capital (tools, infrastructure,
knowledge) are also considered as input. As part of the production process,
factors of production add value to the flowing inputs, resulting in the output.
Selling the output should generate at least sufficient income to cover the
renewal of the flowing inputs as well as sustain the factors of production.
Additional income is surplus. The mode of production represents both the
scale of input and output, as well as the proportional balance between these.
The concept of the ‘firm’ as developed in New Institutional Economics
provides a useful insight into aspects of efficient investment in physical and
human capital, governance and incentives for employees.40 Even in a pre-
industrial context, there was a sufficient legal and practical framework to
facilitate the economic activity of ancient ‘firms’. To be sure, economic activity
is more than just production in this respect; distribution, trade, exchange, and

37
On the ophiolitic clays, see Degryse et al. (2008b). On nucleated workshops, see Peacock
(1982: 9).
38 39
Poblome, Malfitana, and Lund (2001). Murphy and Poblome (2011).
40
Frier and Kehoe (2007: 126–34).
 The Potters of Ancient Sagalassos Revisited 389

consumption are the other side of the coin. Although there is nothing wrong
with studying distribution and consumption patterns of wares, I would like to
argue that such research needs to be linked to production studies to achieve a
more integrated and complete understanding of the operationability of ‘firms’
in antiquity. Distribution studies in Roman ceramology in particular tend to
favour the ‘bigger’ wares. These are important in many ways, but the evidence
is mounting that the local level of pottery production was more than just a
niche. In a pre-industrial context, most goods and services consumed by
communities were typically produced in a local or regional setting. More or
less organized production units came and went in a constant but unsystematic
way. The longevity and sustainability of the production process interplay with
the mode of production.
The region of south-west Anatolia forms a good example in this respect. In
2012, evidence was presented for seven pottery production centres, at or
within the territories of Balboura (Middle Hellenistic and Late Roman),
Kibyra (Late(?) Hellenistic to Early Byzantine), Patara (Late Roman to Early
Byzantine), Xanthos (Late Roman to Early Byzantine), Araxa (Roman imper-
ial), Pednelissos (Late Roman to Early Byzantine) and Sagalassos.41 No doubt,
more production units are still out there to be discovered, but what there is
should be illustrative of the importance of the local/regional level of produc-
tion, and its variation in time, space, and output, as well as how the ‘bigger’
wares form only part of the ceramological picture of this region—in a lot of
cases even a relatively small part. More work needs to be done on aspects of
production, distribution, and consumption of these local/regional south-west
Anatolian wares, integrating hard-core ceramology with archaeometry and
conceptual archaeological approaches.
One of the factors that can help explain the impact of these different types
of local/regional pottery is investment.42 The latter mechanism can be added
to each segment of the production process, from exploiting more clays to
employing more potters or developing technologies. Potters can make invest-
ments themselves, using generated surplus, but the higher modes of produc-
tion can also attract third-party investment.
The open question is whether the production of Sagalassos red slip ware was
entirely in the hands of potters or whether (other) investors were involved as
well. I have argued in the past for a third-party investment model, partly based
on the nature of the raw clay materials.43 My null hypothesis was double: if
each potter at Sagalassos was running his own workshop, variety in raw clay
materials, fabrics, and slips of Sagalassos red slip ware should be expected, and,
if archaeometric evidence proved uniform clay, fabric, and slip compositions,
the total amount of clays quarried from such unique locations should suggest

41 42
Armstrong (2012: 35–8). Poblome, Malfitana, and Lund (2011).
43
Poblome (1996, 2006).
390 Jeroen Poblome

large-scale landownership beyond the traditionally expected means and social

position of ‘simple’ potters. The provenance and the diachronic unchanging
nature of the raw clay materials are sufficiently established in the case of
Sagalassos red slip ware.44 What is more, different, but technically viable clay
resources were identified in the north part of the Çanaklı Valley, located closer
to Sagalassos, which were apparently not used for Sagalassos tableware.45 The
non-selection of the latter clays was interpreted as being linked with land
property rights, where the owner(s) of the lands in the north part of the valley
could or would not consider converting agricultural land to a clay quarry. The
owner(s) of land in the north-west part of the valley, on the other hand, were
considered to have used access to their clay beds as an asset in their ownership
and investment portfolio, but also, taking the scale of the enterprise into
account, to have been in a position to discriminate against other landowners,
such as the ones in the north part of the Çanaklı Valley.
Knowing now that the north-west Çanaklı clays were exploited from Classic-
al/Hellenistic times into the seventh century AD I no longer consider it plausible
to presume such a static, limited, and high-profile pattern of landownership over
a millennium or so. The attested variety in Sagalassos tile production agents can
be seen as an indication to allow more flexibility as far as landownership and clay
exploitation rights are concerned. Tradition of use as well as proven qualities of
the raw materials directed attention mostly to the north-west Çanaklı clays. This
aspect of path dependency can be combined with more varied and evolving
patterns of landownership and estate management, allowing more parties to be a
stakeholder in the Sagalassos tableware production process. Estates could have
been of variable size, and their owner had the choice whether and to what extent
to exploit the raw clay materials on the property. This scenario also gives the
potters more breathing space, in the sense that they could choose from which
landowner to acquire clays, for how much, and for how long. Both parties had to
come to an agreement, but both had options. Possibly, as suggested in the case of
tile production, landownership and the leasing of rights to produce pottery need
not have been only private. Rents from the leasing of public lands are known to
have represented additional sources of income for the cities of Asia Minor, at
times managed by a specific public official.46
In cases where historical information is available on the ownership of clays or
lands on which clays were quarried, the evidence indicates private ownership of
the raw materials by the landlord. Raw clay materials were made available to
potters as part of work or lease contracts.47 In the same way, juridical sources

44
Degryse and Poblome (2008); Degryse et al. (2008b).
45
Degryse et al. (2003).
46
Macro (1980: 684). For Laodikeia on the Lykos, see Corsten (1997: 97–9); for Colossae: IGR
4.870.
47
Mees (2002: 253).
 The Potters of Ancient Sagalassos Revisited 391

indicate that the matrix for usufruct of clays is leasing agreements on private
landownership.48 In his study of the relationship between Roman private law and
the rural economy Dennis Kehoe highlights the imperial policies of privileging
management systems based on small-scale landholding and tenancy on imperial
estates, instead of promoting the social and economic interests of the landowning
elite. Also on private land, state policy promoted the security of land tenure as a
form of stable and long-term management, while striking a careful balance with
protecting the property rights of landowners. Although this was not at all
intended to preclude large-scale landowning by members of the social elite, the
latter mostly put management systems in place based on small-scale tenants
cultivating the estates. In Kehoe’s view, the produce of small-scale farming
represented the core of the Roman economy.49 Therefore, the general pattern
of expectancy, also in the case of the north-west section of the Çanaklı Valley,
seems to be that tenancy played a role in exploiting rural properties and the raw
clay materials these contained and that there was a variety of players in the field.
In this way, our idea of the degree to which the Sagalassos elite were involved in
Sagalassos red slip ware becomes dependent on more general issues of our
reconstruction of the Roman economy.
The Çanaklı Valley has not yet been intensively surveyed. Geo-archaeological
fieldwork located the clay source area in its north-west zone and determined an
extent of 62.5 hectares within which clays were quarried in antiquity and a
wider area of 288.86 hectares within which clays could have been quarried. The
fact that modern brick factories have been extracting substantial amounts of
clays especially in the north-west part of the valley could have destroyed the
archaeological record to some extent, including traces of the presumed farming
estates. The presence of a small, Roman imperial rock-cut necropolis at the
edge of that part of the valley where the clays were quarried, which was
originally associated with the clay-quarrying activities, is perhaps better inter-
preted as the final resting place of the farmers who lived and worked nearby.50
Indeed, the clay-quarrying activities were seasonal at best and did not need
to result in permanent occupation of the area, as the destination of the raw
material was Sagalassos. The farmers, on the other hand, may have considered
the community necropolis as the translation of their permanent link with these
lands. The fact that, although a new Potters’ Quarter was laid out with the early
Roman imperial initiation of Sagalassos red slip ware and that this ware
uniquely used north-west Çanaklı clays for its fabric, there was no intention
to locate the potting infrastructure in this valley, is a further indication that
the potters considered themselves part of the urban community of Sagalassos:
they are likely to have been buried there and not in the north-west Çanaklı
necropolis.

48 49
Wieling (2000: 10-b, 17-b). Kehoe (2007a).
50
For the original interpretation, see Waelkens et al. (2000: 199).
392 Jeroen Poblome

WHERE ARE ALL THE PEOPLE?

It is very difficult to establish the social status of the individuals involved in

pottery production. Moreover, status may have differed between owners and
tenants of land and facilities, workshop managers and craftsmen, while,
depending on scale, these roles could have been combined in one person. In
so far as the evidence of stamps on Italian sigillata is relevant, variation in the
practice and meaning of stamping varied widely between different workshops,
combining slaves, freedmen, and freeborn.51 What is perhaps more important
is how these different social backgrounds were typically combined within
familia or extended households as basic training, production, business, and
juridical units, allowing a fairly flexible, mutually dependent, and to some
extent sustainable management strategy.52 Along with the local/regional scale
of production, the familia as typical operational matrix represented the hall-
marks of the artisanate in Roman antiquity. To be sure, management policies
based on social dependency were restrictive in the sense that genuine social
advancement for the not already privileged was very limited, but this embed-
ded condition was buffered by the fact that investment and initiative could
be shouldered by a socially diverse, yet networked unit, granting relative
operational independence.
The scale at which pottery workshops operated was in balance with this
social context. Throughout the empire, successful tableware brands were
produced by clusters of fairly autonomous workshops, generally quite modest
in their layout. As attested in the case of the transition from the Hellenistic to
the Roman imperial Potters’ Quarter at Sagalassos, some growth was possible,
but this was generally less dependent on direct investment by upper classes in
production infrastructure—they kept their fortunes mostly tied to the land—
and more on how the elite created favourable conditions in the urban econ-
omy in order to market the surplus from their landed estates, as well as the
level of income in the hands of small landholders and tenants, creating oppor-
tunities for enterprising, economically independent (but possibly socially tied)
craftsmen.53
Successful types of tableware did not generate economies of scale or manu-
factories; rather the basic and economically independent production unit of the
workshop was multiplied. When opportunity arose, loosely integrated networks
of artisans and workshops were created based on subcontracting arrangements,
possibly formulated in terms of locatio-conductio.54 Third parties, such as

51
Oxé, Comfort, and Kenrick (2000: 15–24).
52
Frier and Kehoe (2007: 133). For the juridical framework of work in familia context, see
Drexhage, Konen, and Ruffing (2002: 105–7).
53
Kehoe (2007b: 559–66).
54
Drexhage, Konen, and Ruffing (2002: 107–12); Wieling (2000).
 The Potters of Ancient Sagalassos Revisited 393

members of the social elite, could also issue contracts in these terms, possibly in
conjunction with providing access for tenants to raw materials and workshop
infrastructure.
In case sufficient critical mass in craft production was reached, collegia could
be founded. These voluntary associations of professionals traditionally offered
opportunities for sociability and conviviality, communal religious celebrations, a
degree of social competition, and visibility in the local community.55 In New
Institutional Economics terms, collegia could also represent a so-called private-
order enforcement network, representing an alternative organizational form
next to the market and the firm. The entrepreneurial members preferred to
close contracts for goods and services within such networks, not because this was
necessarily cheaper—on the contrary, network membership will have represent-
ed some additional cost—but mainly because the concerns for maintaining
individual professional reputation and mutual respect proved a guarantee for
the execution of the contract, in accordance with the clauses of good faith typical
for locatio-conductio type of agreements.56 Whether collegia worked exactly
according to the logic of private-order enforcement networks is impossible to
prove, but professional associations have sufficient social functions, allowing a
good level of acquaintance between their members.

H O W DI D T H I S W O RK O N T HE GR O U N D ?

Research into the Potters’ Quarter of Sagalassos is a testimony to the

research history of the Sagalassos Archaeological Research Project, as well as,
in some respects, to the evolution of archaeological fieldwork on classical
Mediterranean sites in general. While excavations were intermittent, each
time in answer to a different set of research questions forming part of the
wider interdisciplinary research agenda of the project, the introduction of
large-scale geophysical research around the turn of the millennium laid the
foundations for a more complete understanding of the area.57 By 2013 the
excavations had covered about 3 per cent of the study area. Geophysical
prospecting extended over almost the entire area (Fig. 16.6). Between the first
and the sixth centuries AD, the plateau to the east of the local theatre can be
defined as the eastern suburbs of Sagalassos, referring to how, besides the
making of pottery tablewares, other activities also made it into a busy place.58
The second most commonly documented features were the burial grounds,
gardens, and monuments of the east necropolis, located mainly in the north and

55 56
Van Nijf (1997); Zimmermann (2002). Hawkins (2012).
57
Martens et al. (2012).
58
Excavation and Ph.D. research results by Johan Claeys, University of Leuven.
394 Jeroen Poblome

Fig. 16.6. The results of geophysical prospection in the eastern suburbs. Image: Sagalassos
Archaeological Research Project.

east zones of the quarter.59 These slopes were the most visible to those entering
the eastern suburbs from the east access road into town. The church that was
discovered in the north-east section possibly continued the funerary and com-
munal functions into Early Byzantine times.60 A street network accessed the
various suburban parts, often combined with water infrastructure. Aqueducts
provided fresh water to the centre of Sagalassos and smaller channels and
pipelines serviced the suburbs. In the south-east area a couple of small-scale
limestone quarries were located, and, as mentioned, near the south-east opening
of the so-called central depression of the eastern suburbs ophiolithic clays were
extracted.61 The latter activity stopped in the course of the first century AD, after
which the quarry pits were backfilled and the terrain re-purposed as part of the
necropolis. A core-drilling programme provided indications for (communal?)
waste-dumping practices in the central depression.62 The general excavation
and survey programme at Sagalassos has provided indications of other craft
activities, such as fulling, metal-working, glass-making, and bone-cutting.63

59 60
Köse (2005). Fieldwork coordinated by Femke Martens in 2007. Unpublished.
61 62
Degryse et al. (2008c). Degryse et al. (2003).
63
De Cupere, van Neer, and Lentacker (1993); Lauwers et al. (2005).
 The Potters of Ancient Sagalassos Revisited 395

Although some of these are presumed to have been located in the eastern
suburbs, by 2013 no actual traces had been discovered here.
From the geophysical overview, twenty-five workshop areas can be defined
and eighty-nine kilns. After the excavations, another nine workshops and
eighteen kilns could be added to these totals, with attested average life spans
of about two centuries. These are conservative estimates. The more northerly
reaches of the quarter have not yet been systematically covered by the geo-
physics team, mainly because of the nature of the terrain, which is partly
covered in mountain screes. At Site F in 2011, however, in a location beyond
the geophysical overview, new workshop remains were located, including a badly
preserved kiln.64 It will always remain difficult to present a final complete picture
of the quarter, as excavations have also indicated that kiln remains and walls of
workshops can be blurred for geophysics if they lie deeper in the stratigraphy
or are badly preserved or partly overbuilt by later structures. Moreover, the
geophysical results are non-chronological, revealing no aspects of chronological
allocation or evolution through time. Furthermore, by workshop areas I mean
something different from workshops pur sang. The latter can be fully established
only upon excavation. ‘Workshop area’ represents a more general site definition,
encompassing walls and kilns, within which one or more workshops could have
been operational. The workshop area of the so-called Coroplast Workshops
(Fig. 16.7), for example, revealed the existence of at least six workshop units
upon excavation, while the East Slope Workshops (see Fig. 16.5) are actually two
consecutive ateliers.65 Although no indications have yet been found, it also cannot
be excluded that some craft activities other than tableware production were being
organized in these workshop areas. Be that as it may be, the workshop areas seem
to make up their own clustered neighbourhood on the centrally located, middle
terraces of the eastern suburbs. The question arises as to how far the move away
from the Hellenistic workshops and the dedicated function of the artisanal part of
the eastern suburbs reflected planning and management coordinated by the city
council. Indeed, an artisanal cluster of between 3.5 and 4 hectares can hardly have
grown spontaneously. Lack of relevant comparative material makes it difficult to
judge just how important the quarter could have been in the urban landscape,
society, and economy. The estimation of the Potters’ Quarter compares to the
25.2 hectares reconstructed for residential purposes, the 37.5 hectares for the
urbanized area of Sagalassos, and the c.90 hectares that included the necropolis
and suburban functions. Based on these figures, the local population has been
estimated at less than 10,000—possibly between 2,500 and 3,750 inhabitants.66
Although the early imperial relocation of the tableware craft away from its
Hellenistic quarter and into this new neighbourhood has already been attested
in two excavated workshops and associated pottery waste dumps, it does not

64
Excavations at Site F 2011 were supervised by Johan Claeys. Unpublished.
65 66
Murphy and Poblome (2011). Willet (2012: 182–3).
396 Jeroen Poblome

Fig. 16.7. The so-called Coroplast workshops, fourth to sixth centuries AD. Plan:
Sagalassos Archaeological Research Project.

seem to be the case that the new Potters’ Quarter was packed into all its
corners at once, but that the scenario was one of gradual yet steady growth. As
excavating all the workshops is impossible and unnecessary, the evolution in
production can best be approached from output curves of Sagalassos red slip
ware. In a previous paper, output data were reconstructed based on the urban
survey results and a range of selected excavated deposits from the wider urban
excavation programme.67 Output and presumed amount of workshops peaked
between Flavian and Severan times and again between the Leonid dynasty and
the reign of Justinian. Although the estimated output is lower in the interven-
ing period, there are no typo-chronological indications that production of
Sagalassos red slip ware was interrupted in late Roman times. That being the
case, the local weather conditions, especially in winter, need to be taken into
account, as these made year-round craft production impossible. These condi-
tions must have affected general patterns in demand for goods and services
too, with life running in different cycles in winter. The climate conditions,
together with the social context of production already discussed, imply that
large-scale, vertically integrated manufactories with permanently employed

67
Poblome et al. (2013c).
 The Potters of Ancient Sagalassos Revisited 397

staff were not a feasible management scenario, and that small-sized, flexible
solutions were favoured instead.
This is also the type of workshop that has been documented in the excavations.
The ateliers typically contain one or two kilns each, and a couple of rooms and
spaces with specific functions in the production process (Fig. 16.8). The attest-
ation of sometimes three potter’s wheel installations in one workshop and the
finding of sets of tools with engraved names within another workshop setting
indicate that, generally, a couple of craftsmen were involved in each separate
workshop.68 On the ground floor, no signs that the potters were also living in
these ateliers have been found, and none of the workshops studied seems to have
had an upper floor, all of which implies that the craftsmen rented or owned a
small house somewhere else in town. Private workshop ownership within the
familia as a typical operational nucleus seems plausible.69 As with anything in the
real world, things will have been more complicated in this respect too. The fact
that architectural changes in layout and function of the different workshops
involved in making Early Byzantine mould-made products seem to be happening
more or less simultaneously could possibly hint at a third-party stakeholder in the
production process. Workshops also seem to specialize and restrict their output to

Fig. 16.8. 2013 aerial view of the so-called Coroplast workshops, fourth to sixth
centuries AD. Photo: Sagalassos Archaeological Research Project.

68
Murphy and Poblome (2012).
69
This depends on more general patterns of reconstruction of the ancient economy and how
much non-elite members of society were allowed to breathe. Cf. Scheidel and Friesen (2009).
398 Jeroen Poblome

a specific set of types of the general typology of Sagalassos red slip ware, creating
options for collaboration on orders or contracts.70
So far, no collegium of potters has been attested at Sagalassos, epigraphically or
otherwise. The dedication IGR 3.360 was set up in the later second–third
centuries AD in the area of the south Gate into Sagalassos by ἡ συντεχνία τῶν
βαφέων to support the statue of its local patron and imperial high priest Aelius
Quintus Claudius Philippianus Varus in honour of his funding gladiatorial
games. This is the only local reference to a craftsmen association, but at least it
indicates the phenomenon was not unknown at Sagalassos.71 It is, however,
possible that, if there was a role for public and/or sacred landownership in the
section of the Çanaklı Valley where the clays for the local tableware industry were
dug, a religious association emerged, instead of an association of potters.72
The preliminary interpretation of an unpublished structure excavated since
2011 at the west edge of the artisanal zone within the eastern suburbs could
provide further indications for il fenomeno associativo (Fig. 16.9).73 The

Fig. 16.9. Aerial view of the presumed early Roman imperial Vereinshaus. Photo:
Sagalassos Archaeological Research Project.

70
Excavation and Ph.D. research results by Elizabeth Murphy, Brown University.
71
See Brandt (1992: 137–9) for other craftsmen and professional associations attested in the
ancient regions of Pamphylia and Pisidia, with potters epigraphically represented at Baris.
72
On sacred landownership, see Dignas (2005) and Zuiderhoek (2009: 37–52).
73
Excavations at Site PQ2 2011, 2012, 2013, and 2014 were supervised by Johan Claeys, Sven
Van Haelst, and Peter Talloen. See also <http://journal.antiquity.ac.uk/projgall/poblome342>
(accessed 8 July 2015).
 The Potters of Ancient Sagalassos Revisited 399

original phase of the building was erected around the middle of the first
century AD. It was rectangular in ground plan (10.4  9 m), with its entrance
from the street on the north side. Against the inside of the back wall a water
feature was built (1.7  2.8 m). Around 100 AD, the building was extended on
its south side (10.2  12.3 m) and a new entrance created in the south wall.
After a while, three smaller rooms were created in the north part of the
building, the water feature levelled, and the floor level raised. No direct
evidence related to the function of the building was found, yet its location at
the interface between the town and the artisanal quarter, its simple rectangular
plan forming one (articulated) space inside, the width and finish of its
entrances, the quality of part of its wall construction using ashlars, the
presence of the water feature, and the absence of functional indicators make
it stand out in the building corpus of the area. Moreover, during the second
half of the second century AD, a dump accumulated against the outer east wall
with concentrations of discarded sets of tableware mixed with particular parts
of some goats/sheep, pigs, and mainly cattle, which would have provided low-
quality meat. The waste is indicative of a soup kitchen, and its association with
the building under excavation hints at practices of communal dining and
sociability. That the building was designed to receive groups of people became
clear in 2014, when in three of these rooms enormous quantities of archaeo-
logical material were found, as though they were left only yesterday; in fact,
they date to around AD 270. The finds include objects made of materials such
as metal and carved bone, as well as glass flasks and drinking cups, oil lamps,
and large quantities of ceramic tableware. Many of the latter, especially bowls
and dishes, were preserved intact and discovered upside down, containing
original food remains. These finds seem to relate to ancient communal dining
practices involving dozens of participants. The tableware and food remains
were not tidied away, but rather thrown along the walls of the building and left
behind; this was a one-off event, not resulting from accumulation and seem-
ingly associated with the end of the building as such. Although the excavation
needs to be completed and preferably an inscription found, at this stage
I should like to propose, as a working hypothesis, that the group of people
that made use of this building could have been an association, albeit of an
unidentified nature.74
The final identification of this building should be helpful, but, in the
meantime, the collected evidence on the potters of Sagalassos is already enough
to further sustain recognition of and research into the ancient middle classes.75
Social and economic historians are warming to the idea not only that the main
sector of the Roman economy, agriculture, could not depend solely on elite

74
Published parallels from Asia Minor are unknown to me. In so far as the situation in
contemporary Italy is helpful, see Bollmann (1998).
75
Mayer (2012).
400 Jeroen Poblome

landowners, requiring the crucial involvement of small-scale farmers and

tenants to develop successfully, but also that towns, that other playground of
elite investment in representative projects, could not come to life without the
participation of proficient craftsmen, merchants, professionals, and entrepre-
neurs. The archaeology of Sagalassos will, it is hoped, further develop in this
respect. In the meantime, this community should most of all be considered as
one of many other such provincial towns in the Roman Empire, fostering active
groups of artisans contributing to all aspects of urban life in antiquity.

ACKNOWLEDGEMENTS

The research for this paper was supported by the Belgian Programme on
Interuniversity Poles of Attraction (IAP 07/09), the Research Fund of the
University of Leuven (GOA 13/04), and Projects G.0562.11 and G. 0637.15 of
the Research Foundation Flanders (FWO). The archaeometric analyses were
carried out at the Leuven Centre for Archaeological Sciences. I wish to
acknowledge the most excellent excavation qualities of Elizabeth Murphy,
Johan Claeys, Sven Van Haelst, and Peter Talloen, who shared the adventure
of working in the eastern suburbs. Arjan Zuiderhoek most kindly commented
on the early stages of this chapter’s composition.

REFERENCES

Armstrong, P. (2012). ‘The Survey Pottery: Hellenistic and Later’, in J. J. Coulton (ed.),
The Balboura Survey and Settlement in Highland Southwest Anatolia, 2. The Balboura
Survey: Detailed Studies and Catalogues (British Institute at Ankara Monograph, 43).
London, 31–82.
Arnold, D. E. (1985). Ceramic Theory and Cultural Process. Cambridge.
Bintliff, J. (2011). ‘The Death Of Archaeological Theory?’, in J. Bintliff and M. Pearce
(eds), The Death of Archaeological Theory? Oxford, 7–22.
Bollmann, B. (1998). Römische Vereinshäuser: Untersuchungen zu den Scholae der
römischen Berufs-, Kult-, und Augustalen-Kollegien in Italien. Mainz.
Braekmans, D., Degryse, P., Poblome, J., Neyt, B., Vyncke, K., and Waelkens, M.
(2011). ‘Understanding Ceramic Variability: An Archaeometrical Interpretation of
the Classical and Hellenistic Ceramics at Düzen Tepe and Sagalassos (Southwest
Turkey)’, Journal of Archaeological Science 38: 2101–15.
Brandt, H. (1992). Gesellschaft und Wirtschaft Pamphyliens und Pisidiens im Altertum
(Asia Minor Studien, 7). Bonn.
Corsten, T. (ed.) (1997). Die Inschriften von Laodikeia am Lykos (Inschriften
Griechischer Städte aus Kleinasien, 49). Bonn.
De Cupere, B., van Neer, W., and Lentacker, A. (1993). ‘Some Aspects of the Bone-
Working Industry in Roman Sagalassos’, in Waelkens and Poblome (1993), 269–78.
 The Potters of Ancient Sagalassos Revisited 401
Degryse, P., and Poblome, J. (2008). ‘Clays for Mass Production of Table and Common
Ware, Amphorae and Architectural Ceramics at Sagalassos’, in Degryse and Waelkens
(2008), 231–54.
Degryse, P., and Waelkens, M. (eds) (2008). Geo- and Bio-Archaeology at Sagalassos
and in its Territory (Sagalassos, 6). Leuven.
Degryse, P., Poblome, J., Donners, K., Deckers, J., and Waelkens, M. (2003). ‘Geoarch-
aeological Investigations of the “Potters’ Quarter” at Sagalassos, Southwest Turkey’,
Geoarchaeology 18: 255–81.
Degryse, P., Muchez, P., Sintubin, M., Clijsters, A., Viaene, W., Dederen, M., Schrooten, P.,
and Waelkens, M. (2008a). ‘The Geology of the Area around the Ancient City of
Sagalassos’, in Degryse and Waelkens (2008), 17–24.
Degryse, P., Poblome, J., Viaene, W., Kucha, H., Ottenburgs, R., Waelkens, M., and
Naud, J. (2008b). ‘Provenancing the Slip of Sagalassos Red Slip Ware’, in Degryse
and Waelkens (2008), 255–60.
Degryse, P., Heldal, T., Bloxam, E., Storemyr, P., Waelkens, M., and Muchez, P. (2008c).
‘The Sagalassos Quarry Landscape: Bringing Quarries in Context’, in Degryse and
Waelkens (2008), 261–90.
Dignas, B. (2002). Economy of the Sacred in Hellenistic and Roman Asia Minor.
Oxford.
Dignas, B. (2005). ‘Sacred Revenues in Roman Hands: The Economic Dimension of
Sanctuaries in Western Asia Minor’, in S. Mitchell and C. Katsari (eds), Patterns in
the Economy of Roman Asia Minor. Swansea, 207–24.
Drexhage, H.-J., Konen, H., and Ruffing, K. (2002). Die Wirtschaft des Römischen
Reiches (1.-3. Jahrhundert): Eine Einführung. Berlin.
Frier, B. W., and Kehoe, D. P. (2007). ‘Law and Economic Institutions’, in Scheidel
et al. (2007), 113–43.
Fülle, G. (1997). ‘The Internal Organization of the Arretine Terra Sigillata Industry:
Problems of Evidence and Interpretation’, JRS 87: 111–55.
Hartley, B. R., and Dickinson, B. M. (2008–12). Names on Terra Sigillata: An Index of
Makers’ Stamps and Signatures on Gallo-Roman Terra Sigillata (Samian Ware), vols
1–8. London.
Hawkins, C. (2012). ‘Manufacturing’, in W. Scheidel (ed.), The Cambridge Companion
to the Roman Economy. Cambridge, 175–94.
Jones, A. H. M. (1940). The Greek City from Alexander to Justinian. Oxford.
Kehoe, D. P. (2007a). Law and the Rural Economy in the Roman Empire. Ann Arbor, MI.
Kehoe, D. P. (2007b). ‘The Early Roman Empire: Production’, in Scheidel, Morris, and
Saller (2007), 543–69.
Köse, V. (2005). Nekropolen und Grabdenkmäler von Sagalassos in Pisidien in helle-
nistischer und römischer Zeit (Studies in Eastern Mediterranean Archaeology, 7).
Turnhout.
Ladstätter, S. (2007). ‘Mode oder politisches Manifest? Überlegungen zur Übernahme
römischen Formenguts in die frühkaiserzeitliche Keramik von Ephesos’, in Meyer
(2007), 203–20.
Lauwers, V., Degryse, P., Poblome, J., and Waelkens, M. (2005). ‘Le Verre de Saga-
lassos: De nouvelles preuves d’une production locale de verre’, Bulletin de l’associa-
tion française pour l’archéologie du verre 26–9.
402 Jeroen Poblome
Loots, L., Waelkens, M., Clarysse, W., Poblome, J., and Hübner, G. (2000). ‘A Catalogue of
the Tile Stamps found at Sagalassos’, in Waelkens and Loots (2000), 685–96.
Lucas, G. (2012). Understanding the Archaeological Record. Cambridge.
Lund, J. (2002). ‘The Ontogenesis of Cypriot Sigillata’, in A. Rathje, M. Nielsen, and
B. Bundgaard Rasmussen (eds), Pots for the Living: Pots for the Dead (Acta
Hyperborea, 9). Copenhagen, 185–223.
Macro, A. D. (1980). ‘The Cities of Asia Minor under the Roman Imperium’, in
H. Temporini and W. Haase (eds), Aufstieg und Niedergang der Römischen Welt
II, 7.2. Berlin, 658–97.
Martens, F., Mušič, B., Poblome, J., and Waelkens, M. (2012). ‘The Integrated Urban
Survey at Sagalassos’, in F. Vermeulen, G.-J. Burgers, S. Keay, and C. Corsi (eds),
Urban Landscape Survey in Italy and the Mediterranean. Oxford, 84–93.
Mayer, E. (2012). The Ancient Middle Classes: Urban Life in the Roman Empire 100
BC–250 AD. Cambridge, MA.
Mees, A. W. (2002). Organisationsformen römischer Töpfer-Manufakturen am Beispiel
von Arezzo und Rheinzabern unter Berücksichtigung von Papyri, Inschriften und
Rechtsquellen (Römisch-Germanisches Zentralmuseum. Forschungsinstitut für
Vor- und Frühgeschichte Monographien, 52). Mainz.
Meyer, M. (ed.) (2007). Neue Zeiten–neue Sitten: Zu Rezeption und Integration
römischen und italischen Kulturguts in Kleinasien (Wiener Forschungen zur Arch-
äologie, 12). Vienna.
Mitchell, S. (1993). Anatolia. Land, Men, and Gods in Asia Minor, 1. The Celts and the
Impact of Roman Rule. Oxford.
Mitchell, S., and Waelkens, M. (1988). ‘Cremna and Sagalassus 1987’, Anatolian Studies
38: 53–64.
Muchez, P., Lens, S., Degryse, P., Callebaut, K., Dederen, M., Hertogen, J., Joachimski, M.,
Keppens, E., Ottenburgs, R., Schroyen, K., and Waelkens, M. (2008). ‘Petrography,
Mineralogy and Geochemistry of the Rocks in the Area of the Archaeological Site of
Sagalassos’, in Degryse and Waelkens (2008), 25–52.
Murphy, E. and Poblome, J. (2011). ‘Producing Pottery vs Producing Models: Interpreting
Workshop Organization at the Potters’ Quarter of Sagalassos’, in M. L. Lawall and
J. Lund (eds), Pottery in the Archaeological Record: A View from the Greek World (Gösta
Enbom Monographs, 1). Aarhus, 30–6.
Murphy, E., and Poblome, J. (2012). ‘Technical and Social Considerations of
Tools from Roman-Period Ceramic Workshops at Sagalassos (Southwest Tur-
key): Not Just Tools of the Trade?’, Journal of Mediterranean Archaeology 25:
69–89.
Neyt, B. (2012). ‘Mineralogical Resources for Ceramic Production in the Territory of
Sagalassos: A Geochemical, Mineralogical And Petrographical Study’. Unpublished
Ph.D. thesis, University of Leuven.
Neyt, B., Braekmans, D., Poblome, J., Elsen, J., Waelkens, M., and Degryse, P. (2012).
‘Long-term Clay Raw Material Selection and Use in the Region of Classical/Hellenistic
to Early Byzantine Sagalassos (South-West Turkey)’, Journal of Archaeological Science
39: 1296–305.
Nijf, O. M. van (1997). The Civic World of Professional Associations in the Roman East
(Dutch Monographs on Ancient History and Archaeology, 17). Amsterdam.
 The Potters of Ancient Sagalassos Revisited 403
Oxé, A., Comfort, H., and Kenrick, P. (2000). Corpus vasorum arretinorum: A Catalogue of
the Signatures, Shapes and Chronology of Italian Aigillata (Abhandlungen zur Vor- und
Frühgeschichte, zur Klassischen und Provinzial-Römischen Archäologie und zur
Geschichte des Altertums, 41). Bonn.
Paulissen, E., Poesen, J., Govers, G., and De Ploey, J. (1993). ‘The Physical Environment at
Sagalassos (Western Taurus, Turkey): A Reconnaissance Survey’, in Waelkens and
Poblome (1993), 229–47.
Peacock, D. P. S. (1982). Pottery in the Roman World: An Ethnoarchaeological Approach.
London.
Poblome, J. (1996). ‘Production and Distribution of Sagalassos Red Slip Ware: A Dialogue
with the Roman Economy’, in M. Herfort-Koch, U. Mandel, and U. Schädler (eds),
Hellenistische und kaiserzeitliche Keramik des östlichen Mittelmeergebietes. Frankfurt
am Main, 75–103.
Poblome, J. (1999). Sagalassos Red Slip Ware: Typology and Chronology (Studies in
Eastern Mediterranean Archaeology, 2). Turnhout.
Poblome, J. (2006). ‘Made in Sagalassos: Modelling Regional Potential and Con-
straints’, in S. Menchelli and M. Paquinucci (eds), Territorio e produzioni ceramiche:
Paesaggi, economia e società in età romana (Instrumenta, 2). Pisa, 355–63.
Poblome, J., and Zelle, M. (2002). ‘The Tableware Boom. A Socio-Economic Perspec-
tive from Western Asia Minor’, in C. Berns, H. Van Hesberg, L. Vandeput, and
M. Waelkens (eds), Patris und Imperium: Kulturelle und politische Identität in den
Städten der römischen Provinzen Kleinasiens in der frühen Kaiserzeit. Leuven,
275–87.
Poblome, J., Bes, P., and Lauwers, V. (2007). ‘Winning Hearts, Minds and Stomachs?
Artefactual or Artificial Evidence for Romanization’, in Meyer (2007), 221–32.
Poblome, J., Malfitana, D., and Lund, J. (2011). ‘Investing in Roman Antiquity.
Modelling Moderate Growth’, FACTA: A Journal of Roman Material Culture
Studies 5: 9–13.
Poblome, J., Bounegru, O., Degryse, P., Viaene, W., Waelkens, M., and Erdemgil, S.
(2001). ‘The Sigillata Manufactories of Pergamon and Sagalassos’, JRA 14: 143–65.
Poblome, J., Degryse, P., Viaene, W., Ottenburgs, R., Waelkens, M., Degeest, R., and
Naud, J. (2002a). ‘The Concept of a Pottery Production Centre: An Archaeometrical
Contribution from Ancient Sagalassos’, Journal of Archaeological Science 29:
873–82.
Poblome, J., Degryse, P., Viaene, W., and Waelkens, M. (2002b). ‘An Augustan Pottery
Workshop at Sagalassos’, in V. Kilikoglou, A. Hein, and Y. Maniatis (eds), Modern
Trends in Scientific Studies on Ancient Ceramics (BAR International Series, 1011).
Oxford, 335–41.
Poblome, J., Braekmans, D., Waelkens, M., Fırat, N., Vanhaverbeke, H., Kaptijn, E.,
Martens, F., Vyncke, K., Willet, R., and Degryse, P. (2013a). ‘How did Sagalassos
Come to be? A Ceramological Survey’, in M. Tekocak (ed.), Studies in Honour of
K. Levent Zoroğlu. Antalya, 527–40.
Poblome, J., Braekmans, D., Mušič, B., Van Der Enden, M., Neyt, B., De Graeve, B.,
and Degryse, P. (2013b). ‘A Pottery Kiln underneath the Odeon of Ancient Saga-
lassos, South-West Turkey: The Excavation Results, the Table Wares and their
Archaeometrical Analysis’, in N. Fenn and C. Römer-Strehl (eds), Networks in the
404 Jeroen Poblome
Hellenistic World according to the Pottery in the Eastern Mediterranean and Beyond.
(BAR International Series, 2539). Oxford, 193–204.
Poblome, J., Willet, R., Fırat, N., Martens, F., and Bes, P. (2013c). ‘Tinkering with
Urban Survey Data: How Many Sagalassos-es do we Have?’, in P. Johnson and
M. Millett (eds), Archaeological Survey and the City (University of Cambridge
Museum of Classical Archaeology Monograph, 2). Oxford, 146–74.
Scheidel, W., and Friesen, S. J. (2009). ‘The Size of the Economy and the Distribution
of Income in the Roman Empire’, JRS 99: 61–91.
Scheidel, W., Morris I., and Saller, R. (eds) (2007). The Cambridge Economic History of
the Greco-Roman World. Cambridge.
Six, S. (2004). ‘Holocene Geomorphological Evolution of the Territory of Sagalassos.
Contribution to the Palaeo-Environmental Reconstruction of Southwest Turkey’.
Unpublished Ph.D. thesis, University of Leuven.
Talloen, P., Daems, D., Beaujean, B., and Poblome, J. (2015). ‘The 2014 control
excavations on the Upper Agora of Sagalassos’, ANMtD. News of Archaeology
from Anatolia’s Mediterranean Areas 13: 99–107.
Vanhaverbeke, H., Waelkens, M., Vyncke, K., De Laet, V., Aydal, S., Mušič, B., De
Cupere, B., Poblome, J., Braekmans, D., Degryse, P., Marinova, E., Verstraeten, G.,
van Neer, W., Šlapšak, B., Medarič, I., Ekinci, H. A., and Erbay, M. O. (2010).
‘Pisidian Culture? The Classical–Hellenistic Site at Düzen Tepe near Sagalassus
(Southwest Turkey)’, Anatolian Studies 60: 105–28.
Vermoere, M. (2004). Holocene Vegetation History in the Territory of Sagalassos
(Southwest Turkey): A Palynological Approach (Studies in Eastern Mediterranean
Archaeology, 6). Turnhout.
Vermoere, M., Six, S., Poblome, J., Degryse, P., Paulissen, E., Waelkens, M., and Smets,
E. (2003). ‘Pollen Sequences from the City of Sagalassos (Pisidia, southwest Turkey)’,
Anatolian Studies 53: 161–73.
Waelkens, M., and Loots, L. (eds) (2000). Report on the Survey and Excavation
Campaigns of 1996 and 1997 (Sagalassos, 5). Leuven.
Waelkens, M., and Poblome, J. (eds) (1993). Report on the Third Excavation Campaign
of 1992 (Sagalassos, 2). Leuven.
Waelkens, M., Vanhaverbeke, H., Paulissen, E., Poblome, J., Reyniers, J., Viaene, W.,
Deckers, J., De Cupere, B., van Neer, W., Ekinci, H. A., and Erbay, M. O. (2000). ‘The
1996 and 1997 Survey Seasons at Sagalassos’, in Waelkens and Loots (2000), 17–216.
Wallace-Hadrill, A. (2008). Rome’s Cultural Revolution. Cambridge.
Wieling, H. (2000). ‘Vertragsgestaltung der römischen Keramikproduktion’, in
K. Strobel (ed.), Forschungen zur römischen Keramikindustrie: Produktions-, Rechts-
und Distributionsstrukturen (Trierer Historische Forschungen, 42). Mainz, 9–22.
Willet, R. (2012). ‘Red Slipped Complexity: The Socio-Cultural Context of the Concept
and Use of Tableware in the Roman East’. Unpublished Ph.D. thesis, University of
Leuven.
Zimmermann, C. (2002). Handwerkvereine im griechischen Osten des Imperium
Romanum (Römisch-Germanisches Zentralmuseum: Forschungsinstitut für Vor- und
Frühgeschichte Monographien, 57). Mainz.
Zuiderhoek, A. (2009). The Politics of Munificence in the Roman Empire: Citizens,
Elites and Benefactors in Asia Minor. Cambridge.
 Index

agency 5, 11–12, 16, 109, 214, 222–50 comparison goods 308, 311, 321
agriculture 26, 38, 62, 77, 80–1, 215, 224 competition 8, 116, 119–23, 139, 307–8, 310,
244, 399 342, 344
Alexandria 102, 105–7, 194, 261 consumer city debate 7, 34–9, 79, 90
amphorae 63–4, 80, 86, 100–1, 103, 108–9, consumers 8–9, 15, 97, 118, 121–2, 139–50,
111, 117, 121, 227, 232–33, 236, 239, 243 207, 344
alum 86 consumerism 139–146
Adriatic 64 convenience goods 13, 308, 313, 344–5
garum 236 Corpus Inscriptionum Latinarum 27–8, 200
oil 227, 233, 239, 243
reused 166, 173 D’Arms, John 39, 108, 225, 246–7, 254
wine 64, 101, 111, 232–3 didaskalikai 183–7
Amsterdam 18th century 308–9, 321 disciplina 185–7
Annales School 78 De Robertis, Francesco Maria 57–9, 65–6
apprenticeship 10, 115–16, 139, 183–97, 324, dolia 100, 157, 163
343–4 Düzen Tepe 381–5
Aquincum 13–4, 352–72 dyeing 126, 213, 280, 313–15, 335
Arles 11–12, 192, 254–68 workshops 82–3, 340–1
ars sordid 62, 89 dyers 82, 121, 126, 281–294, 335–7
Asia Minor 8, 42, 66, 119, 127, 195–6, 278–94,
322, 390 Early modern world 26, 104, 115–16, 123–4,
autonomy 12, 15, 255, 392; see also 147, 196, 244
independence. economies of scale 139, 343
Egypt, Roman 13, 32, 42, 64, 120–2, 125,
bakers66, 81, 88, 123–4, 282, 322, 335, 133–5, 139–40, 184–9, 193, 242, 322–5,
340, 347 334–40
bakeries 9, 154–76, 308–9, 313 elite 122, 133, 184, 192–3, 233, 246–9,
Baetica, Hispania 67, 100–1, 233, 237, 258–60, 286, 288
239–40, 243, 262 consumption 15, 36, 145
Baetulo 100–1 control 302–3, 310, 312–13, 324–6
Bianchi Bandinelli, Ranuccio 64 houses 189, 230, 356
Birmingham Jewellery Quarter 305–7 involvement in commerce 36, 38, 88,
bone-working 365–70, 394 107–9, 189, 224, 228–9, 243, 392
bread 8–9, 15, 124–5, 155, 164–5, 281–2, 335. attitude to commerce 1–214, 254–5
Britain, Roman 16, 37–8, 100, 134, 140, 235 (land)ownership 13, 15, 33–5, 79, 111, 392
Bücher, Karl 29–34 embeddedness 3, 302, 392
Bücher-Meyer controversy 30 emporia 102–3, 107–8
butcher 23, 216, 347, 369 Ephesos 105–6, 119, 190, 279, 283–4, 287,
289, 326, 340
canonization 26–9, 44 Eurysaces (baker) 15, 42, 124–5, 160–1
Cicero 25, 81, 109, 121, 184–5, 188–9, 230,
237, 245, 259, 323, 346 family 10–1, 81, 116, 139, 185–91, 196, 206,
De off. I.150–1: 2, 62, 264 211, 213, 218, 227–49, 306, 324
chaîne opératoire 80, 154, 176 felt-making 215, 282
clustering 13–14, 301–27, 334–48, 392, 395 fines 283–5
cognomina 227–8, 233, 235, 258–9, 263, 266 Finley, Sir Moses I. 2–3, 7, 34–7, 43–4, 61,
collegia 2, 24, 28, 33, 41–2, 58, 65–6, 244–5, 223–5, 254–5
311, 358, 393; see also professional footwear 8–9, 117, 132–50,
associations. formalism 320
406 Index
Frank, Tenney 30–36, 43–4, 78 manumission 109, 203–4, 211, 230–1,
freedmen 11, 41–2, 109–10, 138–9, 189–90, 249–50, 260, 344
202–3, 211, 217, 222–50, 254–5, 263, 267 maritime trade 8, 62, 66, 97–111, 229, 244,
frumentarii 61, 98, 228, 262, 322, 334, 258, 265
345–6 market (institution) 5, 8–9, 13, 31, 33, 60, 97,
fullers 3, 88, 121, 217, 282, 311–16, 326, 104, 110–11, 117–23, 125, 127, 138,
335–6, 338, 340–1, 343–5 207–8, 232–9, 303, 306, 309, 345, 348,
fullonicae 340 384, 393
funerary monuments 11, 139, 185, 234, 258, market (place) 133, 148, 314, 317, 322, 325;
263, 279–81, 289–94 see also macellum
furnaces 310, 318–19 marketing 108, 147–8, 232, 245, 383, 388
Marshall, Alfred 301, 342
gender 41, 67, 147, 200–19 Marx, Karl 26, 29
glass-blowing 82, 97, 121, 320, 345, 347, 394 Marxist scholarship 17, 56, 59–61, 64
grain trade 98, 228–9, 240, 245 Medieval Europe 13, 26, 29, 39, 57, 88, 123–4,
guilds, Medieval 13, 302, 348 146–7, 222, 302, 342–3, 347–8
memory management 11, 289–93
heating technology 164–8 mercatores 11, 98, 106, 228–9, 262–5
Hippo Regius 228, 240 Meyer, Eduard 6–7, 29–31, 43
Hierapolis 11–12, 126–7, 278–94 metal-working 78–80, 89, 359, 361, 364–5,
horizontal specialization 117, 124–7 371, 394
middle class 110, 184, 194, 399
iconography 11, 15, 41–2, 200–6 middlemen 102, 104, 243
independence 11, 39, 109, 243–4, 255, 392; mill 155–60, 169–75; see also bakeries,
see also autonomy water-mill
instrumentum domesticum 63 Mommsen, Theodor 6, 23, 27–8
investment 10, 14–5, 34–5, 108–9, 136, Münstersche Beiträge zur Antiken
173–5, 222–50 Handelsgeschichte 39
Istituto Gramsci 59–61
nailsmiths 123–4
Jones, A. H. M. 33–6, 43–4, 218 naukleroi, see naviculari
navicularii 29, 66, 98, 102, 105, 223, 233,
Kellis 338–40 257–63, 268–70
kneading machine 160–4 negotiatores 8, 11, 25, 39, 98–100, 106, 223,
235, 246, 262–3
labour market 184, 196, 204, 214–5, negotiatores vinarii 98, 262
218–9, 360 negotiatores olearii, see olearii
‘lanifricariae’ 310–12, 314, 340–2 New Institutional Economics 4–5, 109, 119,
Late Antiquity 60–2, 66, 80, 188, 345 225, 388, 393
leather 9, 88–9, 132–50, 282, 346, 370; see also niche 8–9, 121–3, 127, 344, 389
tanning.
legal texts 25, 29, 229–30, 235 occupation 118
lenuncularii 256–63 occupational identity 10–1, 15, 67, 216
London, Roman 144 occupational epigraphy 27–8, 39, 58, 62–3,
locatio-conductio 392–3 116, 208, 212, 222, 234
location factors 125–7 Oikenwirtschaft 29–30, 44
Lucian of Samosata 194, 196 olearii 100, 262
luxury goods 31, 33, 82, 133, 203–7, 218, 242, Ostia 8, 98, 106–7, 138, 209–10, 215, 217,
286, 320–1, 343–4 228–9, 233, 240–1, 246–8, 260–2, 340
Lyon 259, 262–5 Piazzale delle Corporazioni 106–7, 223
oven 155–6, 164–8, 171–5
macellum 104, 321 Oxyrhynchos 118, 322, 338, 347
Aquincum 356–61
Pompeii 312–13 Panopolis 338
Madrague de Giens (wreck) 101 papyri 13, 32, 35–6, 39, 64, 118, 123, 125, 187,
manufactory 80, 88, 388, 392, 396 193, 322, 325, 334–48
 Index 407
patrocinium 189–93 Silchester 318–20
peculium 230 silk trade 122, 205–6
perfume production 83–4, 88, 203–4, 320, slavery 26, 60–1, 109, 115–16, 184, 190, 202,
345–6 208, 210–11, 213, 217–18, 229–31
plebeian elite 11, 255–8, 267–8 Smith, Adam 8, 25–6, 118–19, 123–4
Pleket, H. 116, 127 smiths 85, 123–4, 194, 202, 241, 281–2, 290,
Polanyi, Karl 3, 34 292, 324–5, 347
Pompeii 31, 38, 42–3, 120, 123, 217, 244, social networks 3–4, 111, 184, 189, 302, 324
325–6 social status 2–3, 12, 35–6, 108–10,
bakeries 9, 125, 154–76 116, 204, 214, 220–8, 232–3, 255,
clustering 309–14, 340–2 286, 392
excavation 26–7 specialization 8, 31, 88, 97–111, 115–127,
textile industry 36–7, 88, 215 138–9, 197, 206–7, 236–8, 305, 323,
workshops 82–9, 367 342, 397
port cities 8, 35, 97, 108, 110, 248, 254, statio 106–8, 260, 269
262, 265 Statilii Tauri 208
pottery production 14, 27, 31, 78–9, 303, 324, Strabo 126, 280
354, 358–9, 371, 377–400 substantivism 32–6, 115, 223–4, 302
productivity 9, 123–5, 170–1, 342 suburbs 89, 385–7, 393–400
producer city 38, 127
professional associations 4, 10–12, 24, 28, 42, tanning 85–9, 136–8, 367–8
65–6, 80, 121, 138, 228–9, 244, 254–68, tabernae 1, 107, 122, 230, 233, 362
278–94, 306, 340; see also collegia Tarraconensis, Hispania 101
property 325–7, 390 taxes 29, 33, 242
Ptolemais Euergetis 121, 334–7, 343, 347 in kind 136
Puteoli 31, 106–8, 189, 204, 242, 248 tax registers 13, 334–6, 347
purple(-dyeing) 121, 203–5, 235, 280–7 technè 194–5
technology 9, 25, 77–8, 135–8
quinquatrus 4 terra sigillata 34, 323, 384; see also red slip
quinquennalis 228–9, 245 ware
Testaccio, Monte 227
rationalization 9, 90, 170–5 textile production 126, 139, 200–19, 279–82,
red slip ware 377–400 310–18, 334–49; see also Pompeii
reputation 110, 183–4, 187, 193–4, 291, 393 Thyateira 278–87, 294
retail 1, 10, 12, 15, 116, 118, 147, 307–9, 313, Timgad 45, 37–8, 314–8, 326
320–2, 359 tituli picti 121, 227–8, 233, 236, 239, 247
return cargoes 102–3, 238 toponyms 13, 121, 320–1, 345–8
Rodbertus, Johann Karl 29–30, 44–5 trading communities 8, 104–8
Rome 30, 32, 41, 86, 88, 100, 115–6, 119, 121, Treggiari, Susan 40–1, 201–4, 214
125, 138, 200–19, 222–50, 320–3, 345–6 Trimalchio 249, 259, 265
Monumentum Liviae 41, 202–4
Via Sacra 204–6, 241–2, 321 A.Umbricius Scaurus 254
Vicus Tuscus 205, 234, 321–2 unguentarii 204, 303, 321, 346; see also
Rostovtzeff, Mikhail I. 7, 30–5, 43, 254 perfume production
urban space 12–4, 42–3, 259, 301–400
Saepinum 86–7, 367 Urso charter 325
Sagalassos 377–400 utricularii 256–8, 271–2
Saittai 190, 281–7, 294, 347
scale 9–10, 13–14, 29–31, 33, 36–7, 39, 57, 78, vertical integration 240, 396
88, 98, 111, 116, 136–9, 173, 208, 314, vertical specialization 117, 124–5
380–400 Veturii 208
Seneca 62, 81 Vindolanda 134, 137, 140–3
seviri Augustales 29, 122, 233, 246–7, 258–68 Voorburg 141, 144
shipwrecks 101–3
shippers 97, 110, 222–3, 233, 240, 256–61 water provision 343, 394
shoes, see footwear water-mills 87, 156, 281, 292
408 Index
weaving 82, 203, 208, 214, 218, 281–2, women 11, 41, 67, 147, 200–19
339, 343 wood-working 63, 82
Weber, Max 33–4, 117–18, 127 workshops 6–7, 9, 25–7, 36–7, 77–90, 115–16,
wine trade 98–101, 232–3 154, 189, 310–20, 325–6, 340–4, 356–61