Labour, Coercion, and Economic Growth

in Eurasia, 17th–20th Centuries

 Studies in
Global Social History

Series Editor
Marcel van der Linden
International Institute of Social History, Amsterdam, The Netherlands

Editorial Board
Sven Beckert
Harvard University, Cambridge, MA, USA
Philip Bonner
University of the Witwatersrand, Johannesburg, South Africa
Dirk Hoerder
Arizona State University, Phoenix, AZ, USA
Chitra Joshi
Indraprastha College, Delhi University, India
Amarjit Kaur
University of New England, Armidale, Australia
Barbara Weinstein
New York University, New York, NY, USA

VOLUME 11

The titles published in this series are listed at brill.com/sgsh

Labour, Coercion, and Economic Growth
in Eurasia, 17th–20th Centuries

Edited by
Alessandro Stanziani

LEIDEN • BOSTON
2013
Cover illustration: Miniature of Indian craftsmen, painted by Svami in the 18th century. Courtesy of
the Bibliothèque nationale de France.

Library of Congress Cataloging-in-Publication Data

Labour, coercion, and economic growth in Eurasia, 17th–20th centuries / edited by Alessandro
Stanziani.
  p. cm. — (Studies in global social history, ISSN 1874-6705 ; 11)
 Includes bibliographical references and index.
 ISBN 978-90-04-23112-2 (hbk. : alk. paper)—ISBN 978-90-04-23645-5 (e-book)
1. Forced labor—Eurasia—History. 2. Labor—Eurasia—History. 3. Economic development—
Eurasia—History. I. Stanziani, Alessandro.

 HD4875.E83L33 2012
 331.11’730950903—dc23
2012028033

This publication has been typeset in the multilingual “Brill” typeface. With over 5,100 characters
covering Latin, IPA, Greek, and Cyrillic, this typeface is especially suitable for use in the
humanities. For more information, please see www.brill.com/brill-typeface.

ISSN 1874-6705
ISBN 978-90-04-23112-2 (hardback)
ISBN 978-90-04-23645-5 (e-book)

Copyright 2013 by Koninklijke Brill NV, Leiden, The Netherlands.

Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing,
IDC Publishers and Martinus Nijhoff Publishers.

All rights reserved. No part of this publication may be reproduced, translated, stored in
a retrieval system, or transmitted in any form or by any means, electronic, mechanical,
photocopying, recording or otherwise, without prior written permission from the publisher.

Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV
provided that the appropriate fees are paid directly to The Copyright Clearance Center,
222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA.
Fees are subject to change.

This book is printed on acid-free paper.

 contents

Notes on Contributors ...................................................................................... vii

Introduction: Labour, Coercion, and Economic Growth in

Eurasia, Seventeenth–Early Twentieth Centuries .............................. 1
. Alessandro Stanziani

Part one

JURIDICAL MODELS AND LABOUR DYNAMICS

The Duty to Work: A Comparison of the Common Law and

Civil Law Systems from the Eighteenth to the Twentieth
Centuries ......................................................................................................... 29
Simon Deakin

Dutch Imperial Anxieties about Free Labour, Penal Sanctions

and the Right to Strike ................................................................................ 63
Ulbe Bosma

Children and Forced Labour in the Indian Ocean World,

circa 1750–1900 ............................................................................................... 87
Gwyn Campbell

part two

DEPENDANCE AND SERVITUDE AT WORK:

LOCAL CUSTOMS AND GLOBAL DYNAMICS

Factors that Shaped the Organization of Labor and the Labor

Market in Tokugawa Japan: Kyoto and Central Japan ...................... 115
Mary Louise Nagata

Contractual Relations, Tariffs and Customs in the Lyon Silk

Industry in the Nineteenth Century ....................................................... 145
Pierre Vernus
vi contents

The Circulation of Commercial Manpower in an Indian

Worldwide Trading Network in the Early Twentieth Century ....... 175
Claude Markovits

part three

DANGEROUS TIES:
STATE, LANDLORDS AND LONGUE DURÉE SERVITUDES

Constrained Labour in Early-Modern Rural East-Central and

Eastern Europe: Regional Variation and Its Causes ........................... 191
Markus Cerman

Rights and Bondage in Russian Serfdom .................................................... 215

Alessandro Stanziani

Acting As Master and Bondservant: Considerations on

Status, Identities and the Nature of “Bond-Servitude”
in Late Ming China ...................................................................................... 237
Claude Chevaleyre

Public Works and the Question of Unfree Labour .................................. 273

Chitra Joshi

References ............................................................................................................ 289
Index ...................................................................................................................... 323
 NOTEs ON CONTRIBUTORS

Ulbe Bosma is Senior Researcher at the International Institute of Social

History in Amsterdam and published several books, among which are
Being ‘Dutch’ in the Indies: A History of Creolisation and Empire, 1500–1920
(Singapore University Press/Ohio University Press), with Remco Raben,
and articles on colonialism and migration in International Migration
Review, Journal of Southeast Asian Studies and Journal of Global History.
His recent research is on labour regimes, migration and sugar production
in South and Southeast Asia.

Gwyn Campbell is a Canada Research Chair, and Director of the Indian

Ocean World Centre, at McGill University. He has published extensively
on Africa and the Indian Ocean world, including An Economic History
of Imperial Madagascar, 1750–1895 (Cambridge, 2005) and David Griffiths
and the Missionary “History of Madagascar” (Brill, 2012). He is currently
completing a work entitled Africa and the Indian Ocean World from early
times to 1900 to be published by Cambridge University Press.

Markus Cerman is Associate Professor of Economic and Social History in

the University of Vienna. His research concentrates on economic develop-
ment in late medieval and early modern Europe and comparative rural
history. His publications include European proto-industrialization, Cam-
bridge 1996 (with Sh. Ogilvie, eds.); Wirtschaft und Gesellschaft. Europa
1000–2000, Innsbruck 2011 (with F. Eder et al., eds.); Demesne lordship and
rural society in East Central and Eastern Europe, c.1500–c.1800 (Agricul-
tural History Review 59, 2, 2011, special issue) (with R. Hoyle, eds.); Villag-
ers and lords in Eastern Europe, 1300–1800, Houndmills, 2012.

Claude Chevaleyre pursues doctoral studies in Chinese history at the

Ecole des Hautes Etudes en Sciences Sociales (Paris), under the direction
of Professor Pierre-Etienne Will (Collège de France). His dissertation is
dealing with the social status and identity of so-called “bondservants” in
the Ming (1368–1644) and Qing (1644–1911) dynasties. His research focuses
on late imperial social and legal history and pays particular attention to
the Ming-Qing transition period. Among his publications are Bibliographie
d’Etienne Balázs: Œuvres, projets, appréciations”, in Actualité d’Etienne
viii notes on contributors

Balazs (1905–1963). Témoignages et réflexions pour un centenaire. Paris: De

Boccard, 2010, pp. 111–130.
“Under Pressure and out of Respect for Human Dignity: The 1910 Chi-
nese Abolition”. Forthcoming in a collective book, directed by Myriam
Cottias et Marie-Jeanne Rossignol, on the worldwide influence of the
western abolitions of slavery.

Simon Deakin is a Professor of Law at the University of Cambridge. He

specialises in labour law, company law and private law, with a focus on
the empirical and historical dimensions of legal change. He is the co-
author of ‘The Law of the Labour Market’ (with Frank Wilkinson, 2005)
and co-editor of ‘Capacitas: Contract Law and the Institutional Precondi-
tions of a Market Economy’ (with Alain Supiot, 2009).

Chitra Joshi teaches history at Indraprastha College, University of Delhi,

India. She is the author of Lost Worlds. Indian Labour and its Forgotten
Histories (Delhi: Permanent Black, 2003 and London: Anthem Press, 2005)
and of numerous book chapters and articles in i.a. Indian Economic and
Social History Review, International Review of Social History and Studies in
History. She is currently working on the history of roads and labour in
nineteenth-century India.

Claude Markovits is a Senior Research Fellow Emeritus at the Centre

National de la Recherche Scientifique and is affiliated with the Centre of
Indian and South Asian Studies at the Ecole des Hautes Etudes en Sciences
Sociales, Paris. His publications include The Global World of Indian Mer-
chants (Cambridge: Cambridge University Press, 2000), The Un-­Gandhian
Gandhi: The Life and Afterlife of the Mahatma (Delhi: Permanent Black,
2004), and Merchants, Traders, Entrepreneurs (Palgrave Macmillan, Bas-
ingstoke and New York, 2008).

Marie Louise Nagata is associate professor at the department of History,

Francis Marion University. She is the author of Labor Contracts and Labor
Relations in Early Modern Central Japan (Routledge Curzon Press, 2005)
and has published numerous articles in the field of East Asian history,
with a focus on family life and labor issues in Japan.

Alessandro Stanziani is Directeur d’études (full professor) at the EHESS

and Senior Researcher at the CNRS, Paris. He works on Russian, European
 notes on contributors ix

and the Indian Ocean history of Labour, Food and Economic Institutions.
He is the author of:
L’économie en révolution, le cas russe, 1870–1930, Paris, Albin Michel,
1998; Histoire de la qualité alimentaire en France, XIXe–XXe siècle, Paris,
Seuil, 2005; Rules of exchange. French capitalism in comparative perspec-
tive, 18th–20th centuries, Cambridge, Cambridge University Press, 2012.
Bâtisseurs d’Empires. Inde, Russie, Chine à la croisée des mondes, XVe–
XVIIIe siècle, Paris, Liber, 2012.

Pierre Vernus is lecturer in contemporary history at the Univer-

sity Lumière Lyon 2 and deputy director of the Rhône-Alpes Historical
Research Unit (modern and contemporary history). He published Art,
Luxe et Industrie. Bianchini Férier, Un Siècle de Soieries Lyonnaises, 1888–
1992 (2006) and articles about silk industry and business interests associa-
tions. He recently edited with Danièle Fraboulet Genèse des Organisations
Patronales en Europe (XIXe–XXe siècles) (2012).
 INTRODUCTION

LABOUR, COERCION, AND ECONOMIC GROWTH

IN EURASIA, SEVENTEENTH–EARLY TWENTIETH CENTURIES

Alessandro Stanziani

The Main Scope and Argument

The history of the forms of “free” labour is intimately linked to that of

coerced labour. This book shows that between the seventeenth and the
twentieth century, forms of labour and bondage all over Eurasia were
defined and practised in reference to each other. A whole spectrum of
forms of dependence, bondage and labour existed in Russia, India and
Indonesia as well as in Japan, China, Britain, France, Prussia and the
Indian Ocean World.
This was true not only in specific areas—rooted in local institutions,
values and economic relationships—but also on a global scale. Serfdom in
Prussia and in Russia expressed an extreme variant of Western European
notions and practices of labour as domestic service and social obligation.
Similarly, in the European colonies, indentured labour was conceived as
an extreme form of domesticity and servants’ subordination in Europe.
Without the identification of the worker as a servant in Britain, France
and the Netherlands, indentured labour in the British, French, and Dutch
colonies would have not been possible.
Common tendencies and local perspectives on labour relationships
found their sources not only in the global circulation of models, peoples,
goods and institutions, but also in market dynamics. From the seventeenth
and the late nineteenth century, proto-industry, agriculture, trade and
manufacturing experienced unprecedented growth throughout Eurasia.
Mostly labour-intensive, this long-term growth put considerable pressure
on labour resources and contributed to increased coercion and legal con-
straints on labour mobility in both Asia and Europe.

The Uncertain Boundary between “free” and “unfree” Labour

Anthropologists, sociologists and historians have highlighted differ-

ent aspects of labour relationships, according to their disciplines, in an
2 alessandro stanziani

attempt to draw the line between “free” labour and “forced” labour, par-
ticularly slavery. They have discussed social status (membership in or
exclusion from the clan, the family, the local community), religion, legal
status (the form of dependence, freedom of movement, the hereditary
character of such constraints), socio-economic conditions (dependence,
non-economic advantages, coercion, etc.), political rights and legal (and
procedural) rights.1 Researchers have pinpointed several variables, but
without reaching a consensus. These issues have been debated even more
fiercely in the last twenty years as cultural studies and subaltern studies
brought out the relativity of the notions of freedom and coercion. As
a result, the question has now become whether or not a given form of
dependence, bondage, etc. found in a particular society in Africa, Asia, the
Indian Ocean or the Americas could be considered “slavery”. If the answer
is yes, then by implication slavery existed before and independently of
colonialism; conversely, if the answer is no, it means that these forms of
dependence and bondage were specific to a particular place and “imperi-
alist” and revisionist culture would like to call them “slavery” to minimise
the West’s “debt” to the Third World.
The debates have become all the more virulent as they are no longer
confined to academia. One obvious example is the quarrels that have pit-
ted international organisations against countries and firms accused of
practising and legitimising hidden “slavery”.2 It is symptomatic and some-
times paradoxical to see cultural relativism espoused both by academics
critical of “imperialism”, “colonialism” and now globalisation and by local
managers and multinational companies that exploit child labour.
The aim of this book is not to take sides in favour of one or the other
“general” definition of labour and forced labour, but rather to set the
boundary line between free labour and forced labour in specific historical
and institutional contexts and explain why, in a given context, this line
was conceived and put into practice in one way rather than another.3 By
undertaking a radical re-examination of the historical forms of labour and
how they were defined, we are not seeking to relativise and deconstruct
categories in order to assert, for example, that “forced labour did not exist”

1 Testart, L’esclave; Meillassoux, Anthropologie; Finley, Ancient slavery; Miers, Kopytoff,

Slavery in Africa; Williams, Capitalism; Bush, Serfdom and Slavery; Engerman, Terms; Klein,
Breaking; Patterson, Slaver; Lovejoy, Transformations.
2 ILO, International; IPEC, Every Child; Cunningham, Viazzo, Child Labour; Miers, “Con-
temporary forms”.
3 Campbell, Miers, Miller, Children.
 introduction: labour, coercion, and economic growth 3

or that it is an “intellectual invention”. On the contrary, by viewing these

elements in their proper historical contexts, we hope to provide an origi-
nal explanation of the dynamics of forms of labour. Instead of attempting
to establish the moment when “free labour” and “civilisation” emerged, or
conversely, stigmatising the continuation of the “guild tradition” or even
of latent forms of slavery, we want to grasp the dynamics at work in cer-
tain historical forms of labour starting from the historically situated ten-
sion between freedom and constraint.
The first hypothesis underlying this book is that so-called “free” forms
of labour and bondage were defined and practised in reference to each
other, not only within each country and region, but also on a global scale.
In the Anglo-Saxon world (chapter by Deakin), until almost the end of
the nineteenth century, there were fewer “wage earners” than servants,
i.e. workers whose status resembled that of servants in husbandry. These
workers enjoyed limited rights in relation to their employers: the absence
or unjustified breach of a work contract was punishable by criminal pen-
alties. Yet it was precisely the definition of “free” labour, based on the
unequal legal statuses of employers (actually masters) and wage earn-
ers (servants) that enabled several forms of bondage to be considered a
contractual “free choice” at the time.4 In the British empire, the Ameri-
can colonies and later the United States, from the seventeenth century
to around the middle of the nineteenth century, the indenture contract5
which historians today usually classify as a form of forced labour, was in
no way opposed to free labour during this period.6 The indenture contract
was seen as an expression of contractual free will; in India, this form of
servitude continued until the twentieth century.7 It is therefore important
to note the formal and factual link between the forms of bondage and the
definition of “free” labour: if the wage earner had not been defined as a
servant, it would never have been possible to consider an emigrant sign-
ing an indenture contract a “free worker”. That explains why indenture
was not classified as a form of “forced” labour in the colonies until the
collapse of the Masters and Servants Acts, when free labour was redefined
(1875) in Britain.8

4 Steinfeld, Coercion; Deakin, Wilkinson, The Law.

5 Steinfeld, The Invention; Galenson, White Servitude; Northrup, Indentured Labor.
6 Hay, Rogers, English Society; Hay, “Masters and servants”, in Steinmetz, Private Law.
7 Hay, Craven, Masters.
8 Northrup, Indentured Labor.
4 alessandro stanziani

The connection between the forms of labour in the “centre” and in the
colonies was by no means specific to the British empire. It also existed
in the colonies of the Dutch empire, especially Indonesia, where colonial
leaders based their conception of post-slavery labour on the constraints
imposed on servant wage earners in the Netherlands (chapter by Ulbe
Bosma). The conceptions and practices of labour in Europe and its main
colonies influenced each other and entered into global dynamics. Inden-
tured labour was not the antithesis of “free” wage labour but an extreme
form of the peculiar historical meaning that wage labour acquired in mod-
ern Europe.
But then if the definition of “free” labour included the indenture con-
tract and allowed for numerous constraints and penalties placed on wage
earners, what distinguished it from officially recognised forms of bondage,
beginning with serfdom?
As a matter of fact, what has been called the “second serfdom” in East-
ern Europe needs to be redefined.9 Serfs were never legally defined as such
in Prussia10 and Central and Eastern Europe (chapter by Markus Cerman),
or even in Russia (chapter by Stanziani). The documents usually cited as
proof of the introduction of serfdom actually refer to forms of constraint
on mobility and designated those who had the right to own and transfer
inhabited estates, i.e. various categories of nobles.
Forms of bondage beyond those justified by “free contracts” were wide-
spread in India (chapter by Joshi) and in Africa (chapter by Campbell).
These types of bondage close to slavery were based less on European law
than on local customs. Hence, a continuum of solutions appears, with a
variety of forms of debt bondage, indenture contract, and other forms of
bondage contracts between English masters and Indian coolies, as well as
genuine slavery.11
The same complexity was found in China (chapter by Chevaleyre),
where the status of servants and bonded labourers differed from Euro-
pean experiences, sometimes resembling more closely the situations of
Ottoman palace slaves and servants in husbandry. The complexity of sta-
tuses and real conditions and the inadequacy of “European” categories
to express them testify first to the widespread presence of these situa-
tions and second, to their specific character. The same issue arises in

9 Stanziani, “Free labor-forced labor”.

10 Melton, “The decline”; Ogilvie “Communities”.
11 Patnaik, Dingwaney, Chains; Tinker, A New System; Pouchepadass, Paysans; Condom-
inas, Formes extrêmes; Reid, Slavery.
 introduction: labour, coercion, and economic growth 5

every study of slavery as soon as we step outside of the ancient world or

nineteenth-century North America, where in both cases slave status was
passed on to descendants.
Yet local specificities must be put into a global context. In time as well
as in space, continuities and links between free and unfree labour prevail
over clear-cut oppositions. Unlike the conventional thesis—celebrating
the triumphal march towards freedom starting in the nineteenth century
after centuries if not millennia of bondage—recent approaches in histo-
riography underline the changes in slave-trading systems prior to official
abolition and, conversely, the continuation of forms of bondage and slav-
ery after the reforms. The conditions of African-Americans and of “free”
labourers in the colonies are used to confirm this position.12
Similarly, historians of serfdom in Eastern Europe have revealed the
decadence of the system well before the arrival of Napoleon and the
French codes,13 just as French historians have shown how medieval serf-
dom evolved prior to the Enlightenment.14 Both groups have also brought
to light aspects of bondage and service in husbandry that continued after
the official abolition of serfdom.15 New trends have also arisen among his-
torians of wage labour. Anglo-Saxon historians have underscored the coer-
cive nature of the British and American norms applied to labour (Masters
and Servants Acts) until roughly the end of the nineteenth century.16 In
France, the break between the world of guilds and the Old Regime on the
one hand, and the world of free labour on the other, has likewise been
questioned.17 The rise of collective bargaining has been deemed at least
as important in terms of its effects on labour relations as the abolition of
guilds and of labour service (corvée).18
Our approach confirms these new contributions of historiography and
emphasises the continuities rather than the breaks between the late sev-
enteenth century and the early twentieth century in the area of labour and
its institutions. Despite institutional and political shifts, important conti-
nuities in labour regulations and practices are revealed and confirmed by
the cases of Britain (Deakin), France (Vernus), Austria and Prussia (Cer-
man), Japan (Nagata), China (Chevaleyre), and Russia (Stanziani).

12 Miller, Slavery and Slaving; Drescher and Engerman, eds., A Historical Guide.
13 Hagen, Ordinary.
14 Bois, La crise; Duby, Les trois ordres; Bonnassie, From Slavery.
15 Hagen, Ordinary, 2004, Kolchin, Unfree Labor.
16 Steinfeld, The Invention.
17 Minard, La fortune.
18 Didry, Naissance.
6 alessandro stanziani

Continuities were even stronger in colonial empires where a significant

link has been established between the conditions of European wage earn-
ers (mainly servants in husbandry and household servants) and inden-
tured immigrants in Indonesia (Bosma) and throughout the Indian Ocean
World (Campbell).
However, the continuities between free and unfree labour are impor-
tant not only in time but also in space. Surprisingly enough, the recent
research orientations we have just mentioned have never come together:
there has been little dialogue between historians of slavery and histori-
ans of wage labour, and consequently neither group has challenged the
presupposition that these two worlds were and remain separate or even
opposed. The aim of this book is precisely to overcome this fracture by
revealing the connections between these elements underpinned by chro-
nologies that are in fact too common to be unconnected or to have come
about merely by chance. To answer this question, the historical dynamics
of labour must be understood at once in a global dimension and in their
local specificities. We will therefore examine them on these various lev-
els first by studying the circulation of economic and legal knowledge and
second, how it was applied locally.

Circulation of Knowledge and Labour Dynamics

Not all the contributions to this work necessarily adopt a global or com-
parative approach. On the contrary, we think specific national and even
regional and local features should be taken into consideration in order
to understand how the whole system worked. The strength of global his-
tory lies not in collecting second-hand banalities common to a number
of different worlds, but rather in achieving a relevant representation of
this multiplicity through local and national specificities. The comparative
dimension applied at various levels is therefore an integral part of our
method, which is why we propose to place side by side situations and
historical experiences that are usually studied separately: wage labour in
France and England, Russian serfdom and slavery in China, work disci-
pline in India and Japan. This approach reveals striking similarities as well
as the differences between the contexts and forms of labour.
From the standpoint of global history we have adopted, two main
variables are worthy of mention: the circulation of economic and legal
knowledge and economic and institutional dynamics between the seven-
teenth and twentieth centuries. Economic knowledge and legal models
 introduction: labour, coercion, and economic growth 7

circulated along with people and goods. This circulation led not only to
increased homogeneity among systems but also to differentiation and
even hierarchies of areas and countries. For example, the identification
of serfdom with an archaic world in the “East” and even the notion of
“Eastern Europe” itself were pure inventions of the Enlightenment.19 Mon-
tesquieu published The Spirit of the Laws in 1748, soon to be followed by
the first volumes of the Encyclopédie.20 In these works, the serfdom of
absolutist and medieval Europe was contrasted with the free labour of
Enlightenment Europe. The advances of the Enlightenment contributed
to the invention of a historiographic break between “enlightened France”
and the old France on the one hand, and between Western and Eastern
Europe on the other. These two “inventions” mirror each other, e.g., in the
fact that serfdom could be found in Eastern Europe as well as in medieval
Europe.
Similarly, British and Dutch colonisers in India and Indonesia strove to
translate local forms of dependence and bondage into their own categories;
the disparities and similarities between the forms of slavery and bondage
in the two worlds are key factors in grasping the complexity of the defini-
tions and of the labour practices themselves (chapter by Ulbe Bosma).
If labour practices had not been translated into English legal terms, the
forms of post-slavery and post-colonial bondage would have developed
differently.21 Conversely, the use of written contracts rather than informal
arrangements reveals the need on the part of important Indian traders in
Cairo to be able to bring their shop assistants and servants before English
courts if necessary (chapter by Markovits). The contracts sought to trans-
late the relationships between caste, type of labour and remuneration into
terms and categories derived from British norms.
Yet the circulation of ideas and practices did not necessarily indicate a
relationship of dependency of the so-called “periphery” on the core; colo-
nial discourse aside, forms of bondage indeed existed in Russia, India,
Africa and the Americas well before the arrival of Europeans.22 And even
during the colonial period, dependency was not simple and straightfor-
ward. To return to the example of Indonesia, servitude was defined and
put into practice based on indigenous notions and practices that had

19 Stanziani, “Free Labor”.

20 Duchet, Anthropologie.
21 Prakash, Bonded Histories.
22 Coquery-Vidrovitch, Moniot, L’Afrique noire; Thornton, Africa; Meillassoux, L’esclavage.
8 alessandro stanziani

c­ ontinually influenced those discussed and applied in the Netherlands

itself (chapter by Bosma).
In like manner, in the eighteenth and nineteenth centuries, there were
constant references to the case of Russia in debates over work discipline
in Europe, where the assessment of the conditions of “serfs” led people to
take a position regarding wage labour, which was viewed either in opposi-
tion to serfdom or, on the contrary, considered a new form of slavery.23
From this same point of view, the specificity of Japan did not consist
in simply adapting the Western model. In traditional historiography,
Japanese development has always been tied to Meiji reforms and West-
ern influences. On the contrary, recent analyses including Marie-Louise
Nagata’s contribution here, demonstrate that Japan’s economic and social
dynamism dates back to the eighteenth century.
To sum up, the circulation of ideas and models is important, but it can-
not be taken exclusively as a synonym of dependency because 1) the “cen-
tre” and its colonies often influenced each other, 2) bondage and slavery
did in fact exist, although in different forms, before and after colonisation
and 3) finally, emancipation did not come about solely under pressure
exerted by the “advanced West”. The role of the circulation of knowl-
edge therefore has to be associated with economic and social trends as
a whole.

Global Intensification of Labour

How can we possibly justify the increasing limitations placed on free

labour throughout Eurasia between the seventeenth and the nineteenth
centuries? Was this not a period of strong demographic growth and
increasing capital development?
Logically those processes should have led to reducing constraints on
labour rather than strengthening them. This line of reasoning fails, how-
ever, to take into account the above-mentioned values and notions of
labour as a social service and an obligation on the one hand, and the fact
that this period witnessed unprecedented labour-intensive growth all over
Eurasia. Indeed, over the past two decades at least, economic historians
have moved away from the interpretation developed in the early nine-
teenth century by the so-called “classical” economists (Smith, Ricardo)

23 Stanziani, “Free Labor”.

 introduction: labour, coercion, and economic growth 9

and Marx, according to which the agrarian revolution and later the indus-
trial revolution were accompanied by the substitution of labour by capital
and the concentration of production units. In fact, they occurred mainly
in the twentieth century. Before that date, in most of Eurasia, proto-
industry, agriculture and even industry rather followed a labour-intensive
path. First, with regard to agriculture: conventional histories of economic
growth stress the relative decline of agriculture during the industrialisa-
tion process; at the same time, agriculture was supposed to provide goods
to feed a growing urban population. This outcome required greater pro-
ductivity and yields, which in turn were obtained through a shift in the
organisation of economic units and in the relative weight of the factors
(decreasing labour and land and increasing capital). These views have
been increasingly challenged, starting with the role of capital in agri-
culture. Recent empirical analyses show that in many parts of England
livestock densities were stable throughout the modern period until the
mid-nineteenth century;24 conversely, there was more livestock in Asia
and Eastern Europe at the time than is commonly believed.25 In other
words, in Britain, the increase in yields that occurred before 1800 cannot
be explained by rising livestock numbers.26 On the rest of the continent as
well, the long-term trend of rising wheat prices (roughly between 1680 and
1815) led to reducing the surface area devoted to livestock and livestock
feeding while increasing the acreage for wheat cultivation.
The role of machines and new sources of power is also undergoing revi-
sion: after steam became the dominant form of power employed in manu-
facturing, the major sources of energy available to farmers continued to
be men, animals, wind and water.27 Mechanisation in farming proceeded
slowly because agricultural operations were more separated in time and
space than industrial processes. Mechanisation was therefore a relatively
unimportant component of the changes in agriculture technology up to the
mid-nineteenth century28 when the appearance of commercial fertilizers
and the development of mechanical harvesting equipment began signifi-
cantly to affect methods of production.29 Until the machine age, i.e. after
1850, much of the rise in productivity and the growth of ­output depended

24 Allen, Enclosures.
25 Pomeranz, The Great.
26 Allen, “Tracking”: 226.
27 O’Brien, “Agriculture”.
28 O’Brien, “Agriculture”.
29 Grantham, “Agricultural Supply”.
10 alessandro stanziani

more on the intensive use of known technology than on novel methods.

So-called “new husbandry” was indeed not so new30 and it required more
labour, not less.31 Only in agricultural systems like those in the United
States, characterised by the high opportunity cost of labour, did economic
pressure to mechanise come to resemble that experienced by some sec-
tors of industry. On the contrary, in Russia and France as well as Prussia,
from the seventeenth century through the second half of the nineteenth
century, labour was not only the major input in agriculture, either directly
or embodied in land improvements, but its weight even increased dur-
ing this period.32 Recent analyses converge towards the same conclusion:
labour and labour intensity are now identified as the main source of agri-
culture growth before 1850, with human and physical capital playing a
secondary role.33 Labour-intensive techniques linked to the dissemination
of knowledge and attractive markets (with rising agriculture prices) were
widely used between the seventeenth century and the last quarter of the
nineteenth century, when this trend reversed (agricultural prices went
down and wages went up).34
However, increasing labour demand in agriculture had to compete with
similar processes in proto-industry and manufacturing. Theories of proto-
industrialisation originally associated proto-industry with demographic
growth and proletarianisation on the one hand, and a decline of urban
guilds and feudal institutions on the other.35 Further analyses have called
these assertions into questions and nowadays it is widely accepted that
on a comparative scale no single, uniform link can be established between
proto-industrialisation and any of the aforementioned variables. Proto-
industry developed in Western, Central and Eastern Europe from the end
of the seventeenth century in response to market demand and demo-
graphic pressure.36 It retained its central position all over Europe at least
until the mid-nineteenth century. After that date, and only after, some
areas declined and manufactures and industries replaced the ­putting-
out system.37 However, the shift was by no means complete, and in

30 Federico, Feeding.
31 O’Brien, “Path Dependancy”.
32 Clark, “Productivity growth”.
33 Grantham, “Agricultural Supply”; Allen, Enclosures; O’Brien, “Agriculture”.
34 Thompson, “The Second”.
35 Mendels, “Des industries rurales”; Mendels, “Proto-industrialization”; Kriedte, ­Medick,
Schlumbhom, Industrilization before industrialization.
36 Hagen, “Capitalism in the countryside”.
37 Ogilvie, Cerman, European proto-Industrialization.
 introduction: labour, coercion, and economic growth 11

many European areas and districts, proto-industry continued to play a

leading role during the second half of the nineteenth century and even in
the twentieth century.38
This timing is even more relevant for Asia: authors such as Lee and
Sugihara have maintained that the “industrious revolution” identified by
De Vries in Europe39 was also visible in areas of Asia.40 As in Russia, the
success of proto-industry in Japan (chapter by Nagata) and China, too,
particularly rural proto-industry, was at the root of a labour-intensive path
of growth.41
In all these areas, as in most Russia regions (chapter by Stanziani), agri-
culture did not turn into a simple supplier of produce and labour-force for
industry; on the contrary, estates and peasants took part to the develop-
ment of local and national markets for both wheat and proto-industrial
products. In Russia (Stanziani), Japan (Nagata), and Central Europe (Cer-
man), commercialization of produce was not the consequence of extra-
economic coercion but responded to market dynamics.42 As in many
areas of Russia and Western Europe, increases in agricultural output and
income in Japan led to growth in demand for manufactured goods, which
was met by an expanding rural industry using labour-intensive technol-
ogy. The resulting growth in rural non-agricultural activity in turn gener-
ated increased incomes for rural households and hence greater demand
for agricultural output.43
Similar results are now available on India. Contrary to the usual view,
which stressed the decline of cottage industry under British rule and
the growth of international markets, new research shows that the use of
“traditional” labour-intensive techniques has remained widespread up to
the present day. Thanks to their flexibility, these techniques have allowed
a labour-intensive pattern of growth linked to family units to remain
integrated in both agriculture and industrial markets.44
The persistent, global strength of agriculture and proto-industry had an
unanticipated effect, however: urbanisation and the supply of labour for
urban manufacturing were mostly seasonal.

38 Sabel, Zeitlin, Worlds of possibilities.

39 De Vries, “The industrial revolution”.
40 Sugihara, “Labour-Intensive”; Lee, “Trade and Economy”.
41 Frank, Rural Economic; Pomeranz, The Great.
42 Sugihara, “Labour-Intensive”.
43 Frank, Rural Economic; Saito, “The Labour Market”.
44 Roy, The Economic History; Perlin, “Proto-Industrialization”.
12 alessandro stanziani

Until the mid-nineteenth century, double employment (rural and

urban) was the rule rather than the exception not only in Russia (Stan-
ziani), Japan (Nagata) and France,45 but also in Britain. According to
Lindert and Williamson, for the larger occupational groups such as “agri-
culture”, “commerce” and “manufacturing trades”, the census and statisti-
cal error margins are probably within the range of minus 40 to plus 66
per cent!!46
Seasonal needs in agriculture were the crucial variable here. In Rus-
sia (Stanziani), China and Japan (Nagata) as well as Britain, France, and
Central Europe (Cerman), the solution to seasonal and local shortages
of manpower lay in interregional migration and, later in the nineteenth
century, in the transformation of hand harvesting techniques and tools.47
In fact, the labour requirements for harvesting were especially high since
labour input peaked sharply at harvest time.48 Thus, as Markus Cerman
clearly shows in his chapter, landowners in Brandenburg satisfied part of
their manorial labour needs by requiring the children of their peasant ten-
ants to serve on the manor for a period of three years. However, at harvest
time, the labour services provided by peasants and their children fell short
and recourse to seasonal day labour was widespread. The proportion of
rural dwellers that worked part time for corn growers depended on what
they did outside the peak season.
These activities were extremely problematic for rising manufacturing
and industry which, like the other parts of the economy, mostly relied
upon labour. Because the price of capital was still high, urban employers
sought to cope with labour shortages by putting legal pressure on labour-
ers, who were not allowed to leave before the end of their terms, as well
as on competitors through strong penalties for unfair competition, etc.
The first industrial revolution in Britain was far from the overall substitu-
tion of labour with capital. Feinstein’s estimations show that in Britain,
capital and labour grew at about the same rate from 1760 through 1830,
so that there was effectively no change in the capital-labour ratio during
these seven decades. In the last three decades, the ratio slowly rose, as
capital per worker increased at a rate of about 1/2% per year.49 The rate

45 Postel-Vinay, “The Dis-Integration”.

46 Lindert, Williamson, “Revising”; Lindert, Williamson, “English workers”.
47 For Britain: Collins, “Migrant labour”; for france: Postel-Vinay, “The Dis-integration”;
For Germany: Melton, “Population structure”.
48 Grantham, “Divisions of labour”.
49 Feinstein, “Capital Formation”.
 introduction: labour, coercion, and economic growth 13

of capital formation in Britain was relatively slow on the whole until the
mid-nineteenth century and the capital-labour ration increased sharply
only afterwards.50
By 1850, relatively few workers were employed in factories: only a small
proportion worked in technologically advanced industries such as cotton,
iron and steel, and metalwork, and the full impact of steam power in
transport and production was yet to be felt.51 This view breaks with the
traditional scenario of the first industrial revolution and strongly reduces
the gap between Britain and other countries, particularly France. Histori-
ography long considered France “backward” in comparison with Britain,
precisely because of its slow rate of concentration and capital intensifica-
tion in industry.52 Recent analyses have consistently modified this view:
if British capital intensification and the increasing capital-labour ratio in
industry have been revised downwards, on the contrary, French dynamics
have been corrected upwards. From this standpoint, the relatively slow
pace of capital growth in France, and the increasing ratio of labour to
capital industry are no longer seen as exceptional or “inefficient”.53
This means that not only in Russia, but also in France and most of the
European countries, economic and industrial growth in the eighteenth
and nineteenth centuries remained on a small scale and was labour inten-
sive. Growth was most often achieved within the same production func-
tion whose scope expanded slightly until the mid-nineteenth century.54
There is evidence that much of the productivity increase was not associ-
ated with specific innovations, but rather with workers operating more
machines.55 Christine Mac Leod reveals that the most frequently declared
goal of innovation was either improving the quality of the product or sav-
ing on capital, not labour. And if inventors were not particularly intent
on saving labour, those who judged their inventions were even less so.
In other words, economic actors did not wish to substitute labour with
capital and the final outcome for the whole economy was an increasing
demand for labour.56

50 Craft, British Economy; J. Williamson, “Why was British growth”; Harley, “British
industrialization”. Deane, “Capital Formation”; Feinstein, Pollard (eds.), Studies in capital
formation; Craft, British Economic Growth.
51 Deakin, Wilkinson, The law: 20.
52 Crouzet, British ascendant; Lévy-Leboyer, Bourguignon, L’économie française.
53 Craft, British Economic; O’Brien, Economic Growth in Britain and France.
54 Craft, British Economic.
55 Clark, “Productivity Growth”.
56 Mac Leod, Inventing: 158–181.
14 alessandro stanziani

Overall, in British industry during the second half of the eighteenth cen-
tury, labour input grew at about 1.2–1.3 per year, 2/3 of which was caused
by a larger population, and the remaining third coming from longer work-
ing hours.57 Following E.P. Thompson,58 legions of historians have endea-
voured to show that working time actually increased with the industrial
revolution. Indeed, this process started much earlier, in the seventeenth
and eighteenth centuries, with the “industrious revolution”59 and the mul-
tiplication of farming and industrial activities.60 Thus, the global history
of the eighteenth and nineteenth centuries in all the countries we men-
tioned shows a similar evolution of working time, i.e. workdays became
longer as labour became more intensive.61 The innovations and technical
improvements recorded in agriculture, industry and trade did not take
place at the expense of labour but actually fostered employment. This
mechanism is confirmed in the study of manufactures in France and Eng-
land, Japan and India. From the point of view of labour institutions and
the relationship of labour to other production factors, it thus became a
common “wave of capitalism” (to borrow Braudel’s expression but not his
chronology) that lifted Eurasian capitalism as a whole from the seven-
teenth to the mid-nineteenth century.
Hence questions concerning working time deserve to be studied within
a more complex set of dynamics: if labour was already becoming more
intensive in every sector during the preindustrial era, then how could
presence at work be controlled when families, landowners, traders and
manufacturers were competing for it?

Child and Female Labour

In labour relations, children have always played a crucial role in every

context as the focus of controversies relating not only to economics
but also the social order, religion and institutions. Of course it is neces-
sary to consider how “child” was defined, which changed according to
the period and place62 (chapter by Gwyn Campbell). At the same time,

57 Voth, “Time and Work”.

58 Thompson, “Time, work-discipline”.
59 De Vries, The industrious; Voth, “Time and work”.
60 Craft, British Economic Growth; Mokyr, The Economics.
61 Cross, Worktime; Cross, A quest; Fridenson, Reynaud, La France et le temps. Schmiechen,
Sweated Industries.
62 Ariès, Centuries.
 introduction: labour, coercion, and economic growth 15

this ­relativism does not rule out identifying similarities and differences
as well as continuities and breaks between these definitions and child
labour. The differences concern the status of children relative to that of
their parents: the children of slaves were not always or necessarily slaves
themselves. Conversely, pawnship, i.e. the pawning of children by debtors,
was a form of non-hereditary belonging, at least from a formal standpoint.
Above all, unlike slavery in the strict sense, pawnship expressed a form
of integration in society rather than the alienation of a slave in another
society.63 Hereditary status influenced the evolution of families and soci-
eties before, during and after Western expansion. Slavery as practised
by Westerners in Africa led first to an intensification of local forms of
bondage. Later on, when the Westerners themselves decided to prohibit
pawnship, which was considered slavery, new forms of dependency were
introduced in various African societies. This interaction with the Western
world was essential: the status of children in eighteenth century England
was linked to the authority of the head of the household, the employer
and children’s rights. Children—as children or as apprentices—had fewer
rights in relation to their fathers/employers or their masters outside the
family. These unequal rights accompanied and supported the industrious
revolution64 and later the industrial revolution in the strict sense, both
in Europe (chapter by Deakin) and in Japan (Nagata). Child bondage was
not incompatible with the growth of the industrial world and that con-
nection was transplanted to the colonies, where slaves were immediately
treated like children with fewer rights. When slavery was abolished, first
the British (1832–1842), then the French (1848–1860) and towards the end
of the nineteenth century the Dutch (chapters by Bosma and Campbell)
assigned temporary “apprentice” status to their former slaves.65 This status
was naturally used to demonstrate that former slaves were not fully civi-
lised, but also and more precisely, the fact that, as apprentices in Europe,
they were still subject to the authority of their masters/employers. The
special status attributed to children and former slaves underpinned the
encounter between European countries and Asian and African societies in
the nineteenth century. These worlds had a reciprocal influence on each
other, which led to perpetuating a special status for children within the
societies concerned.

63 Lovejoy, Fayola, Pawnship. Lovejoy, Transformations.

64 De Vries, The Industrious.
65 Davis, The Problem; Engerman, Terms; Lovejoy, Transformations.
16 alessandro stanziani

Women, too, were subject to unequal rights. Keeping in mind the

fundamental differences between matrilineal and patrilineal societies
underlined by Laslett, Meillassoux and many others,66 it is nevertheless
important to think about how these forms of hierarchy interacted and
how people were placed in situations of dependency or slavery. In sev-
eral African and Asian societies, even before the arrival of the Europe-
ans, female slaves were sold for a specific use. Concubines and servants
certainly played a role, but the rise of rural and later plantation slavery
changed the context. The type of crops influenced the choice of male or
female slaves, and hence, the price of women. Once again, the status of
women was related to the form of economic activity in the broad sense of
the term in both “free” and slave-holding societies. In Europe, women had
fewer rights than men and these inequalities were intensified rather than
reduced by the advent of urbanisation and industrialisation in the eigh-
teenth and nineteenth centuries. The diminished rights of women work-
ers, not to mention widespread workhouses in Britain as well as France,
Germany and Russia, testify to these continuities.67 This trend was not
reversed until the twentieth century, and even then, mainly during the
second half.68
In view of these facts, it is possible to conclude that, not only in the
seventeenth but also the eighteenth and much of the nineteenth centu-
ries, the continuing importance of cottage industries and frequent migra-
tion between the city and the country created a strong link between two
types of constraint: presence at work and compliance with working hours,
on the one hand, and competition among employers (including heads of
households), manufacturers, landowners and trader-entrepreneurs for the
control and appropriation of labour, on the other. This is where presence
at work and discipline encountered institutions and labour law. Along
with the rules governing workshops, agricultural estates and plantations
in Europe and Asia, a set of provisions was devised to control mobility
and presence at work, such as the worker’s booklet, laws against poaching
workers and begging, forms of bondage, etc. Constraint in the organisa-
tional sense became linked to institutional constraints; presence at work
conveyed concerns about internal organisation, competition and public

66 Meillassoux, Anthropologie; Laslett, The world.

67 On France: Perrot, Les femmes; on Britain: Burnette, “An Investigation”; on the USA:
Goldin, “The Quiet Revolution”; on Russia: Alpern Engel, Women in Russia; on Germany:
Biernacki, The Fabrication.
68 Orloff, “Gender and the Social Rights”.
 introduction: labour, coercion, and economic growth 17

order. Yet if this was the case, what were the specificities of each area and
how did they enter into global dynamics?

Empires of Eurasia: Global Dynamics/Centres of Development

Why Eurasia? We chose to focus on this area for a number of reasons,

first of all because studies of the evolution of labour in “Western” Europe
has been excessively dominated by a Eurocentric approach that views the
industrial revolution and the French Revolution as the major breaks. On
the contrary, we would like to show that those turning-points were merely
partial, at least as far as labour institutions were concerned, and that the
dynamics at work in France, England, Russia and Europe in general can be
grasped only in their interaction with other parts of the world.69
Our decision to focus on Eurasia and only in one paper (Campbell) on
Eastern Africa in no way indicates a lack of interest in the rest of Africa
and the Americas. The choice was partly based on the existing bibliogra-
phy pertaining to slavery and labour. Researchers, especially in the last
few years, have written at length about the interaction between Europe
and the Americas and between Europe and Africa, whereas the relation-
ships between Europe and Asia have received relatively less attention.
Examining Europe and Asia together also has the advantage of avoiding
retrospective thinking about Europe; we will see that labour institutions
and practices in Europe were connected to what was happening in its
colonies and in Asian empires.
Of course it was necessary to select particular areas. We have included
a great number of regions of Asia and Europe: Russia, Great Britain,
France, Prussia and Austria, India, Japan, China, Indonesia and the Indian
Ocean region extending from China to Eastern Africa. These choices were
made not because the regions are statistically representative, but rather
because they are especially relevant to the questions we are asking. Thus,
the French case is of interest not because it was the land of Colbertism
opposed to liberal England, or because nineteenth century France was the
country of free, codified law compared with Germany, which still lagged
behind. On the contrary, France is of interest because its labour norms in
the nineteenth century were actually quite well suited both to a capitalist
economy and to the heritage of the Old Regime. This case is all the more

69 Pomeranz, The great.

18 alessandro stanziani

likely to raise new questions as, contrary to a widespread preconcep-

tion, common law in England was in fact accompanied by a considerable
degree of regulation and state intervention and labour remained subject
to criminal constraints until the end of the nineteenth century (chapter
by Deakin). Highlighting the case of France and comparing it with Eng-
land leads us to question the differences between liberalism and regula-
tionism or between free labour and guilds within the capitalist world, and
from there to narrow the distance separating so-called “free” labour from
the varieties of bondage.
We have paid special attention to India, in particular, because the time
is ripe for a new analysis of the forms of dependence on England and
its Empire specifically based on labour status. Unlike traditional analyses,
these new studies, including the excellent examples presented here, are
drawn from archival sources and provide a concrete picture of how norms
and the labour market actually worked. We will show that while British
norms and perceptions translated into various forms of bondage and slav-
ery in India, and thereby helped perpetuate slavery well after its official
abolition, the latter nevertheless existed prior to any British interven-
tion. The solution adopted in India and the practices that were accepted
did not result solely from British influences, but rather from interaction
between those influences and local traditions. From this point of view,
the chapters presented here by Joshi and Markovits concur with those of
Campbell on the Indian Ocean and Bosma on Indonesia as well as with an
increasingly extensive bibliography demonstrating the complexity of non-
Western area “dependence” in the nineteenth century, thereby explaining
the rise of proto-industry and then industry in India from the nineteenth
century to the present day.70
Finally, the case of Japan (chapter by Nagata) stands out because the ste-
reotyped image of its economic dynamism as a reproduction of the West-
ern model adapted to its own institutions and local culture is countered
here by an analysis that emphasises the specificity of Japanese capitalism
within a global dimension. Proto-industrial growth in the eighteenth and
nineteenth centuries and the role of the family in those dynamics were
part of a more general trend of the worldwide economy.
The French, Japanese and Indian examples, with the similarities and
differences we have just mentioned, are put to the test by comparing them
with other regions in Asia and Europe. The Russian empire is interesting

70 Roy, The Economic History.

 introduction: labour, coercion, and economic growth 19

(Stanziani): although serfdom and its history have been abandoned for
decades, they constitute a kind of intellectual icon, apparently unshake-
able in spite of new studies on slavery and on serfdom in Eastern Europe
and above all despite new data in Russian economic history revealing
considerable economic and demographic growth at the time of serfdom.
We explain this data by showing that Russian serfdom was never actually
institutionalised and was in fact more flexible and more complex than is
usually assumed.
These conclusions coincide with those regarding Central and Eastern
Europe where the evolution of institutional relations, social relationships
and economic dynamics did not wait for Napoleon to adopt a market ori-
entation. As the chapter by Markus Cerman shows, there was sizeable
proto-industrial growth at work in these regions.
Once serfdom—especially the second serfdom—has recovered its place
within the comparative history of forms of labour, we will take another
look at the differences and similarities in relation to other possible con-
figurations of the labour world. Within the structure of this book, China in
particular seems to pose a problem.71 The chapter by Chevaleyre demon-
strates the specificity and complexity of the forms of bondage and the spe-
cial legal status of bonded people compared with the familiar social and
legal forms in the West during the same period. Like the chapters devoted
to other contexts, the one on China attacks the traditional breaks pro-
posed by historians. In particular, it re-examines the transition from the
Ming dynasty to the Qing dynasty around 1644 precisely from standpoint
of the status and condition of serfs, which encompassed various situa-
tions ranging from servitude and debt bondage to hereditary slavery. The
continuities exceed the so-called changes in Chinese history. Indeed, his-
torical research72 has underlined a Chinese specificity compared with the
rest of the Eurasian continent: the studies show that slavery and the most
restrictive forms of bondage were weaker and eliminated sooner in China
than in the West. This outcome is often explained by the demographic
factor, namely, the abundance of available labour in the country.73 This is
an altogether traditional, rather widespread argument in the literature on

71 I acknowledge my debt to Kenneth Pomeranz, Christine Moll-Murata, Yves Chevrier,

Christian Lamouroux, Choi Wai for having helped me on this point.
72 Zurndorfer, Change; Hansson, Chinese Outcast; Campbell, Lee, “Free and unfree”;
Cartier, “Travail; Schottenhammer, “Slaves”.
73 Pomeranz, The Great, 82–84.
20 alessandro stanziani

the historical forms of forced labour. Following Herman Nieboer,74 count-

less scholars have explained forced labour by population scarcity.75 This
argument is now being challenged by historians of serfdom and slavery
and of wage labour. The latter have shown that coercion in wage rela-
tions increased rather than declined between the early 1750 and the mid-
nineteenth century Britain, at a moment when the population growth was
already perceptible.76
On the other hand, the cases of Australia and Canada testify to the
fact that the colonisation of new territories did not necessarily call upon
massive imports of slaves, as in the United States.77 To be sure, in Aus-
tralia, convicts initially played an important role but their number were
small compared to forms of indentured and contract labour immigration.78
These last forms also prevailed in Canada in the eighteenth and early
nineteenth centuries. Although “free” and “unfree” immigration can no
longer be opposed as such,79 and conditions for indentured immigrants
were harsh (in particular for Asian immigrants during the second half of
the nineteenth century), they still cannot be assimilated to slaves because
their legal status was never passed on to their children and they could
always return home. However difficult their situation, they did not endure
the radical exclusion of slaves.80 Thus, while Canada and Australia were
not lands of freedom, they never experienced massive slave immigration
like Brazil or the United States. Indeed, for indentured contracts in gen-
eral, the forms of immigration to Australia and Canada seem to have been
linked far more to immigrants’ motivations than to population scarcity.81
Last but not least, the case of Russia (chapter by Stanziani) also shows
that the introduction and implementation of serfdom rules was never
connected to population density and the availability of labour.
Thus, from a reverse angle, it seems legitimate to have doubts on this
subject with regard to China as well. In comparison to the rest of world,
the demographic rise in China, already considerable in the sixteenth and

74 Herman Nieboer, Slavery as an Industrial System. Cambridge: Cambridge University

Press, 2010 (Original: 1900).
75 Just a few example in a huge bibliography using this argument: for Russia: Blum,
Lords; for the USA: Fogel, Engerman, Time.
76 Deakin, hereafter.
77 Engerman, Terms, “Introduction”.
78 Northrup, Indentured Labor.
79 Lucassen and Lucassen (eds.), Migration. Eltis, Coerced and free migration.
80 Northrup, Indentured Labor.
81 Galenson, White servitude.
 introduction: labour, coercion, and economic growth 21

seventeenth centuries, even accelerated in the eighteenth century.82 At

the same time, there is no proven correlation between the increase of
population and the abolition of bondage. There were significant regional
differences in mobility throughout the nineteenth century. In particular,
though restrictions on the movements of a portion of the rural population
were soon abandoned—in certain regions and for certain peasants, espe-
cially those on government-owned estates—, constraints nevertheless sur-
vived for a long time, at least until the end of the nineteenth ­century.83
In general, a kind of caste separation between liangmin (good people)
and jianmin (lowly people) continued until the nineteenth century. These
different statuses were hereditary; the category of lowly people included
servants, prostitutes and musicians as well as craftsmen and wage earn-
ers. These categories enjoyed fewer rights than the other social groups.
They were also subject to more severe criminal punishments and greater
discrimination in labour relationships.84
In this context, debt bondage, bonded servants and rural labourers
were quite widespread, not to mention slave prostitutes.85 Though there
are no hard figures to quantify this phenomenon, legal disputes on the
subject suggests it was a considerable problem throughout the nineteenth
century.86 And even until the late twentieth century, serfdom and rural
slavery were widely practised by the Yi in the mountains of Sichuan and
Yunnan.87
In the cities, too, the heritage of the guild system, which guaranteed
fair competition between employers just as it did in Japan (chapter by
Nagata), conditioned the mobility of labourers by laying down rules of
apprenticeship and remuneration.88
All the same, it would be worth doing detailed studies on the forms
of labour in China in the eighteenth and nineteenth centuries until the
beginning of the twentieth century to determine whether forms of con-
straint and bondage were maintained or actually disappeared.
In short, it is possible to identify similarities and differences in work
constraints in Eurasia. The similarities lie mainly in the dissemination of

82 Lee, Feng, “Malthusian”.

83 Campbell, Lee, “Free and unfree”.
84 Hansson, Outcasts.
85 Schottenhammer, “Slaves”.
86 Hansson, Outcasts.
87 Hui, “Les esclaves”.
88 Moll-Murata, “Chinese Guilds”.
22 alessandro stanziani

those constraints from the end of the seventeenth century to the early
twentieth century, along with an intensification of labour during peri-
ods of agricultural, trade, proto-industrial and industrial growth. Labour
intensification in support of growth, rather than a workforce shortage as
such, accounts for the constraints. Continuity in the forms of labour
regulation during this period was particularly evident in Japan and Great
Britain. The differences concern primarily the strength of proto-industry,
which was greater in Asia than in Europe, but also the early liquidation
of the forms of bondage in China. This liquidation, which no doubt
should be qualified, was in any case accompanied by a strengthening
rather than the disappearance of the forms of infra- and inter-village
solidarity, especially during migrations to the city. At the time, the same
forms of solidarity were found in Russia, which, unlike China, neverthe-
less maintained “serfdom” well beyond the period of eighteenth century
growth.
Finally, the forms of labour in Japan stood in stark contrast to these
models and were equally far removed from the solution adopted in West-
ern Europe around the end of the nineteenth century, i.e. the social state.
Japanese enterprises encouraged workforce loyalty and set up internal
systems of social protection, both of which were among the main arrange-
ments developed to support the transition from small to large industry.
At the same time, these comparisons at the national and imperial level
are valid only as a rough approximation. No doubt legal rules (civil, tax
and customs laws) refer to the national dimension of these phenomena,
yet those rules were only one of the components of economic action, along
with symbolic, cultural and political aspects. Hence, we cannot ignore the
importance of local components and the strong differences between the
dynamics of different regions within a single country. Let us look at that
dimension now.

Global versus Local Labour Dynamics

The forms of forced labour—the existence of bondage or even slavery

alongside “free” labour—often varied from one city to the next and from
one place to another. This observation is especially relevant in our case
as the institutions and economic activities in the world we are study-
ing were extremely fluid, multiple and local from the eighteenth to the
early twentieth century. Several institutions coexisted at the local level
and even when a process of national unification took place, institutional
 introduction: labour, coercion, and economic growth 23

­ luralism continued.89 It was more widespread at the level of empires

p
where legal pluralism was an important instrument of economic and
political action.90
Local practices and customs played an important role and they were
recognised in nineteenth century Russia with regard to property, in Islam
with regard to transferring ownership of slaves, in Kyoto and Lyon with
regard to work discipline. These elements account simultaneously for
common phenomena (restrictions on labour mobility), the diverse ways
they were expressed and their source (worker’s booklet, Russian serfdom,
criminal punishment in the British Empire). They also explain the dif-
ferences in the dynamics of Lancashire, the south of France and western
Russia as well as those between individual English or Japanese factories
and villages. Different solutions were adopted within a few miles of each
other and similarities developed more frequently with factories in dis-
tant regions than with those nearby. The solutions adopted concerning
working time testify to local irreducibility within a space that was nev-
ertheless global. The chapters by Nagata and Vernus confirm this phe-
nomenon. In Japan as in France, despite national norms and theoretically
common techniques, the solutions that were adopted differed from one
region to the next, often significantly. On the other hand, surprising simi-
larities can be detected between certain districts of these two countries.
These results confirm the similarities between certain Chinese regions
and English districts, which have been the subject of recent research,91
just as those between proto-industrial districts in Europe.92 These resem-
blances express the close link between technical solutions, economic
and local institutional heritage, national discipline and international
phenomena.

Book Outline

This book is divided into three parts, reflecting the coexistence of sev-
eral levels of analysis. The first part reveals how the circulation of eco-
nomic knowledge and legal models was related to economic behaviour
and dynamics through case studies of England, Indonesia and the Indian

89 Royer, Histoire.
90 Benton, Law.
91 Pomeranz, The great.
92 Cerman, Ogilvie, European.
24 alessandro stanziani

Ocean World. Simon Deakin shows that, contrary to its reputation as the
country of common law and free labour, Great Britain multiplied the stat-
utes and constraints governing labour, unlike France where guild-based
constraints proved less important than is usually acknowledged. Thus
it becomes easier to understand the interface between the categories of
European colonisers and those at work in the colonies. Ulbe Bosma dis-
cusses the attempts by Dutch colonial elites to establish certain labour
relationships in Indonesia using the specificity of that region compared
with their homeland as a lever. Similarly, the chapter by Gwyn Campbell
takes a very long-term view in examining the forms of labour, bondage
and slavery in the different societies of the Indian Ocean. They contrib-
uted to the proto-industrial growth experienced in other regions, which
raises the question of their different trajectories over the very long term,
even though the dynamics in which they took place were identical.

The second group of articles examines in greater detail the way local forms
of dependence interacted with global dynamics. Marie-Louise Nagata
explains how families and production units in Japan under the Tokugawa
regime coped with proto-industrial growth. Taking the example of silk
production in Tokyo, she brings out the interaction between family ties,
the market and working time. This serves as a basis for comparing its simi-
larities with and differences from Lyon, another silk production region.
Pierre Vernus studies the solutions developed in this industrial district in
the face of rapidly growing international markets in the nineteenth cen-
tury and the undermining of local practices concerning labour and prod-
uct quality. Claude Markovits looks at these same constraints in Indian
trading companies. This case opens up at least three interesting perspec-
tives. First, it fills a gap in the study of labour in India, which is too often
focused on manufacturing. Next, it shows that Indian emigration was not
necessarily limited to coolies and bonded labourers but also involved
trading networks. Third, the presence of Indian trade organisations at the
international level reflected international economic dynamics in which
Asia, far from being a mere spectator of European development, in fact
played an active role.93 In his study of silk marketing by Indian firms,
Markovits provides an important supplement to the chapter by Vernus
by bringing out the connection between the dynamics of silk production

93 Chaudhury, The Trading; Tracy, The Rise.

 introduction: labour, coercion, and economic growth 25

in nineteenth century Asia and the rise of local trade networks that were
solidly integrated in the international markets.
Taken together, the three cases of Kyoto, Lyon and an Indian trading
firm in Cairo testify to the importance of forms of dependence rooted in
family networks, which was to a certain extent common to Europe and
Asia. This aspect is the core of the third and final part of the book, devoted
to what Braudel described as “long-term prisons”. We will be discussing
these constraints more from the standpoint of labour relations than of
food. We have selected four cases: Central-Eastern Europe, Russia, China
and India. The first two cases lead to questioning the “second serfdom” in
Central and Eastern Europe, and from there, to re-examining the slow evo-
lution of these societies. Instead of a shift from feudalism to capitalism, in
keeping with the favoured tradition of theoreticians and historians, we will
emphasise the interaction over the very long term between forms of bond-
age, constraints and market dynamics. The obvious differences between
these situations, which we have already underlined, cannot conceal the
common problem they all faced, namely economic and territorial expan-
sion and the conflict between elites that it produced. From this point of
view, the polymorphous legal status of “serfs” corresponded to the differ-
ence in status separating servants, coolies, shop assistants and agricultural
labourers from their masters. The status of the different forms and degrees
of service in husbandry and bondage reflected different economic condi-
tions. Work discipline and its organisation were never a simple contrac-
tual matter but rather took place at the intersection between the private
order and the public order of the markets. That is why it is important to
compare the evolution of the second serfdom with that in China and later
in colonial India. In the case of China, we will also learn in the chapter by
Claude Chevaleyre that the existence of a multitude of forms of depen-
dence ultimately adapted quite well to the order of civil servants, “castes”
and naturally the markets. In the end, these elements will be compared
with the forms of constraint in colonial India, particularly in the area of
public works (article by Chitra Joshi); to meet workforce requirements,
the British government resorted as readily to (Indian) criminals as to the
“free” population. Underlying these measures was the notion of labour
as social service. This connection was possible due not only to the hazy
distinction in English law between a servant and a bonded labourer, but
also to the fact that in Mogul villages in India, peasants were traditionally
hired for public works projects (road construction, labour during periods
of famine, etc.). It was the convergence of these different traditions that
made it possible to mobilise compulsory labour to build imperial roads.
26 alessandro stanziani

The same convergence of private and public interests—of the state and
the village economic community—, already noted in France and Russia,
accounts for the emergence and duration of the system.
We have thus come full circle: we find the link between freedom and
constraint discussed in the first section in the organisation of labour
and the movable boundary line between the private order (contractual)
and the public order (regulatory and legal) of labour.
 Part one

Juridical Models and Labour Dynamics

 The Duty to Work: A Comparison of the Common Law
and Civil Law Systems from the Eighteenth to the
Twentieth Centuries

Simon Deakin*

Introduction

Through the concept of the ‘contract of employment’, the law defines and
regulates the individual employment relationship within the context of a
market economy. The idea that, in a capitalist economic framework, work
relations should take a ‘contractual form’ is nevertheless a surprisingly
recent one. In the first stages of industrialisation in Europe and north
America, work relations were not uniformly ‘free’ and the obligations of
the parties were not, on the whole, defined using the language of contract.
The model of employment that we are familiar with today only covered a
small part of the working population, those who were ‘salaried’ managers
or professionals. The story of how, from these beginnings, the model of
the contract of employment came to provide the foundation for the legal
account of work relations is of interest not just from the point of view of
legal history and comparative law, but for the insights it provides into the
evolution of law under industrial capitalism.
In both the common law countries and those of the civil law, the ‘con-
tract of employment’ that we know today is essentially an invention of
the late nineteenth and early twentieth centuries. It emerged alongside
the large, integrated industrial enterprise and the modern welfare state.
Nevertheless, the process took different forms in various national systems.
Each legal system developed its own distinctive concept of the employ-
ment relationship, based in part on its own juridical culture, as well as on
the social and economic context provided by the path it took to industri-
alisation. The evolution of labour law both reflected and reinforced these
national differences.

* University of Cambridge. This paper builds on and extends my work with Frank
Wilkinson, first published in our joint work, The Law of the Labour Market (Oxford: OUP,
2005), and my chapter, ‘The Comparative Evolution of the Employment Relationship’, in
Guy Davidov and Brian Langille (eds.) Boundaries and Frontiers of Labour Law (Oxford:
Hart Publishing, 2006). I am grateful to the publishers and editors of these earlier works
to draw on them here.
30 simon deakin

The Evolution of Concepts in Labour Law

The ‘contract of employment’ (in English law: the equivalent of ‘contrat de

travail’ in French labour law) is a highly complex amalgam of many differ-
ent kinds of norms and values surrounding the employment relationship.
We should not expect the lawyer’s term ‘contract of employment’ to have
a point-for-point correspondence with the employment relationship as it
is observed and experienced in a social or economic context. The legal
concept must obey certain internal rules of system-congruence which are
specific to the juridical order.1 At the same time, the legal system will
function well or badly according to how far it can increase the volume
of external information which it can handle, while maintaining the inter-
nal mechanisms (specifically legal acts and processes) through which its
reproduction is ensured. In that sense, the capacity of the legal order to
process the information which it receives from the economy is in issue.
If the legal concept of ‘employment’ were to bear no relationship at all to
the ‘economic’ one, its functioning would, at some point, be impaired.
This leads on to the observation that in the process of evolution of legal
concepts, we simultaneously observe continuity and change. Adaptations
are explained by a reference back to an existing conceptual maxim or
principle. The rate of conceptual evolution is therefore much slower than
the rate at which the content of the rules themselves changes. We can,
nevertheless, identify conceptual mutations, in particular during those
periods of rapid change when one paradigm displaces another in the
manner of ‘punctuated equilibrium’.2
The interpretive task of ‘decoding’ of legal concepts has the effect of
highlighting the contingent nature of implicit legal values; they can be
seen to be the result of historically specific circumstances, which may
no longer hold.3 But at the same time, the reshaping of the law which
accompanies attempts at reform virtually never starts from a blank sheet.
Concepts already available are put to new uses. This does not mean that
the legal forms themselves to do not change; however, they change much
more slowly than shifts in the substance of the law, and in ways which
reflect the previous pattern of development. Legal evolution, then, is
­path-dependent. Conceptual adaptations are piled one on top of another,
with the result that the structure of legal thought at any given point in

1 Luhmann, Law as an Autopoietic System.

2 Eldredge and Gould, “Punctuated equilibria”.
3 Atleson, Values and Assumptions.
 the duty to work 31

Table 1 Legal Classifications of Work Relationships from the Eighteenth

Century to the Mid-Twentieth Century
1600–1800
Servant Worker (typically unmarried) engaged in service
under a yearly hiring, entitled to payment in cash or
in kind whether or not there was work during the
period of the hiring, and with a right to a poor law
settlement after the hiring ended
Labourer Daily or casual manual worker in agriculture or the
unregulated trades
Master, journeyman, Worker in trades protected by guild regulation
apprentice
1800–1875
Servant Manual worker in industry or agriculture under the
disciplinary regime of the master and servant legis-
lation, with little security or wages or employment
Employee Clerical, managerial or professional worker outside
the master-servant regime, with a degree of contrac-
tual income and employment security
Independent contractor Independent artisan outside the scope of master
and servant legislation
1875–1950
Workman Manual worker subject to the semi-disciplinary
provisions of the Employers and Workmen Act 1875,
increasingly protected as the period went on by col-
lective bargaining, workmen’s compensation and
social insurance legislation
Employee At the beginning of the period, a non-manual
worker with managerial or professional status; by
the end of the period, a wage or salary-dependent
worker, either manual or non-manual
Self-employed Independent worker not employed under a contract
of employment

time incorporates forms which, although in some sense superseded, nev-

ertheless continue to shape the path of the law.
Table 1 indicates the main classifications of labour relationships in Eng-
lish law at three different periods in its historical development.4 From this

4 See further, Deakin, “The contract of employment” and ‘‘The personal employment
contract”.
32 simon deakin

it can be seen that the modern distinction between the ‘employee’ and
the ‘self-employed’ worker (which is roughly equivalent to that between
‘travail subordonné’ and ‘travail indépendant’) is a relatively recent one.
It did not exist in its modern, paradigmatic form until the 1950s. Prior
to that point, elements of the same idea can be found, but there were
equally powerful divisions in the law, in particular between ‘servants’ and
‘employees’, a distinction drawing on notions of class and status as much
as function.
A ‘genealogical’ analysis helps us see that the modern employment con-
tract has been superimposed on an older notion, the ‘master-servant’ rela-
tion, which had few of the features which compensate today’s employee
for entering into a situation of personal and economic dependence upon
the employer. The key to understanding this process lies in the analy-
sis of the legislation and case-law surrounding the terms ‘servant’ and
‘employee’. At the mid-point of the twentieth century, it was believed
that the prevailing legal concept for defining the employment relationship
during the nineteenth century had been the ‘control’ test. Kahn-Freund
influentially suggested that the control test developed in the context of
the common law relating to the employer’s vicarious liability for torts of
a servant acting in the course of employment.5 It is certainly true that, in
the post-1945 period, this was one of the contexts in which the control
test was still being applied. However, the nineteenth-century authorities
which were (and are) cited for the ‘control’ test were not concerned with
the issue of tortious liability.
Close examination of origins of the concept of the contract of employ-
ment suggests that it was not the common law of vicarious liability which
provided the context for the emergence of that concept, but, rather, social
legislation dealing with taxation and national insurance. One of the most
widely cited cases for the control test, Yewens v. Noakes,6 concerned the
definition of a live-in servant under tax legislation. This case did not turn
on the distinction between ‘employees’ and the ‘self-employed’ which is
familiar to modern employment lawyers and which Kahn-Freund was
writing about in the 1960s, nor even to the roughly equivalent nineteenth
century distinction between ‘servants’ and ‘independent contractors’. The
court’s decision was based on its refusal to believe that a salaried clerk
earning a substantial salary could be a ‘servant’, since, according to the

5 Kahn-Freund, “Personal scope”.

6 (1880) 6 QBD 530.
 the duty to work 33

court, such a person was more clearly akin to ‘the manager of a bank, a
foreman with high wages, persons in the position almost of gentlemen’.
Yewens v. Noakes, then, was not concerned with the modern distinction
between employment and self-employment for which it is still, even today,
being cited. Rather, it was concerned with a completely different status-
based divide, that between ‘servants’ and labourers in manual employ-
ment, on the one hand, and those employed in higher-level occupations
and managerial and clerical work, on the other.
To see why this distinction was more important at that point than it is
today, a now-vanished feature of the nineteenth century legal landscape
must be borne in mind. The distinction between manual and non-manual
work had been central to the operation of the nineteenth century master-
servant legislation; only ‘servants’ were subject to fines and imprisonment
for breach of service contracts. This same distinction was carried over as
a kind of ‘frozen accident’ into early social legislation concerning work-
men’s compensation and social insurance. It was in this context of the law
of the emerging welfare state that the ‘control’ test was established in a
series of early twentieth-century decisions.7
Why did twentieth-century courts light upon the otherwise obscure
decision in Yewens v. Noakes? The (re)discovery and adaptation of the
control test in the 1900s and 1910s was a doctrinal innovation which was
introduced at the same time as the courts were being called on to define
the boundaries of what was then regulatory legislation of a wholly novel
type. Nor was this judicial innovation particularly welcoming to the new
legislation. The element of compulsion in social legislation went strongly
against the grain of prevailing common law values. As a result, the courts
regularly held that professional and managerial workers were outside the
scope of these new laws.8 The control test, as applied by twentieth cen-
tury courts, was also linked to disputes about employer’s liability in the
context of the widespread practice of internal contracting. The contract
system of hiring labour through an intermediary was still the predomi-
nant form of industrial organisation in road building, construction, ship-
building, mining and quarrying, and iron and steel. The adoption of the

7 Simmons v. Heath Laundry [1910] 1 KB 543; Scottish Insurance Commissioners v. Edin-

burgh Royal Infirmary 1913 SC 751; Hill v. Beckett [1915] 1 KB 578; Underwood v. Perry [1923]
WC & I Rep. 63. The relevant statutes were the Workmen’s Compensation Acts of 1897 and
1906 and the National Insurance Act 1911.
8 See, Waites v. Franco-British Exhibition (1909) 25 TLR 441; Bagnall v. Levinstein [1907]
1 KB 531; Dow v. McNeil [1925] WC & I Rep. 32; re South Dublin Union Officers [1913] WC
& I Rep. 245.
34 simon deakin

control test enabled employers to avoid responsibility for the social risks
of illness, injury and unemployment which it had been the aim of social
legislation to impose, at least in part, upon them. In short, the rise of the
control test in the early years of the twentieth century tells us much about
prevalent employment disputes at that time clash between freedom of
contract and the welfare state, and about shifts in the structure of the
business enterprise which were also going on at this time.
The ‘unitary’ model of the contract of employment which came to
extend to all categories of wage-earners, including salaried and clerical
workers, was only clearly adopted when further reforms were enacted to
social legislation, in particular the extension of social insurance which
took place in the National Insurance Act 1946, and when the internal
contracting system gave way to integrated management, a development
which in some industries, such as coal mining, occurred as late as the
post-war nationalization process. A major aspect of the Beveridge Report
of 1942 was the abolition of distinctions between different categories
of employees: henceforth, all wage or salary earners, regardless of their
annual income or of their professional status, would come under the same
contributory classification.9 It was in the context of this new situation that
the courts abandoned the old distinction between low status and high sta-
tus employees when seeking to define the contract of service.10 The con-
trol test itself came to be regarded as excessively artificial, and gave way to
the more recognizably modern tests of ‘integration’ and ‘business reality’.
At around the same time, the term ‘servant’ mutated into the modern
‘employee’. By these means, a more inclusive notion of the employment
relationship came to be established for the purposes of determining the
scope of employers’ liabilities in respect of personal injuries, employment
protection and social insurance.
To understand this process, it is necessary to go back to the early mod-
ern roots of labour regulation in the British context. In the following sec-
tion, the role of master-servant legislation in the British case is considered
in some detail, with a brief account of the American case providing a
point of contrast.

9 Social Insurance and Allied Services Cmd. 6404, November 1942, at para. 314.
10 See, in particular, Stevenson, Jordan & Harrison v. McDonald & Evans [1952] 1 TLR 101.
 the duty to work 35

The Evolution of the Contract of Employment in the Systems of the

Common Law: England and the United States

England: The Emergence of the Master-Servant Model

The institutional origins of the market economy in the case of England
can be traced to the later middle ages. At that point, England possessed
a unified legal system that operated on a national scale. The first laws on
the work relationship, adopted in the wake of the Black Death of 1346,
demonstrated the growing importance, at that point, of legislation as a
mechanism of economic control. Their passage also indicates the pres-
ence of a system of labour contracting, operating alongside the compul-
sory labour of serfs and villeins.11 The Statute of Labourers of 1351 did not
just formalise the regulation of wages; it triggered the development of the
action of action of assumpsit, forerunner of the modern law of contract.12
By the fifteenth century, as the role of the Church as a source of social pro-
tection was disappearing, the regulatory role of the state expanded, with
the passage of numerous poor law statutes and of the Statute of Artificers
of 1562. This statute was to provide the basis for wage fixing, the consti-
tution of the urban guilds and the regulation of the agricultural labour
market into the nineteenth century.13
Formally, labour under the Statute of Artificers was not ‘free’: service
in agriculture was obligatory for those who had no independent means of
subsistence. Within the urban guilds, relations between masters, journey-
men and apprentices more closely resembled those of producers subject
to the common rules of the trade, than they did those of capitalist entre-
preneurs and subordinated workers. The putting-out system consisted of
networks of independent contractors, linked to merchants and other com-
mercial intermediaries, rather than a class of wage-dependent labourers.
Thus we have to exercise great care when considering the meaning given
to the terms ‘labourer’ and ‘servant’ at this time.
The century after 1750, generally associated with the period of the
‘industrial revolution’, was a time of legal transformation to match that of
the changes which were taking place in the wider society and economy.

11 ‘As villeinage ceases, the poor law begins’ was Tawney’s assessment (Tawney, The
Agrarian at p. 47), and the point can be extended to wage regulation more generally.
12 See Palmer, English Law in Age of Black Death.
13 See Deakin and Wilkinson, Law of the Labour Market, 44–51.
36 simon deakin

For Arnold Toynbee, writing in 1870, the essence of the industriali rev-
olution was the displacement of medieval regulation by the forces of
­competition.14 Competition in the labour market was marked by the abo-
lition of the wage-fixing clauses of the Statute of Artificers in 1813 and
by the repeal of the apprenticeship laws in 1814. The Statute was by no
means a dead letter at this point. The repeal of the apprenticeship laws
was accompanied by a series of strikes. Prior to the final repeal, the judges
had been weakening the law through a series of restrictive interpretations
of its provisions.15 The willingness of the eigthteenth century courts to
uphold guild rules protecting the ‘mystery’ of the trade16 gave way to judi-
cial astonishment that the apprenticeship laws might be applied to the
owners of the newly emerging industrial firms.17 By these legal means,
a major institutional constraint on capitalist forms of enterprise was
removed.
Even this was only a step towards the ‘contractualisation’ of work
relations. In certain professional and managerial occupations, a type of
employment contract emerged, into which the judges began to imply a
number of terms which reflected the idea of reciprocal obligations of the
parties. At the start of the nineteenth century we also see the growing rec-
ognition by the courts of the common law action for damages for wrong-
ful dismissal. This contractualisation of the employment relationship did
not however apply to workers falling under the statutory régime of the
Master and Servant Acts: in their case, breach of contract was a crime,
under which thousands of workers were successfully prosecuted every
year between the 1820s and the 1870s.
The first of the Master and Servant Acts was enacted in 1747. This Act
gave the justices jurisdiction to examine and rule on disputes between
masters and servants in husbandry and between masters and ‘artificers,
handicraftmen, miners, colliers, keelmen, pitmen, glassmen, potters and
other labourers employed for any certain time, or in any other manner’
whether or not any rate or assessment of wages had been made for them
in that year. They were also supplied with the power to order payment of
wages due and to punish the servant or labourer for any ‘misdemeanour,
miscarriage or ill behaviour’ by abating wages or committing him to the

14 Toynbee, Lectures on the Industrial Revolution, 92.

15 Prothero, Artisans.
16 Hobbs v. Young (1689) 1 Show KB 267.
17 Kent v. Dormay (1811), Kingston Assizes, August 14 reported in Chitty, A Practical
Treatise, 122.
 the duty to work 37

house of correction for up to a month; they could also discharge the ser-
vant from the contract. An Act of 1758 extended their jurisdiction to cover
servants in husbandry hired for less than one year18 and that of 1766 made
it an offence for the servant to quit before the end of the agreed term.19
This last provision was an attempt to bring up to date the similar prohibi-
tion in section 15 of the Act of 1563. An Act of 1823 established new crimes
of absconding from work and refusing to enter into work under a con-
tract of hiring, and provided for imprisonment of workers for up to three
months.20 Thus far from abandoning this particular part of the Elizabe-
than labour code, therefore, Parliaments of the eighteenth and nineteenth
centuries significantly strengthened and extended it.
Complementing the Master and Servant Acts were numerous mea-
sures passed specifically to deal with theft and embezzlement by servants,
labourers and outworkers. At this point, merchants and middle-men who
put out goods for finishing up remained the legal owners of the mate-
rial or produce throughout the process; these laws gave them powerful
procedures for enforcing discipline against the rural manufacturers whom
they normally employed on a nominally independent basis.21 Precedents
for these measures exist from the seventeenth century;22 again, the pace
of legislative change increases after the mid-eighteenth century. Under
Acts of 1740, 1749, 1777, and 179223 it became an offence punishable by
imprisonment for persons employed in manufacturing various goods to
divert or sell materials sent to them for finishing up, or to detain them
for more than twenty-one days after completing the work agreed. Justices
could issue a search warrant for the inspection of premises of those con-
victed or charged of these crimes. Those buying or receiving stolen goods

18 31 George II c. 11, s. 3.

19 6 George III c. 25, s. 4. This was an attempt to impose a general restraint of the
kind previously enacted around this time for particular trades, including the journeyman
shoemakers (9 George I c. 27, s. 4, 1722), woollen manufacturers and framework knitters
(12 George I c. 34 s. 2, 1725), glove and shoe manufacturers (13 George II c. 8, 1739, s. 8).
The 1766 Act was stated to apply to any ‘artificer, calico printer, handicraftsman, miner,
collier, keelman, pitman, glassman, potter, labourer or other person contracting for any
time or times whatsoever’.
20 4 George IV c. 34, ss. 3, 4 and 5.
21 On the so-called ‘proto-industrialisation’ model based on the putting-out to produc-
tion to household forms of labour, see Berg, Age of Manufactures; the evidence for and
against ‘proto-industrialisation’ as a distinct mode of production is assessed by Daunton,
Progress and Poverty ch. 6.
22 See 13 & 14 Charles II c. 15, 1672–3; 20 Charles II c. 6, 1677; 8 & 9 William III c. 36,
1696–7.
23 13 Geo. II c. 8; 22 Geo. II c. 27; 17 Geo. III c. 56; 32 Geo. III c. 44.
38 simon deakin

from servants also committed crimes. An Act of 1800, noting that ‘it often
happens’ that colliers and miners breach their contracts ‘to the great and
lasting prejudice of their employers’, made it an offence for persons con-
tracted to raise coal or minerals to do so ‘in a different manner to [the
owner’s] stipulations thereto, and contrary to the directions and against
the will of the owner’.24
The use of the criminal law to enforce the contractual obligations of
labourers and artisans found ready justification. ‘Imprisonment may be
viewed as a mode of compelling the performance of contracts’, wrote
J.E. Davis in his account of the Master and Act of 1867.25 ‘In some cases,
damages might recompense a master for the breach of a contract by his
servant, but the latter is seldom in a position to pay damages, and there-
fore, in the absence of any other remedy, he might set his employer at
defiance’.26 It was expected that the jurisdiction over employment dis-
putes would be very widely used, and that the regular courts would not
be able to cope:
No substitute . . . can be found for the jurisdiction of the magistrates, stipen-
diary or otherwise, in the case of master and servant. The County Court
does not sit sufficiently often for this purpose. In populous districts it is held
once a month, so that, to say nothing of frequent adjournments of cases
from one court to another, persons would be almost without the means of
having their cases heard for many weeks together, instead of having the
magistrates’ court to resort to every week-day in the most populous districts,
and in rural districts seldom less than once in every week.27
Court procedure was streamlined and, generally, beneficial to the
employer who would normally be the one to bring the complaint. In
complaints of neglect of work, which took the form of proceedings for
a conviction as opposed to an order, the servant could not be a witness
in his or her own defence.28 The Act of 1867 remedied this and replaced
imprisonment with fines as the principal remedy for breach of the Act,
although imprisonment remained a possibility for, among other things,
‘aggravated’ ­misconduct.29 Fines also offered an advantage over discharge,

24 39 & 40 Geo. III c. 77.

25 Davis, The Master and Servant Act 1867, 6.
26 Ibid., 7.
27 Ibid., 8.
28 Ibid., 10.
29 30 & 31 Vict c. 141. The principal remedy envisaged by the Act was compensation for
non-performance of the contract of service (s. 9), but there was the possibility of imprison-
ment for up to three months for non-payment of compensation (s. 11) and for aggravated
 the duty to work 39

which ‘although at first sight a desirable course to adopt, was impracti-

cable as a punishment (and therefore impracticable as a remedy) owing to
the demand for labour in many branches of manufacturers and the ability
of men to get work elsewhere’.30
The courts implied into the contract of service employers’ obligations
to provide work and to maintain the relationship in being through depres-
sions in trade, as the necessary complements to provisions for extended
notice or duration. Without such terms, a worker’s agreement to serve the
employer exclusively for a period of years would be void on the grounds
that it was in restraint of trade. The contract might provide for payment
on the basis of piece rates or time rates. In an agreement for exclusive
service for twelve months with payment on piece rates and provision for
three months’ notice on either side, the court found a
necessary implication that the employer shall find reasonable work and pay
for the articles manufactured . . . The necessity of giving notice clearly shows
that there is some obligation on the part of the employer. What is that? To
find reasonable employment according to the state of the trade. That is not
a unilateral agreement, but a mutual agreement with something to be done
on both sides.31
However, as contract terms for the employee’s protection, these obligations
were more or less notional. In the principal, reported cases of this period,
the higher courts consistently rejected employee’s claims for wages based
on the employers’ duty to find work.32 They found that employer had an
implied right to lay off without wages, even in the case of an annual pit
bond binding the workers to a year’s exclusive service.33 In this sense,
long-service agreements effectively benefited only the employer; the
worker was bound without having the protection of security of income
or employment. The principal purpose of finding mutuality was to trigger
the disciplinary provisions of the Act against the worker, or to form the

misconduct (s. 14). No wages were to by payable during a period of imprisonment (s. 17).
Servants and their spouses were deemed to be competent witnesses by s. 16. On the exten-
sive use, in practice, of powers of imprisonment under this Act, see Steinfeld, Coercion and
Contract, 81.
30 Davies, Master and Servant Act 1867: 10.
31 R. v. Welch, (1853) 2 E & B 356, per Lord Campbell, above; see also Pilkington v. Scott,
(1846) 16 M & W 657.
32 This is in contrast to case law of the late medieval and early modern period in which
claims by servants for wages due in return for being willing to serve frequently succeeded,
via an action on the Statute of Labourers of 1351: see Ibbetson, Historical Introduction, 75.
33 Williamson v. Taylor (1843) 5 QB 175; see Steinfeld, Coercion, Contract, 102–124, for
discussion of this case and other decisions in the same vein.
40 simon deakin

basis for an action by the employer against another employer for entic-
ing away the servant. Moreover, the courts’ construction of the contract
of employment differed according to the statutory context which they
were considering. ‘Butty’ workers—subcontractors in coal mining—were
excluded from the coverage of the protective Truck Acts on the grounds
that they were independent workers, while being simultaneously subject
to the Master and Servant Acts as servants.34 Nor did the concept of mutu-
ality extend to the continuation of the master’s traditional obligations of
care under the service relationship. In this sense, the nineteenth century
model of master and servant had little or nothing in common, beyond
the use of certain terminology, with persistence of reciprocal obligations
which can still be seen in parts of the eighteenth century case-law on the
annual hiring. Thus at the turn of the century, the courts passed over the
old authorities to find that a master had no obligation to maintain a ser-
vant or to provide him or her with medical care and expenses in the event
of sickness or injury.35 These cases suggest that the courts at this time had
no consistent conception of the contract of employment as a legal institu-
tion governing the reciprocal obligations of industrial workers and their
employers; the classification of work relationships was determined not so
much by contractual practices, as by the forms of regulatory legislation
which operated on the service ­relationship.
This juridical transformation of the service relationship needs to be set
in the context of the organisational and technological transformation of
production which was taking place during the late eighteenth and early
nineteenth centuries. Although guild-based and independent forms of
production were in decline at this time, a process which, as we have seen,
was hastened by legislative change and judicial interpretations, they did
not give way straight away to directly employed labour. Direct labour rela-
tions were by no means the inevitable outcome of technological change;
at this time, many of the new technologies, such as those associated with
steam power, were used in small workshops and in artisans’ cooperatives
as well as in factories.36 Moreover, certain of the pressures favouring the
adoption of subordinated labour were not technological in nature. Fac-
tory owners took steps to undermine forms of profit-sharing between

34 Sleeman v. Barrett (1863) 2 H & C 934, where the court explicitly rejected an earlier
suggestion (made in Bowers v. Lovekin (1856) 6 E & B 584) that the Truck Acts and the
Master and Servant Acts should be interpreted in tandem.
35 Newby v. Wiltshire (1784) 2 Esp 739; Wennall v. Adney (1802) 3 B & P 247.
36 See generally Berg, Age of Manufactures.
 the duty to work 41

independent producers: ‘some capitalist entrepreneurs had had to fight

for a system in which they could keep all the decision-making, all the
control, and all the subsistence of the workers in their own hands, and
had had to destroy earlier and more equable systems, such as the laws
and organisations of the free miners in the Stannaries, the Mendips, Der-
byshire, the Forest of Dean, Alston Moor and similar areas of Scotland’.37
Close supervision and control of the work process was a feature, by con-
trast, of conditions in the growing number of pauper factories and work-
houses in which those dependent on poor relief were set to work. Pauper
apprentices were employed on a similar basis in private industry, often as
a condition of their parents continuing to receive relief from the poor law
authorities.38 Thus it was not surprising that independent artisans and
outworkers associated factory labour with the workhouse and, as far as
they could, resisted it, drawing on a long cultural tradition of hostility to
the loss of autonomous status.39 In sectors where employers did establish
direct employment relations, they sometimes took a form not far removed
the ‘unfree’ labour of the workhouse. This was exemplified by the institu-
tion of the pit bond in the north-eastern coal fields, by which colliers were
bound by long-term contracts incorporating complex disciplinary codes.
Pit-bonds were regularly enforced in the courts as a means of breaking
strikes and instilling work discipline. This was a nineteenth century prac-
tice but it had roots going back to the early modern and late medieval
period, when long-term employment contracts akin to a form of indus-
trial serfdom had been widespread in the mining industry of Scotland and
parts of the north east of England.
The traditional pit bond was, however, exceptional, and was in any
case in decline by the 1830s.40 In most industries, when the adoption of
integrated factory production took place, it did not lead immediately to
the establishment of direct employment relations between owners and
­workers. Factory sites often began as mills to which outworkers and arti-
sans brought goods for finishing up; in practice, there was often an unclear
boundary between merchants putting materials out for production,
and employers of direct labour. When the transition from ­independent

37 Pollard, Genesis of Modern Management, 39.

38 Ibidem, 161.
39 Hill, “Pottage”.
40 See Steinfeld, Coercion, Contract, 167–182, on the evolution of contracts in coal min-
ing away from annual bonds and monthly hirings to short-term ‘minute contracts’ between
the 1830s and 1860s.
42 simon deakin

­ roduction occurred it was mediated in most industries by the institution

p
of internal contracting, through which artisans or supervisors contracted
with the works owner or entrepreneur for the performance of a job in
return for piece rates, and were themselves responsible for the employ-
ment of the labourers and underhands on time rates.41
Against this background, the significance of the master and servant
legislation lay in providing employers with a mechanism for imposing
discipline on workers who otherwise had only a loose organisational con-
nection to the firm, and who would often be in a position to take advan-
tage of labour shortages to push up wages. Most of the more contentious
mid-nineteenth century decisions on the coverage of the Master and Ser-
vant Acts were precisely concerned with the position of skilled artisans
and intermediaries such as butty workers; the central issue concerning
the scope of the Acts was how far they could be used by employers to
discipline skilled workers who otherwise retained a semi-independent
status. This was reflected in the practice of enforcement of the Act. Pros-
ecutions played an important role in relation to the business cycle, by
providing employers with a disciplinary weapon to counter the effects of a
tight labour market; prosecutions for absconding or refusing to work rose
at times of high levels of activity, as workers sought better paid or less
dangerous work. In terms of numbers of prosecutions alone, the mate-
rial impact of the Acts was considerable, in particular in the mining and
engineering trades and in certain regions, such as the Potteries.42 In Eng-
land and Wales, numbers of prosecutions never fell below 7,000 per year
between 1858 and the repeal of the Acts in 1875; a peak of over 17,000
prosecutions was reached in 1872. The historical evidence suggests that
the disciplinary mechanism of the Acts were widely used as instruments
of economic regulation during a period when modern managerial tech-
niques had yet to develop, and when shifts in the business cycle could
lead to considerable fluctuations in the bargaining power of employers
and workers.
In sum, the master-servant model was not a hang-over from the corpo-
rative regime of the Statute of Artificers; on the contrary, the core disci-
plinary powers of employers and of the courts were enacted in legislation
passed in the century from around 1750 as Parliament responded to the

41 Daunton, Progress and Poverty, ch. 7, for an account of the ‘coming of the factory’.
42 Simon, “Master and servant”; Woods, “Master and Servant Acts in the Black Country”;
Hay, “Master and servant”; Steinfeld, Coercion, Contract, ch. 2.
 the duty to work 43

interests of the new employer class. The model of subordination embed-

ded in the Master and Servant Acts was transferred over to the common
law, with the result that long after the repeal of the Acts in 1875, the lan-
guage of ‘master’ and ‘servant’ remained, a feature which was to weigh
heavily on the future of labour law in Britain and the wider Empire.43

A Comparative Perspective: The American Case

During the nineteenth century the American systems followed their own
route, which also eventually led to the emergence of a general model of
the employment contract covering all work relations, but differing in sig-
nificant respects from that which developed in England. By the start of the
twentieth century, virtually all US states had adoped the rule of employ-
ment at will, according to which either party could terminate the con-
tract without giving a reason for doing so, on minimal notice. Thus the US
courts did not develop a common law action of wrongful dismissal similar
to that which emerged in England. An important turning point was the
1870s. Before that, courts on both sides of the Atlantic had inserted pro-
vision for notice periods into the contract of employment, based on the
type of payment (by the hour, day, week, month and so on). Employees
with longer notice periods enjoyed a kind of legal job security; those with
minimal notice rights were in effect employed at will.
However, the US courts gradually came to regard the at-will form of
employment as the default, which could only by evidence of clear inten-
tion of the contracting parties. This rule originated in Payne v. Western &
Atlantic Railroad 44 and began to take shape in the context of claims aris-
ing from termination of employment such as Martin v. New York Life Insur-
ance Co.,45 a case concerning a salaried worker who in England at the same
time would have benefited form the presumption in favour of extended
notice periods. A further reason for the embedding of at-will employment
was the debate, within constitutional law, concerning the legitimacy of
legislation concerning workers rights to form trade unions. The legal right
of the employer to dismiss at will was a crucial weapon in resistance to
unionisation. The courts struck down legislation protecting union rights,
for example by outlawing the ‘yellow dog’ contract under which workers

43 Hay and Craven (eds.) Masters, Servants and Magistrates.

44 81 Tenn. 507 (1884).
45 148 NY 117 (1895).
44 simon deakin

agreed not to become union members as a condition of their employ-

ment, by invoking at-will employment.
In Britain, by contrast, only a segment of the wide range of work
contracts was presumed, as a matter of construction, to be at will. The
constitutional context which reinforced the movemtn towards at will
employment of US law was missing. The Trade Disputes Act of 1906 con-
firmed the support of the state for collective bargaining and the freedom
to take strike action in support of union membership and recognition,
and the courts gradually adjusted the content of employment law to this
new context.

The Emergence of the Contract of Employment in British Labour Law:

The Role of Social Security Legislation

In common with other jurisdictions, Britain repealed its master and ser-
vant legislation in the second half of the nineteenth century. However, the
legacy of the master and servant code was the assimilation by the com-
mon law of a hierarchical, disciplinary model of service. The persistence
of this model owed much to the Employers and Workmen Act 1875, which
provided statutory backing for a wide view of managerial prerogative and
also for the application of quasi-criminal sanctions against certain catego-
ries of workers well into the twentieth century.
The removal of criminal sanctions from the individual employment
relationship in the 1870s was soon followed by the first legislative inter-
ventions of the welfare state. This wave of ‘social legislation’ began with
the Employers’ Liability Act 1880, which mitigated the effects of the doc-
trine of common employment. This rule held that an employer could not
be held vicariously liable in tort where negligence by one employee, act-
ing in the course of employment, caused personal injury to another. The
basis for the rule was the fiction that the contract of service contained an
implied term under which the employee consented to the risk of being
injured by a fellow worker. The 1880 Act made a limited exception, by
excluding the common employment rule in a case where the plaintiff
was injured as a result of the negligence of a fellow employee exercising
managerial or supervisory responsibilities. From this small beginning, the
principle that the employer should assume responsibility for social and
economic risks arising from the employment relationship began to take
shape. The first Workmen’s Compensation Act was introduced in 1897
and the first National Insurance Act in 1911. The workmen’s compensation
 the duty to work 45

scheme imposed liability on employer for workplace-related injuries and

disease, and prompted the widespread use of employers’ liability insur-
ance to spread the risks in question. Social insurance spread the more
general risks of interruption to earnings through illness, unemployment
and old age throughout the working population through the device of
state-run insurance funds. National insurance contributions, levied on
employers and employees, introduced a kind of fiscal regulation of the
employment relationship which was later extended through the income
taxation scheme.
These interventions, by their purpose and nature, presupposed a cer-
tain degree of stability and regularity in the employment relationship, so
reinforcing tendencies towards the growth of firms. Once the enterprise
became a mechanism for the redistribution of social and economic risks,
the resulting imposition of regulatory and fiscal costs on to employers
created economies of scale which favoured larger firms at the expense of
smaller productive units. The welfare state extended its influence in lock-
step with increasing vertical integration of production and the emergence
of the public sector as a significant employer.
This process can be observed at the juridical level in terms of the deci-
sions of courts called on to determine the scope of social legislation. Early
social legislation did not adopt the ‘unitary’ model of the contract of
employment, which emerged only after 1945; instead, it persisted in strati-
fying the labour force by reference to distinctions drawn from restrictive
legislation and from the common law. Hence salaried, non-manual work-
ers were distinguished from hourly or daily-paid wage earners, who were
distinguished in turn from casual workers and others with irregular wage
relationships. At the core of this process of stratification was the ‘control’
test, the predominant judicial test of the time for identifying the service
relationship.
The Employers’ Liability Act of 1880, which made a limited exception
to the doctrine of common employment, and the Workmen’s Compensa-
tion Act of 1897 both adopted the term ‘workman’ rather than ‘employee’
to describe the workers within their scope. As we have seen, the term
‘workman’ had a specific meaning, related to manual labour, under the
Employers and Workmen Act 1875; to this extent it was a legacy (if in
somewhat attenuated form) of the status distinctions drawn under the
master and servant regime. The 1880 Act explicitly adopted the mean-
ing of the term under the Act of 1875, and the two Acts—the one pro-
tective, the other restrictive—were subsequently interpreted in tandem.
46 simon deakin

In a ­further reinforcement of status distinctions, supervisory employees

were also expressly excluded from the protection of the 1880 Act. The
term ‘workman’ was also applied to truck legislation following the Truck
Amendment Act of 1887.
Following the passage of these Acts, the term ‘workman’, previously
given a wide reading, began to receive a more restrictive interpretation
in the courts. In Morgan v. London General Omnibus Company the Court
of Appeal, in a claim for personal injuries compensation under the 1880
Act, held that a bus conductor ‘who earns the wages becoming due to
him through the confidence reposed in his honesty’, was not a workman:
‘labourer’ cannot in its ordinary acceptation include an omnibus conduc-
tor . . . The mere fact that a man works with his hands is not enough to con-
stitute a workman within [the Act].46
Similarly, a locomotive guard was not within the Truck Acts,47 a sales
assistant48 and a hairdresser49 were outside the Employers and Workmen
Act, and a tramcar driver was outside the Employers’ Liability Act.50 As
a result of these decisions, a large class of employees, who were neither
supervisory employees within the meaning of the Employers’ Liability Act
nor managerial employees under the Truck Acts, were excluded from the
protective scope of those statutes.
The first Workman’s Compensation Act of 1897 adopted a wider defini-
tion, referring to a workman as ‘[including] any person who is engaged in
an employment to which this Act applies, whether by way of manual labour
or otherwise’. The employments covered by this Act were restricted to the
railways, mining and quarrying, and factory and laundry work.51 However,
the attempt to give the Act an extended scope failed to survive the Court
of Appeal judgment in Simpson v. Ebbw Vale Steel, Iron & Coal Co.52 Here,
a widow of a colliery manager sued his former employers for compensa-
tion in respect of his death in an underground accident. Her claim was

46 See (1884) 13 QBD 832, 834 (CA), and (1883) LR 12 QBD 201, 207 (DC).
47 Hunt v. Great Northern Railway Co. [1891] QB 601.
48 Bound v. Lawrence [1892] 1 QB 226.
49 R. v. Louth Justices [1900] 2 Ir. R. 714.
50 Cook v. North Metropolitan Tramways Ltd. (1887) 18 QBD 683, not followed in Scot-
land: Wilson v. Glasgow Tramways & Omnibus Co. (1878) 5 SC (4th. Ser.) 981; while in Smith
v. Associated Omnibus Co. [1907] 1 KB 916 the court distinguished Cook by deciding that a
bus driver who had to start his engine with a hand pump and used spanners and wrenches
to fix it if it broke down was, for these reasons, a ‘workman’.
51 WCA 1897, s. 7.
52 [1905] 1 KB 453.
 the duty to work 47

rejected, the court concluding that while it was theoretically possible for a
non-manual worker to be within the Act, he ‘must still be a workman’. The
Simpson decision was based on the assumption that protective legislation
of this kind was inappropriate for a high status (salaried) worker:
It presupposes a position of dependence; it treats the class of workmen as
being in a sense ‘inopes consilii’, and the Legislature does for them what
they cannot do for themselves: it gives them a sort of State insurance, it
being assumed that they are either not sufficiently intelligent or not suf-
ficiently in funds to insure themselves. In no sense can such a principle
extend to those who are earning good salaries.53
Despite the hostile language used here, it should not be thought that it
was the courts alone which perpetuated these status distinctions. Such
distinctions were deeply embedded in the law as a result of the influ-
ence of the Employers and Workmen Act, and the levels of compensa-
tion established by the Workmen’s Compensation Act were low enough
to justify a belief that the Act was principally intended for the manual
industrial workforce. Lawyers could be explicit about the class distinc-
tions underlying the Acts; in Simpson counsel argued that ‘the Legislature
were contemplating a class of workers who may be described as belonging
to the working class in the popular sense of the term—a wage earning
class’.54
The Workmen’s Compensation Act of 1906 established yet another
scheme of definition, which in due course formed the basis for the clas-
sifications drawn by the legislation of the pre-war and inter-war periods.
A ‘workman’ was now defined as ‘any person who enters into or works
under a contract of service or apprenticeship with an employer, whether
by way of manual labour, clerical work or otherwise, and whether the
contract is expressed or implied, is oral or in writing’; but there were sig-
nificant exceptions, including non-manual workers employed on annual
remuneration greater than £250, casual workers employed ‘otherwise
than for the purposes of their employer’s trade or business’, outworkers
and family workers. This scheme, then, excluded high status workers at
the top and casual workers at the bottom.
In similar vein the National Insurance Act of 1911 applied its health
insurance provisions to ‘employed’ persons, who were defined principally
as those employed under a contract of service or apprenticeship, but

53 [1905] 1 KB 453, 458 (Collins MR).

54 [1905] 1 KB 453, 456.
48 simon deakin

e­ xcluding non-manual workers on an annual salary of more than £160,

casual workers not dependent on their employer’s business, commission
agents, and a large class of public sector workers including civil servants,
military personnel and teachers, on the grounds that they were better
served by state schemes already in existence.55 Male outworkers were
included in the Act,56 but most female outworkers were initially excluded
by the provision excepting outwork employment ‘where the person so
employed is the wife of an insured person and is not wholly or mainly
dependent for her livelihood on her earnings in such employment’.57 The
Insurance Commissioners later exercised delegated powers58 to bring
female outworkers within the Act,59 but both male and female outworkers
remained outside the unemployment insurance part of the scheme. Male
industrial workers were also the principal beneficiaries of the unemploy-
ment insurance scheme. The 1911 Act applied only to a specified number
of industrial trades;60 women workers were only covered in practice after
the extension of insurance in the Unemployment Insurance Act of 1921,
and even then the legislation set differential contribution and benefit
rates which had the effect of providing women workers with less exten-
sive protection than men.
The hierarchical distinction between clerical/managerial workers,
industrial workers and casual workers which is to be found in this legis-
lation filtered through into the common law in the form of the ‘control’
test for identifying the contract of service. ‘Control’ is traditionally seen
in labour law as a test inherited from pre-industrial traditions of employ-
ment, and in particular the forms of domestic and artisanal production.
This was the sense referred to by Kahn-Freund when he described the
control test as:
based upon the social conditions of an earlier age . . . It reflects a state of
society in which the ownership of the means of production coincided with

55 See NIA 1911, s. 1 and Schedule 1.

56 NIA 1911, Sch. 1, Part I, para. (d). The courts held that master tailors were within
the Act even if they themselves hired others: re Master Tailors as Outworkers (1913) 29
TLR 725; and outworkers were held to come under the Trade Boards Act 1909 (Street v.
Williams [1914] 3 KB 537). Statutory definitions of the term ‘homeworker’ subsequently
acknowledged that outworkers employing others could come under protective legislation.
See Wages Act 1986, s. 26(1); National Minimum Wage Act 1998, s. 35.
57 NIA 1911, Sch. 1 , Part II, para. (j).
58 Contained in s. 1(2) of the 1911 Act.
59 SR&O 1912/921; SR&O 1914/880.
60 See NIA 1911, Sch. 6. The regulated industries were named as: building, construction,
shipbuilding, mechanical engineering, ironfounding, vehicle construction, and sawmilling.
 the duty to work 49

the possession of technical knowledge and skill and in which that knowl-
edge and skill were largely acquired by being handed down from genera-
tion to the next by oral tradition and not by being systematically imparted
in institutions of learning from universities down to technical schools. The
control test postulates a combination of managerial and technical functions
in the person of the employer.61
Certainly, the control test was close to the test of ‘exclusive service’ which
was the principal test of subordinated employment under the Master and
Servant Acts; however, there is little to suggest that it was a remnant of
pre-capitalist forms of employment. ‘Control’ was not an important test
until the later nineteenth and early twentieth centuries, and when it
began to be used it was in the context of the social legislation of that
period and in particular the status distinctions which were being drawn
at that time.
An early case referring to ‘control’ in the context of the master’s vicari-
ous liability for the torts of the servant is Sadler v. Henlock (1855),62 in
which Crompton J. defined the relevant test as ‘whether the defendant
retained the power of controlling the work. No distinction can be drawn
from the circumstance of the man being paid at so much a day or by the
job.’ In contrast to a servant, an independent contractor ‘chooses the mode
in which the work is done, and the persons who do it’. This was a case
concerning a labourer employed by the defendant to clear a drain on his
land. It is unlikely that the control test, as such, was clearly established as
a general test of status at this time; up to the repeal of the Master and Ser-
vant Acts the judicial focus was on ‘exclusive service’ and after that on the
definition of the term ‘workman’. Although ‘control’ may have been more
important in the area of tortious liability for personal injury, the main
issues there concerned the scope of the defences—common employment,
consent (volenti non fit injuria) and contributory negligence—rather than
the question of workers’ status.
The control test was only clearly asserted later, in cases concerning
not the common law of vicarious liability but the scope of social legisla-
tion. The leading cases in which the test was established were decided in
the twentieth, not the nineteenth century—Simmons v. Heath Laundry63
and Underwood v. Perry64 in the field of workmen’s compensation, and

61 Kahn-Freund, “Servants and independent contractors”, 505.

62 4 E. & B. 570.
63 [1910] 1 KB 543.
64 [1923] WC & I Rep. 63.
50 simon deakin

­Scottish Insurance Commissioners v. Edinburgh Royal Infirmary65 and Hill

v. Beckett 66 in national insurance. The nineteenth century cases which
were most frequently cited in support of the test had nothing to do either
with vicarious liability or with contractual disputes between employer and
employee. The 1873 case of R. v. Negus,67 in which Blackburn J. referred to
control as the principal test, was concerned with the definition of ‘servant’
under the Larceny Act 1868. The subsequently much-cited case of Yewens
v. Noakes68 involved a claim for exemption from inhabited house duty
under the Customs and Inland Revenue Act 1869, under which premises
used for trade only would be exempt from the tax if it was occupied merely
by a caretaker, defined by the Act as ‘a servant or other person . . . for the
protection thereof ’. The Court of Appeal held that a clerk, earning £150
per annum, did not fall into this category.
In both these cases, the judges were clear that the status of ‘servant’
was distinct from that of high status ‘employee’, and that the main basis
of differentiation was not the knowledge of the trade or craft but the pres-
ence of a wide-ranging power of the master to give orders to the servant,
something the courts had previously associated with the idea of exclu-
sive service. Blackburn J.’s judgment in Negus reflected this, when he said
that ‘the test is very much this, whether the person is charged and bound
to obey the orders of his master. He may be so without being bound to
devote the whole of this time to this service; but if bound to devote his
whole time to it, that may be very strong evidence of his being under
control’.69 Lord Justice Bramwell evidently had the same distinction in
mind when he said in Yewens v. Noakes that ‘a servant is a person subject
to the command of his master as to the manner in which he shall do his
work’.70 Lord Justice Thesiger thought it obvious that a salaried clerk was
not a servant, any more than would be ‘the manager of a bank, a foreman
with high wages, persons in the position almost of gentlemen’.71
Why did courts, several decades later, light upon these cases as author-
itative guidance to the classification of employment relationships? The
answer would not be problematic if the control test had become well

65 1913 SC 751.
66 [1915] 1 KB 578.
67 (1873) LR 2 CP 34.
68 (1880) 6 QBD 530; see Merritt, “Control v Economic Reality”, 113: ‘the “control” test
has been misapplied almost from the moment of its emergence’.
69 (1873) LR 2 CP 34.
70 (1880) 6 QBD 530, 532–533.
71 (1880) 6 QBD 530, 538.
 the duty to work 51

established in tort law or in the interpretation of labour statutes, but as we

have seen, this does not appear to have been the case. A more convincing
explanation is that rediscovery and adaptation of the control test was a
doctrinal innovation which was introduced at the same time as the courts
were being called on to define the boundaries of regulatory legislation of
a novel type. The element of compulsion in the legislation went strongly
against the grain of prevailing common law values, as the judgments of
the Court of Appeal in Simpson v. Ebbw Vale make clear. It was in the
context of employers’ efforts to limit the scope of the legislation and a
sympathetic judiciary that the control test was taken up.
The particular effect of the control test was that it reinforced status
distinctions between the ‘labouring’ and ‘professional’ classes, on the one
hand, while excluding casual and seasonal workers, to whom the employer
made a limited commitment, on the other. Simmons v. Heath Laundry72
concerned the employment status of a young laundry woman who gave
piano lessons in her spare time. It was held that in respect of the music
lessons she was not a ‘workman’ [sic], so that when she injured her hand
at the laundry she could not claim for diminution of earning capacity as
a pianist. Rather than saying that she was an independent contractor for
the purposes of the lessons, the Court of Appeal used the control test in
order to place her in the category of professional workers: ‘the question
to be asked is what was the man [sic] employed to do; was he employed
upon the terms that he should within the scope of his employment obey
his master’s orders, or was he employed to exercise his skill and achieve
an indicated result in such manner as in his judgment was most likely to
achieve success?’.73 In other cases in this line, a lecturer, chemist, nurse,
doctor and poor law officers were held to be outside the Act.74 This exclu-
sion reflected the view that compulsory insurance was inappropriate for
salaried or professional employees; it was also a means of reducing the
potential liability of poor law guardians and hospitals, at a time when
funding was more precarious than it subsequently became.75

72 [1910] 1 KB 543.
73 [1910] 1 KB 543, 553 (Buckley LJ).
74 See, respectively, Waites v. Franco-British Exhibition (1909) 25 TLR 441; Bagnall v.
Levinstein [1907] 1 KB 531; Dow v. McNeil [1925] WC & I Rep. 32 (although not followed
Wardell v. Kent County Council [1938] 2 KB 769; re South Dublin Union Officers [1913] WC&I
Rep. 245.
75 In similar vein, the control test could even be used to hold that a nurse was an
employee or ‘servant’ of a hospital for one purpose (administrative duties) but not another
(work on the ward): Hillyer v. St. Bartholomew’s Hospital [1902] 2 KB 820, a tort case, not
52 simon deakin

The position of casual workers was complicated by the prevalence of

sub-contracting in many industries. The contract system of hiring labour
through an intermediary was still the predominant form of industrial
organisation in road building, construction, shipbuilding, mining and
quarrying, and iron and steel. As we have seen, this meant that there was
unlikely to be a contractual nexus between workmen hired by the butty
worker or foreman, and the ultimate owners of the site, plant or mate-
rials on which they worked. It was for this reason that the Factory and
Workshop Acts of the period imposed obligations not upon employers
in respect of their employees, but upon occupiers or owners of factories
in respect of workers on their premises, regardless of whether there was a
contractual relationship between them.76
The control test, as applied by twentieth century courts, often had
the effect of classifying foremen as independent contractors, given their
responsibility for hiring their own gangs;77 while the gang workers or
labourers themselves had no claim against the ultimate users of their
labour since the latter did not ‘control’ the performance of their work.78
Piecework payments were also treated by the courts as strong evidence
of independent contractor status, notwithstanding clear statutory signals
that this was not to be the case.79 Share fishermen—inshore trawlermen
who were paid on a proportion of the profits from individual voyages—
were found to be outside the social insurance legislation.80
Seasonal and casual workers who were directly employed by the ulti-
mate employer were the subject of particular statutory provisions, speci-
fying that they were to be included in the legislation where they were eco-
nomically dependent on their employer’s business. This did not prevent
considerable litigation arising. The courts held that a single, one-off hiring

followed in the workmen’s compensation case of Wardell v. Kent County Council [1938]
2 KB 769, and overruled for tort law purposes (but with Greer LJ dissenting) in Gold v.
Essex County Council [1942] 2 KB 293.
76 See Factory Act 1844, ss. 41, 73; Factory and Workshop Act 1878, ss. 93, 94.
77 The courts reached opposing outcomes in different cases. Butty workers and fore-
men were found to be within the protection of the Workmen’s Compensation Acts in
Evans v. Penwelt Dinas Silica Brick Co. (1901) 18 TLR 58 and Paterson v. Lockhart (1905) 42
SLR 755, but outside them in Simmons v. Faulds (1901) 17 TLR 352, Hayden v. Dick (1902)
40 SLR 95 and Vanplew v. Parkgate Iron & Steel Co. [1903] 1 KB 851.
78 Crowley v. Limerick County Council [1923] 2 Ir. R. 178; Littlejohn v. Brown & Co. Ltd.
1909 SC 87; although cf. M’Ready v. Dunlop (1900) 37 SLR 779; Doharty v. Boyd 1909 SC 87.
79 WCA 1906, s. 13; NIA 1911, Sch. 1, Part I, para. (a).
80 Scottish Insurance Commissioners v. M’Naughton 1914 SC 826.
 the duty to work 53

was not necessarily outside the scope of the Acts,81 and were also prepared
to include seasonal workers who returned to their employer on a regu-
lar basis82 and part-time workers with long service.83 However, in other
cases they excluded workers under task contracts,84 temporary workers85
and casual workers with short-term service.86 Trainees and unemployed
workers receiving instruction and work experience at government train-
ing centres were also beyond the scope of the legislation.87
The adoption of the control test enabled employers to avoid responsi-
bility for the social risks of illness, injury and unemployment which it had
been the aim of social legislation to impose, at least in part, upon them.
The drawing of fine distinctions to defeat the legislation in this way was
not universally accepted. The Scottish courts, influenced, perhaps, by the
civilian tradition of giving a purposive interpretation to social legislation,
were on occasion reluctant to allow employers to contract out of the Acts
by these means (just as, in the nineteenth century, they had been reluc-
tant to accept the doctrine of common employment which held that the
worker impliedly accepted the risk of a fellow servant’s negligence).88 In
Paterson v. Lockhart,89 a case concerning the status of a foreman, Lord
McLaren considered that
it would be a serious restriction of the scope of the Act if it were possible by
introducing some condition into an agreement to take it out of the category
of a pure contract of service and so to avoid liability under the Act . . . It
is quite in accordance with custom for a superior workman to choose his
own assistants. An engineer may choose his own fireman or a mason his
hodman. But that does not prevent their being servants paid by a common
employer.

81 Boothby v. Patrick & Son [1918] W.C. & I. Rep. 340.

82 Smith v. Buxton [1915] W.C. & I. Rep. 126.
83 Dewhurst v. Mather [1908] 2 KB 754.
84 Alderman v. Warren [1916] WC & I Rep. 266.
85 Stoker v. Wartham [1919] 1 KB 499.
86 Knight v. Bucknill [1913] WC & I Rep. 175; Withams v. Larsen Ltd. [1928] WC & I
Rep. 323.
87 Broome v. Ministry of Labour [1927] W.C. & I. Rep. 232; Watson v. Government Instruc-
tional Centre [1929] W.C. & I. Rep. 265; re Leeds Corp. and Chadwick (1928) 44 TLR 797
(cf. Unemployment Insurance Act 1927, s. 15; Unemployment Insurance Act 1927, s. 15);
McGeachy v. Dept. of Health for Scotland 1938 SC 282.
88 It was not until the House of Lords decided Bartonshill Coal Co. v. Reid (1856) 4 Macq.
266 that the Scottish courts accepted the doctrine, which had its origins in decisions of
the English courts.
89 42 SLR 755, 757.
54 simon deakin

In Dunlop v. M’Cready,90 Lord Adam said of a group of gang workers

that ‘to call them independent contractors is a mere playing with words’.
These were, however, isolated examples of a more tolerant judicial atti-
tude. More typical of the case-law was the finding of the Court of Session
in Littlejohn v. Brown91 to the effect that a rivet boy, hired by the head
riveters in a shipyard, had no contract of service with the main employer:
‘if the test is direct and immediate selection, payment and control and
power of dismissal, the evidence is all one way, namely, that [the plaintiff ]
was in the service of Gammell and Lacey, John Brown & Company being
only indirectly connected with the boy’s employment’.
The ‘unitary’ model of the contract of employment which, in modern
labour law, extends to all categories of wage-earners, only came into being
when further reforms were enacted to social legislation, in particular the
extension of social insurance which took place in the National Insurance
Act 1946. A major aspect of the Beveridge Report was the abolition of dis-
tinctions between different categories of employees: henceforth, all wage
or salary earners, regardless of their annual income or of their professional
status, would come under the same contributory classification.92 Accord-
ingly, the 1946 Act established two principal classes of contributors: Class
I covered ‘employed earners’, defined as ‘any persons gainfully occupied in
employment . . . being employment under a contract of service’, and Class
II covered those employed on their own account.93 The latter paid a lower
rate of contribution and were excluded from the unemployment insur-
ance part of the scheme. In this way the fundamental division between
employees and the self-employed was established. The same distinction
was adopted for the purposes of income taxation94 and, in due course,
under the employment protection legislation which was introduced first

90 (1900) 37 SLR 7797, 782.

91 1909 SC 169, 174.
92 Social Insurance and Allied Services Cmd. 6404, November 1942, at para. 314.
93 NIA 1946, s. 1(2).
94 The modern division between self-employment and employment in tax law emerged
gradually during the inter-war period. The principal division in the Income Tax Act 1918
was between earnings from public-sector employment, which fell under Schedule E, and
earnings and profits from all other employment, which fell under Schedule D. Schedule D
therefore covered both the self-employed and certain groups of private-sector employees.
The Finance Act 1922, s. 18, then transferred earnings from private-sector employment into
Schedule E, leaving only the self-employed in Schedule D. The Pay-As-You-Earn system of
automatic deductions of tax from the earnings of employees was applied to most Sched-
ule E employments by the Income Tax (Employments) Act 1943.
 the duty to work 55

in the early 1960s.95 The ending of the old divide between manual and
non-manual workers was epitomised by the merging of the concepts of
the contract of service and of employment: for statutory purposes, these
were now synonymous with each other.96
Faced with this new situation, the courts abandoned the old distinction
between low status and high status employees when seeking to identify
the contract of service.97 The control test itself came to be regarded as
excessively artificial, and gave way to the tests of ‘integration’ and ‘busi-
ness reality’. These stressed economic as opposed to personal subordina-
tion as the basis of the contract of employment. The test of the worker’s
‘integration’ into an organisation was used to explain how professionals
such as doctors and journalists could be classified as employees not-
withstanding the high degree of autonomy they enjoyed in their work.98
‘Economic reality’ had the effect of extending protection to casual work-
ers and outworkers who were dependent on the business of another, as
opposed to being entrepreneurs with a business and employees of their
own.99 By these means, a more inclusive notion of the employment rela-
tionship came to be established for the purposes of determining the scope
of employers’ liabilities in respect of personal injuries, employment pro-
tection and social insurance.100
Thus the rise of the welfare state helped to confirm the movement
towards the contractualisation of employment relations. Nevertheless, we
can see the preservation of elements of the master-servant model during
the period of transition from the poor law to social security, a process
which retarded the devleopment of British labour law. It is only in the
mid-1950s that the clear division between the ‘employee’ and the ‘self-
employed’ worker was recognised in labour law. There were some prec-
edents for this approach in tax law and social security law from the 1910s
onwards, but even here there was a tendency to sub-divide the class of

95 The first such statute was the Contracts of Employment Act 1963; the relevant provi-
sion is now contained in the Employment Rights Act 1996, s. 230(1).
96 See, in the context of social insurance, Vandyk v. Minister of Pensions and National
Insurance [1955] 1 QB 29.
97 See, in particular, Stevenson, Jordan & Harrison v. McDonald & Evans [1952] 1 TLR 101.
98 Cassidy v. Minister of Health [1951] 2 KB 343; Roe v. Minister of Health [1954] 2 QB 66;
Beloff v. Pressdram Ltd. [1973] 2 All ER 241.
99 Market Investigations Ltd. v. Minister for Social Security [1969] 2 QB 173; Lee Ting
Sang v. Chung Chi-Keung [1990] ICR 409.
100 See Mitchell and Howe, “Evolution of the contract of employment in Australia” for
discussion of how far a similar process of conceptual evolution to that described in the
text occurred in Australian labour law during the same period.
56 simon deakin

wage-dependent workers by reference to occupational status and income

level. It was only with the passage of the National Insurance Act 1946,
implementing key elements of Beveridge’s postwar scheme for social
insurance, that the modern legal division between employment and self-
employment became clear. This distinction was then carried over into the
first employment protection statutes of the 1960s. The term ‘contract of
employment’ which dates from this time truly is a recent development.

The Overseas Diffusion of the Master-Servant Model

The transmission of master-servant legislation to British colonies began

as early as the seventeenth century (in the case of certain American
states and the West Indies) and carried on into the twentieth century (in
the case of some African countries), well after the repeal of the original
laws in Britain. Over the course of the three hundred years or so, ‘almost
2000 statutes and ordinances made their appearance in more than 100
colonies, developing a colonial master and servant law that drew upon,
elaborated, and often subverted the metropolitan models’.101 The ending
of slavery in the Empire in the early nineteenth century and the formal
instantiation of freedom of contract was accompanied by the enactment
of pass and police laws and vagrancy legislation. Systems of plantation
labour in the West Indies and Assam, and mine labour in South Africa,
were underpinned by the coercive powers which master-servant law pro-
vided to ­employers. There was also a link to migration and racial segre-
gation. Legislation was passed in several colonies in the 1830s and 1840s
for the purpose of regulating the waves of indentured labour (millions of
workers) moving around the Empire at that time. These laws created new
forms of status around racial and cultural categories. Legislation of this
type was being enacted in east Asia, the Carribean and Africa into the
middle decades of the twentieth century.
Whereas in Britain, master-servant laws were used to stabilize the labour
supply, reduce the bargaining power of workers and shore up managerial
prerogative in the mainly small-scale manufacturing enterprises which
were characteristic of the industrial structure of the parent system, in the
colonies the same types of laws were used to assist in the dispossession
and separation from the land of indigenous populations and to maintain

101 Hay and Craven, “Introduction”, in Hay and Craven (eds.) Masters, Servants and
Magistrates.
 the duty to work 57

the supply of cheap labour which was essential in plantation and mining-
based economies. Plantation systems simply were not profitable without
criminal enforcement of labour contracts; and the prosecution rates in
some plantation societies were fully fifty times as high as those in parts of
England, such as the west Midlands, where the master-servant laws were
most heavily relied on by employers. In late nineteenth century Trinidad,
around one fifth of all criminal convictions involved breach of contract.
Enforcement took the form not simply of fines and imprisonment, but of
judicially-supervised performance of the work contract, and the addition
of extra periods of service as compensation for breaches by the worker.102
The end of master-servant laws did not come about through economic
development. These laws lingered longest in systems without the demo-
cratic suffrage and without recognition of basic labour rights. Britain’s own
experience had been similar: it was only as the franchise was extended
that the political conditions for the repeal of master-servant laws were
gradually established. Across the Empire, notwithstanding some liberalis-
ing moves from the Colonial Office which local employers often resisted,
it was the pressure of the International Labour Organisation in the 1920s
and 1930s and then the decolonisation process itself which brought about
the abandonment of penal laws.103
The influence of the master-servant model is a good example of a legal
origin effect which occurs when transplantation is combined with colo-
nialisation to produce strong path dependencies.104 This case is, however,
further repudiation of the claim that common law systems were more
inherently disposed to a liberal contractual model of work relations than
civil law ones. Master-servant law was essentially status-based; that is
to say, it preserved distinctions based on class, in the British case, and
race, in the colonies. This can be seen from the evolution of the law in
South Africa. The master-servant laws there were facially neutral, but
with the passage from the middle decades of the twentieth century
of industrial relations legislation regulating the mainly (and later exclu-
sively) white occupations, the master-servant regime was applied the
agricultural and domestic sectors which were traditionally non-white.105

102 Turner, “British Caribbean”.

103 Banton, “Colonial Office”.
104 Deakin, “Legal origin”.
105 Le Roux, “Evolution of the Contract of Employment in South Africa”.
58 simon deakin

There were prosecutions of tens of thousands of workers each year into

the 1950s.106

The Civil Law Systems: France and Germany

The law followed a different path in the civil law systems. The codes of
the post-revolutionary era made use of the traditional concept, derived
from Roman law, of the locatio conductio for the purpose of defining the
new forms of waged work. In adopting the idea of the locatio, the drafters
of the civil codes were aligning work relations with other types of con-
tract, an indication that they regarded the employment relationship as
ultimately founded on an exchange. Thus labour, or labour power (Arbe-
itskraft), was treated as a good or product to which the contract attached
a ‘price’ (not necessarily a ‘wage’). It was classified as part of the law of
things rather than the law of persons: the codes made no mention of the
‘subordination’ of the worker.107 The reality of work relations was very dif-
ferent : all the civil law systems recognised the power of the employer to
give orders, to fix the rules of the workplace, and to to retain the worker
in his service. The laws underpinning managerial prerogative were now,
however, regarded as part of private law; they were seen as part of specia-
lised regulatory or police powers, a feature which helps to explain their
near invisibility at the level of case law and doctrine.
The French Civil Code of 1804 used two versions of the locatio
­conductio.108 The louage d’ouvrage, loosely based on the locatio conductio
operis of Roman law, applied to piece work or to the engagement of a
worker to perform a specific task. The louage de services was modeled on
the locatio conductio operarum and as such was based on the principle of
the hire of services. In truth, any link to the original content of Roman law
was tenuous; the concepts used in the Code were genuine adaptations.109
The ideological aim in the use of the locatio was to avoid any hint of the
subjection of the worker; rather, the work relation was to be based on the
idea of the formal equality of contracting parties. Partly for this reason,
the Code did not admit the idea of a permanent or indefinite hiring.

106 Chanock, “South Africa”.

107 Simitis, “Case of the Employment Relationship”.
108 Code civil, art. 1780.
109 Veneziani, “Evolution of the Employment Relationship”.
 the duty to work 59

It is possible to see in the distinction between the louage d’ouvrage and

the louage de services the same fundamental division as that in English
law between servants and independent contractors. However, the binary
divide may not have operated as straighforwardly as that. The concept of
the louage de services tended to be applied to day workers and domestic
workers, with the louage d’ouvrage covering other categories. In the case
of the louage de services, the employer was the sole judge of whether the
contracted for work had been performed.110 The division of risks between
the parties was not greatly different in either case, and it is likely that in
practice they both were closer to the practice of waged labour than to that
of commercial relationships.111
It is therefore inappropriate to see the contract of employment as a
direct descendant of either the louage de services or the louage d’ouvrage.
Such a direct line of descent was maintained by nineteenth century jurists
who thereby hoped to legitimise the newly emerging notion of the con-
trat de travail (the term began to be used in the 1880s).112 This process
has more recently been termed a ‘mystification’.113 The contrat de travail
served the purpose, from the employer’s point of view, of providing a basis
for the worker’s duty of obedience. The concept was also used around this
time in legislation governing liability for workplace accidents.114 Commis-
sions of jurists charged with the task of establishing a framework for the
law relating to collective bargaining and health and safety at work helped
to spread its use.115 All these developments saw the idea of ‘subordination’
as lying at the core of the new concept. This implied an open-ended duty
of obedience. In exchange, the employer accepted certain responsibility
for social risks.
The German Civil Code adopted in 1896, the BGB, used an apparently
similar terminology to that of the French code at the start of the cen-
tury. The Dienstvertrag (literally, ‘contract for service’) applied to wage
labour while the Werkvertrag, contract for work, to subcontracting. In
other respects, however, the BGB broke with the model of the locatio. The
two contract forms were placed in separate parts of the Code, indicat-
ing not just the fragmentation of the locatio, but, more fundamentally,

110 Code civil, art. 1781.

111 Petit and Sauze, “Une lecture historique de la relation salariale”; Cottereau, “Droit
et bon droit.”
112 Ibid.
113 Cottereau, “Industrial tribunals”, 20.
114 Veneziani, “Evolution of the Employment Relationship”, 64.
115 Ibid., 68.
60 simon deakin

separate conceptions of the work relation. Within the structure of the

Dienstvertrag, the employer’s duty of care (Fürsorgepflicht) was balanced
by the worker’s duty of loyalty (Treuepflicht). Gierke had argued that the
Code should make reference to principles of solidarity which he claimed
to have found in pre-modern antecedents of the Dienstvertrag. By these
means, the employment relationship was aligned with the law of persons,
and with an emergent communitarian notion of the enterprise. Even so,
the BGB did not clearly recognise the ‘binary divide’ between subordi-
nated and independent work.116 The Dienstvertrag covered employment
contracts, but also some forms of the supply of services.117 This came later,
with the appearance in the inter-war years of the concept of the Arbeits-
verhältnis, a development associated with the rise of collective bargaining
and the welfare state.
In both countries, the form of the contract of employment was a reflec-
tion of prevailing juridical cultures.118 The French system was modeled
on the idea of the power of the state to regulate basic conditions of work
(ordre public social). The implicit logic in this approach was that the state,
in acknowledging the formal contractual equality of the parties, came
under a duty to provide protectoin to the weaker party, the employee,
who was thereby placed in a position of ‘juridical subordination’. By con-
trast, the German approach was based on the communitarian model of
the firm, reflecting the influence of Gierke who had inveighed against the
‘individualistic’ and ‘capitalistic’ orientations of the French Civil Code.
In German labour law, the subordination of the worker took the form
of their adhesion to the enterprise (Tatbestand), a process conferring
a status akin to membership of a community.119 These two, alternative
conceptions of the work relationship greatly influenced the subsequent
development of the law governing the employment relationship in con-
tinental Europe, giving rise to a ‘structural ambivalence’ at the core of
European labour law.120

116 Ibid., 59; Sims, Good Faith, 85–86.

117 BGB, para. 611.
118 Mückenberger and Supiot “Ordre public”.
119 Supiot, Critique du droit du travail, 18.
120 Ibid., 19.
 the duty to work 61

Conclusion

Without the welfare state and collective bargaining, there would be no

contract of employment in the sense that we recognise today. The history
of labour law teaches us that this concept serves two distinct functions:
on the one side, organising, from a juridical point of view, the coordina-
tion of the enterprise and, on the other, partitioning and redistributing
social risks arising from labour market participation. Solidararistic collec-
tive bargaining and social security systems are not, perhaps, inevitable
features of a market economy, but they are compatible with it. While
these models arguably reached their high point in western Europe and
north America in the middle decades of the twentieth century, it would
wrong to see them as specific to these systems or to a particular point in
the development of industrial capitalism. Labour law continues to evolve
alongside and in response to shifts in modes of production and societal
structure.
 Dutch Imperial Anxieties About Free Labour,
Penal Sanctions And The Right To Strike

Ulbe Bosma1

Introduction

Thanks to massive inputs of forced labour the Netherlands-Indies emerged

as a major plantation colony in the nineteenth and twentieth century.
Crucial in this process was the Cultivation System (1830–1870) introduced
by Count Johannes van den Bosch and designed as a policy of compulsory
cash crop growing imposed on the rural population of Java. The Cultiva-
tion System encroached upon the freedom of the Javanese landholding
peasants who no longer could assign their land to the crops they deemed
to be most profitable. This was in contrast to the tradition of free labour
that had existed since the Dutch Republic and at variance with the prin-
ciples of colonial rule that had been announced, shortly after the Nether-
lands Indies had been returned to the Dutch by British in 1816. The history
of colonial Indonesia shows that if there was any movement from unfree
to free labour it was a trajectory that was far from linear but filled with
twists and turns.
Van den Bosch’s intention, I will argue, was to relieve the Javanese from
feudal servitude and from the thralls of moneylenders, and make him a
productive subject of the Dutch empire. Over time the emerging planta-
tion economy seemed to vindicate Van den Bosch’s policies. Ever larger
numbers of Javanese became coolies in the rapidly growing complex of
cash crop production. In the 1860s the burgeoning sugar sector even led to
tightening labour markets in East Java, where sugar factories had to offer
advance payments to attract coolies. Again, the issue of forced labour
resurfaced on the colonial agenda. This time to prevent coolies from run-
ning away with their earnest money without performing work. For that
reason the factories demanded and got from the colonial government a
penal sanction on breaches of labour contracts, in spite of the fact that it

1 I would like to thank Jan Lucassen, at the International Institute of Social History, for
his suggestions with regard to the intricacies of labour law in the Dutch Republic and the
Kingdom of the Netherlands.
64 ulbe bosma

was a violation of Dutch labour law. A few years before when the Dutch
Parliament had formalized the right to collective action it had already
decided that there was no place for penal sanctions in ordinary labour
contracts. It could not come as surprise therefore that it did not accept
that the government of the Netherlands Indies had introduced the penal
sanction as a measure to enforce labour discipline.2
And yet, only one year later the penal sanction was introduced for the
Outlying Provinces, and for East Sumatra in particular. In this case, how-
ever, it did not originate from a labour regulation but as an addendum
to a decree ordering civil servants to check whether coolies had been
engaged on a voluntary basis. The first version of a regulation on recruit-
ing practices dated from 1868. In 1880 penal sanctions were added to these
regulations and the whole set was given the name of Coolie Ordinance,
which had to safeguard coolies against deceit and the employers against
their labourers running away before they had worked off their recruit-
ment costs. The penal sanction in practice enforced the already existing
immense inequality of power between employers and labourers, the for-
mer being European citizens, the latter colonial subjects.3 In the early
twentieth century the continued existence of these coerced labour condi-
tions were defended by employers who insisted on the crucial role of the
Sumatra plantations for the colonial and imperial economy.
The Sumatra employers compared the position of their coolies to
Dutch sailors, whose freedom of movement was also severely restricted
by law. This was an argument that began to backfire in a rather peculiar
way once the ILO (International Labour Organization) was established in
1919. As part of its campaign against slavery and forced labour, the ILO
also began to critically review the internationally customary penal sanc-
tions in the sailors’ labour contracts. There is some irony in the fact that
while international legislation could not enforce the abolition of the penal
sanction in the Netherlands Indies, it did compel the Dutch metropolitan
government to attune its own maritime labour legislation to more liberal
standards of international law.
This paper consists of four parts each covering a particular phase in the
development of labour relations on Java from the early nineteenth century
to the end of Dutch colonial rule. I will first describe how Van den Bosch
developed his ideas about forced labour as a bridge between vagrancy and

2 Kalma, Het arbeidscontract, 11.

3 Bosma, Karel Zaalberg, 100.
 dutch imperial anxieties 65

civilisation in the Dutch metropolitan context. The second part focuses

on the implementation of the Cultivation System on Java and details how
the existing systems of traditional tributes were reshaped into a colonial
system of land rent and forced labour. As a result a fast growing army of
marginal and landless peasants found employment in the emerging plan-
tation economy on a wage basis. The third section addresses the question
how to discipline this labour force of rural workmen and women on the
plantations, which was no longer tied by links of bondage or dependency
to Javanese (village) elites. The final stage is after World War I, when the
Indonesian nationalist movement emerges. In this phase the right to col-
lective action becomes a contested one and is eventually practically sup-
pressed by the colonial government in 1923. This was done with reference
to the paramount interest of preserving the colonial order and economy
in particular. Likewise, ‘paramount economic interests’ and ‘extraordinary
circumstances’ were used as arguments by colonial enterprise and govern-
ment to preserve the penal sanction to regulate labour relations on the
Sumatra plantations.

From Free Wage Labour to Work As Fulfilment

The intricate connection between personal freedom and the rise of free
labour (in this context labour free from non-pecuniary compulsion)
has been amply discussed, for example by Steinfeld in his book on free
and unfree labour in the United States and Great Britain.4 Contrary to
the picture Steinfeld is giving us about the Anglo-Saxon world, where the
concept of free labour only emerged in the nineteenth century, in the
Netherlands labour migration was free, labour contracts were free and
there were no penal sanctions on breach of contract during the days of the
Dutch Republic. Particularly its western, maritime and highly urbanised
provinces was one of the few parts of early modern Europe where unfree
labour was absent. Many artisans’ and a limited number of journeymen’s
guilds existed. Moreover, labour conflicts, including strikes have been
documented.5

4 See Steinfeld, Coercion, Contract.

5 According to Lucassen ‘the Dutch labour market was one of the few, or maybe the
only labour market of the Ancien Régime which was completely free’. See Lucassen, “Labour
and early modern economic development”, 395. See also Lucassen “Mobilization of labour”
and Lucassen, De Moor and Van Zanden, “Introduction”.
66 ulbe bosma

Under the Dutch Republic two types of labour contracts existed. One
regulated the position of domestic servants, who were part of the house-
hold and the other the subcontracting of all other wage earners. Although
local in character, because of the federal structure of the Republic, and
therefore showing substantial differences, some features were common.
Subcontracting labour was seen as an integral part of merchant law, i.e.
a contract between equal parties and was sanctioned by civil law. Urban
servants were free to engage in wage labour and both parties had access to
the local courts for enforcement or dissolution of the contract. Sanctions
were of a pecuniary nature.6
The free wage labour conditions that were ubiquitous in the highly
urbanised western part of the Dutch Republic were more or less extended
to the urban enclaves in the colonies.7 Sailors of course being the excep-
tion to the rule, an exception that was based upon labour regulations that
were adhered to by a number of European seafaring nations and dated
from the medieval times. Interestingly enough, however, the Dutch East
Indies Company (voc) did not differentiate between Asian and European
sailors as far as their labour conditions were concerned.8 In the Asian
port cities of the voc freedom was a privilege for city-dwellers who were
exempted from corvée service and for persons who were in the position
to exact services from other people. By the late seventeenth century half
of Batavia’s (contemporary Jakarta) population consisted of slaves.9 The
Dutch East Indies Company itself was the most important slave trader
in the archipelago if not in the entire Indian Ocean. The total volume
of the Dutch Indian Ocean slave trade has been estimated by Vink as
being as large as 15–30 per cent of the Atlantic slave trade.10 In the urban
environment they were usually part of the household, worked as artisans
or clerks, and could easily become part of a grey zone between slavery
and freedom, where he or she could earn money to pay for his or her
­manumission.11 The slave trade on Java was abolished in 1811 by the British
during their interregnum over the island. Slavery in the European realm

6 In the nineteenth century the Dutch versions of the Code Napoléon featured some
novelties like the abolishment of the guilds and in fact the impossibility of any sort of
‘combination’: trade unions and strikes were officially allowed only from 1855, respectively
1872, although de facto much earlier, see Brugmans, Arbeidende klasse, 184–9, 249–52.
7 Lucassen, “Labour and early modern economic development”, 383.
8 See Lucassen, “A Multinational and its Labor Force”, 20–1.
9 Vink, “ ‘The World’s Oldest Trade’,” 148.
10 Ibidem, 168.
11 Bosma and Raben, Being “Dutch” in the Indies, 47–9; Blussé, Strange Company.
 dutch imperial anxieties 67

of Java ­continued to exist until 1 January 1860, as a matter of fact, but

was a marginal phenomenon and largely confined to domestic services
in coastal towns.12 To the institutes of slavery and servitude under the
Javanese rulers I will return in the next section.
The first Constitution of the Netherlands Indies of 1818 was still written
in a liberal mode and inspired by the ideas of free labour and free labour
contracting.13 But this liberal spirit did not last long, primarily not because
of developments in the Netherlands itself. Widespread unemployment and
indigency reigned over the newly created Kingdom of the Netherlands. In
this situation the issues of freedom and (unfree) work became redefined
by a peculiar blending of Enlightenment ideas with Ancien Regime and
Calvinist notions. The result was the idea of work as fulfilment, which
served as the motivating force behind Benevolent Societies. These asso-
ciations wanted to relieve people from poverty, which they considered
to be a dehumanizing condition, and to educate them to an arbeidzaam
(industrious) existence. This was done, for example, by establishing beg-
gar colonies in the Netherlands. From 1818 onwards indigent from Amster-
dam, Rotterdam and other large Dutch cities found themselves relocated
to endless moors in the northern provinces of the Netherlands in so-called
‘peat colonies’.
These peat colonies would become directly linked to the Dutch colo-
nial policies, as they were created by Johannes van den Bosch (1780–1844).
In 1830 he became Governor-General of the Netherlands Indies, where he
inaugurated the forced Cultivation System. Van den Bosch was inspired
by Philipp Emanuel von Fellenberg, the Swiss ‘Volkserzieher’ whom he
even visited at his model farm near Bern. The idea of coercion was alien
to these experiments to edify the paupers, but Van den Bosch made a dis-
tinction in this respect between paupers, beggars and vagrants. The latter
category being defined as indigent without permanent address. Begging
was prohibited in the Netherlands since the Napoleonic times. Under the
influence of the Enlightenment the idea took root in Dutch society that
pauperism had to be banished, begging criminalised, and that indolence
was a cause of poverty that needed to be cured. This very idea that the
government had the right to put people to work for their own good and

12 An exception is the employment of slaves at the nutmeg plantations of the Banda
islands, at the time of abolition there were still some 1200 plantations slaves here. See for
example Bosma and Raben, Being “Dutch” in the Indies, 155.
13 See GB of 1819, Staatsblad van Nederlandsch-Indië, 5 January 1819, no. 10.
68 ulbe bosma

the commonwealth was transplanted to the Netherlands Indies by Van

den Bosch.
To put the record straight: Van den Bosch was in favour of the abolition
of slavery. Moreover, he was criticized by conservative contemporaries
who condemned his attempts to emancipate the poor by giving them
work, which in their vision was a subversion of the social order created by
God, which was best served by charitable institutions. On the other hand,
more liberal minded contemporaries considered Van den Bosch to be a
conservative who refused to distinguish between coerced labour and free
wage labour. As most of his upper class contemporaries, Van den Bosch
saw free labour only existing under conditions of economic independence,
which in fact made every form of wage labour qualify as unfree.14 This in
fact was a deviation from the notions of labour contract and free wage
labour that had existed since the Dutch Republic.
Again, the Dutch situation is at variance with the widely accepted
notion that from the early nineteenth century citizenship and free labour
contracting became linked to each other in a progressive way. It is pre-
cisely at this assumed linearity that Prakash directs his well-known cri-
tique about the European binary of free and unfree, which in his view
had been written on Indian history like it had been on African or Javanese
history for that matter.15 Prakash perceives a Eurocentric linear perspec-
tive that projects slavery onto antiquity and freedom onto the future, a
perspective that was taken for fact by civil servants in different European
colonies.16 The Dutch trajectory, however, shows that Europe was far from
homogeneous on this account. The emergence of new social ideas and
economic concepts or the occurrence of unprecedented poverty, as was
the case in the Kingdom of the Netherlands, could drastically shift percep-
tions on labour issues and even legitimize new types of forced labour. In
the Dutch case the early nineteenth century saw a shift in emphasis in
thinking from labour that can be contracted to labour as a constraint. Van
den Bosch’s involvement in Benevolent Societies and the Cultivation Sys-
tem bears witness to the fact that he was familiar with the idea of labour
as a constraint, a necessary condition to reach an acceptable level of civili-
sation both at the individual as well as the collective level. As Schrauwers

14 The notion that all wage labour was in a way coerced labour was also shared by
many working class people in the United States and Great Britain according to Steinfeld.
See Steinfeld, Coercion, Contract, 13.
15 Prakash, Bonded Histories 8–9.
16 Banaji, Theory as History, 150.
 dutch imperial anxieties 69

has noticed: ‘Viewed in this light, Van den Bosch’s colonial regime estab-
lished an alternative trajectory toward the creation of a free labour market
than that documented by the English poor law and anti-slavery agitation’.17
The complexities of Javanese society would make Van den Bosch’ trajec-
tory towards free labour a rather contorted one, however.

The Pedagogy of the Forced Cultivation System on Java

Van den Bosch’s Cultivation System was imposed under global market con-
ditions and went along with the massive claims on Javanese land. On top
of that came the unequal protection under law. The European employers
were the citizens and guardians of the colonial order making the island
profitable for the Dutch exchequer, promoting the welfare of the colonial
subjects and protecting them from abuse. Whereas Dutch citizens in the
Dutch East Indies fell under Dutch law, the Javanese were merely colonial
subjects, which in the nineteenth century meant amongst other things
that they did not enjoy the same legal entitlements such as recourse to
an independent judiciary, fell under different tax regimes and hardly got
any education. Attempts at creating legal protection for the peasants and
other workers by well-meaning colonial civil servants therefore faltered
on the unequal legal status between colonial subject and European citizen
and the economic power of the latter.
The Cultivation System basically subsumed existing systems of corvée
and dependence. Systems of slavery were practically irrelevant in this
respect. Beyond the cities and domains of the voc on Java slavery was
rather marginal. Victorious rulers and war lords within the disintegrating
Empire of Mataram could carry away populations to put them work at
the construction of palaces or fortifications.18 Pawnship (cacah-kawula)
for food was a rather common way of enslavement. One could also be
sold to the Sultan’s palace to be put to work there as a servant or con-
cubine. Slavery was however not an economic factor in Javanese society,
as Moertono argues, because the rulers could rely on corvée labour that
was abundantly available.19 These ‘traditional’ burdens on the Javanese
­peasantry were quite significant both in political as well as economic

17 Schrauwers, ‘The “Benevolent” Colonies’, 299.

18 De Graaf and Pigeaud, De eerste moslimse vorstendommen, 219.
19 Moertono, State and Statecraft in old Java, 147.
70 ulbe bosma

terms.20 In pre-colonial times the Javanese polities were always in need

of sources to feed themselves and in need of soldiers to expand or defend
their polities.
Until the eighteenth century Java was sparsely populated and the state
apparatus of the Mataram Empire was organised in a such way that each
department and thus each high official had its own district, or better a
group of people, from whom he exacted services. According to custom the
peasantry not only had to deliver 2/5 of their harvest to the Sultan and his
officials and of the remaining 60 percent another 1/5 for the desa (village)
notables, but also to maintain roads and partake in the retinue of the
Sultan’s army. This left the cultivator theoretically with about half the pro-
duce of his lands. The actual situation was quite different however. Only
sawah (irrigated land) holding peasants were due to these taxes, which
were in fact based upon a cacah—originally the smallest unit that could
supply military later one a unit producing agricultural crops. The cacah-
holding peasant, the sikep, had usually a number of dependents to whom
he could shift part of the burden. What made taxes less onerous than
paper, finally, was that the Javanese rulers simply lacked the apparatus to
lay claim to all lands, and particularly not to newly developed lands. The
peasants were therefore able to conceal substantial parts of their income.21
Things might have slightly changed for the worse, however, when by the
late seventeenth century forced deliveries to the Sultan were transformed
into a monetized land tax that were quite often collected by Chinese tax
farmers.22
From the seventeenth century onwards the voc had used the rights it
had obtained via treaties or as the new sovereign to conscript labour to
build and maintain infrastructural works. Each year 400 to 450 Javanese
labourers were conscripted to clean the canals in Batavia, for example.
Gradually a system of contingents of deliveries of cash crops was extended
over West-Java that tapped into the existing patrimonial administrative

20 In my view unfree labour before the arrival of the colonial state, is not as Prakash
and Miers & Kopytoff would have it, just a social and cultural phenomenon within the
confines of a personal relationship. The argument of Jan Breman that the patron-client
relation between landowners and landless in pre-colonial and a colonial India had the
extra dimension of a permanent claim on labour of the latter, seems to me quite convinc-
ing and also applicable to the situation on Java. Likewise, the slave trade and slavery in
Africa was institutional and part of economic life. Prakash, Gyan, Bonded Histories 7; Miers
and Kopytoff, Slavery in Africa; Breman, Labour Bondage in West India; Lovejoy, Transfor-
mations in Slavery, 24.
21 Ibidem, 73.
22 Hoadly, Towards a feudal mode, 142.
 dutch imperial anxieties 71

tax system.23 The bupati and mantri, the high regional officials of the
waning Mataram Empire, were ordered to deliver quantities of coffee and
European horticultural experts were attached to these functionaries. The
whole system of forced cultivations and contingents that had been intro-
duced under the voc was at variance with the process of monetization of
taxes that had already taken place under the Mataram Empire. A moneti-
zation, however, that had also demonstrated its dark sides. For the smaller
peasants it was difficult to pay theses taxes or land rents in cash, and
here the Chinese and other moneylenders came in to provide advances
on the crops against high interest rates of course.24 It was precisely for
this reason that Van den Bosch, unlike most of his contemporaries, did
not denounce the voc system of contingents. Conversely, he severely
criticized the system of land rent imposed during the British interregnum
under Lieutenant-Governor-General Thomas Stamford Raffles (1811–1816).
The peasantry had great difficulty to produce these taxes, which made
them vulnerable to moneylenders. Moreover, the corvée for local officials
had all but disappeared.
The hodge-podge of taxes, land rent, bondedness and different corvées
that existed in early nineteenth century Java was a disaster according to
Van den Bosch. But as a remedy he did not expect anything from the
introduction of sound liberal economic principles, because in his view
these would not work in a Javanese society that was steeped in despo-
tism. His own proposal to impose a system of forced labour and forced
cultivation—he literally used these words—at first sight could hardly be
considered as less despotic, but he made it palatable by presenting the
Cultivation System as an apprenticeship of some sorts, a social remedy
against feudalism.25 He allegedly had declared in 1830 that forced cultiva-
tions should have been applied just for ten years, after which ‘free labour’
should have become the norm.26 I have no evidence to corroborate this,
but what we know for sure is that a rapid introduction of wage relations
in the forced sugar cultivation took place from the early 1830s onwards
when Van den Bosch was still in office as Governor-General. Moreover,
wages and monetization were a basic feature of the Cultivation System.

23 After the Priangan in West-Java had been officially handed over to the voc by the
Sultan of Mataram, the Susuhunan, in 1705, forced cultivation of coffee was introduced.
Hoadley, Towards a feudal mode of production 48, 106. See also Breman, Koloniaal profijt.
24 Carey, “Waiting for the ‘Just King’,” 99.
25 Van den Bosch, Nederlandsche bezittingen ni Azia, Amerika en Afrika i, 222.
26 Vitalis, Opmerking omtrent den loop der suiker-industrie, 62–3.
72 ulbe bosma

The Javanese cultivators had to pay their land rents from their crop pay-
ments. In practice the former was often deducted from the latter, in prin-
ciple however the cultivator would be enabled by the government to pay
for his land rent.27 This aspect seems to me an underrated element of Van
den Bosch’s ‘pedagogy of forced labour’ as it was crucial to introduce the
Javanese peasantry into the workings of the wage economy.
When Van den Bosch presented his plan for the Cultivation System
to the Dutch King, he was aware of the fact that it would be against the
existing communis opinio in the Netherlands. But for most of the Dutch
proletarians, living under constant fear of starvation, the whole concept
of free labour did not mean anything, he argued.28 In his analysis most
of the labourers in densely populated countries like England and the
Netherlands had practically no access to means of production, which in
practice reduced them to the same social status as slaves.29 This point of
view had been foundational for the Dutch peat colonies and now became
extended to Java. What Van den Bosch believed to see was that the Java-
nese peasantry only needed a few hours work per day for subsistence. The
Cultivation System was an exercise in teaching the Javanese population
industriousness, and at the same time a mechanism to make the Nether-
lands Indies profitable for the Dutch exchequer rather than a permanent
loss. What Van den Bosch was aiming for was to unchain an industri-
ous revolution in the tropics, so to speak, and to replace local circuits of
moneylenders and tax farmers by an economic circuit between colony
and metropolis. The semi-governmental Dutch Trading Society played a
central role in connecting Dutch and Javanese markets and workers: sup-
plying the Javanese cultivators with advances, bringing commodities to
Europe and calicos back to Java and thus encouraging the cotton industry
in the Eastern part of the Netherlands.30
At any rate the colonial requisitioning of labour did not right away
replace ‘traditional’ dependency relations. In practice, the rural popula-
tion of Java suffered under the double burden of cultivation conscription
and traditional corvée well into the nineteenth century. The Javanese
local power brokers did not give up their rights easily, and there was no
urge for them to do so, because the Dutch colonial government had to
rely on them. The resistance and also massive desertions from West and

27 See Fasseur, The politics of colonial exploitation.

28 [J. van den Bosch] “Nota”, 316–317.
29 See Berends et al., Arbeid ter Disciplinering.
30 See Schackmann, De proefkolonie.
 dutch imperial anxieties 73

­Central Java to the more sparsely populated areas of East Java amply dem-
onstrated that the Javanese peasantry did not consider the introduction
of the Cultivation System as a first step out of despotism into the realm
of freedom.31 In fact, it turned out to be far more burdensome than past
arrangements. More land and more persons were included in cultivation
conscription (cultuurdiensten) than in the corvée of Javanese rulers or the
landrent schemes that had been imposed by Raffles. Only the elderly, sick
and priests were exempted, whereas peasants who did not hold sawahs
were also subject to cultivation conscription, although to a lesser degree.
The Cultivation System became thus far more encompassing than past
taxation schemes, and though cheating the authorities by underreport-
ing the number of villagers and the acreage of land did continue, control
over the rural areas became tightened up in the course of the nineteenth
century. Moreover, the burden of growing cash crops often turned out to
be underestimated, as for example the soil became depleted from grow-
ing indigo or much more cane had to be brought to the factories than
initially had been reckoned with. And although the Javanese officials, the
lurah, mantri and bupati, got their share of the so-called cultuurprocenten
(cultivation bonuses) they still levied their tributes from the people in
their resort. It took time to transform the patrimonial Javanese state into
a modern bureaucracy.32
Another aspect that might seem to contradict Van den Bosch’s intention
to familiarize the Javanese with the pursuance of his own economic inter-
ests is the fact that the Cultivation System both strengthened communal
tendencies and inequalities in village life. The village elite had been put in a
more powerful position towards their fellow desa men and women than ever
before. In 1838 European agricultural and other enterprises were allowed
to negotiate contracts with the ‘eldesten en most prominent’ members of
the Javanse village (the desa) for the delivery of labour. This was a funda-
mental departure from the principles of the Constitution of Netherlands
Indies of 1818, from which the regulation about individual labour contract-
ing had been derived.33 Burden sharing was also strongly promoted by the
Cultivation System, without becoming egalitarian however. J.H.F. Sollewijn
Gelpke, a civil servant who had made extensive study of Javanese village

31 Bosma, “The discourse on free labour and the forced Cultivation System,” 400.
32 See Sutherland, The making of a bureaucratic elite.
33 GB Stbld. 25 December 1838, no. 50, Van Geuns, De suikeronderneming Djatiroto, 28,
35. Although the possibility of hiring coolies on an individual basis continued to exist, it
meant a firm deviation from the principles of the GB of 1819, Stbld. 5 January 1819, no. 10.
74 ulbe bosma

in the late nineteenth century, mentions that sugar planting desa’s were
quite willing to accept additional villagers to become sawah holders, but
only after they had first been working on the village cane fields.34
The Cultivation System indubitably featured communalisation of land-
holding, but it should not detract us from a far more fundamental develop-
ment, namely the proliferation of wage labour, which became particularly
visible in the sugar industry.35 Under the Cultivation System sugar manu-
factures were contracted to process the cane that was delivered from the
villages assigned by the colonial government to grow cane. The factories
had to sell the sugar back to the government, which via the Dutch Trading
Society took care of the shipment to the Netherlands. In the sugar sector
Van den Bosch’s vision of creating a wage economy was making headway
in the 1830s, as we will detail at the following pages. Whereas in the initial
years of the Cultivation System practically the entire payment by sugar
factories had been in crop payments plant (a combination of wage and
lease), in 1886 the sugar industry paid 3/4 in coolie wages and only 1/4 in
land leases. The share of wage payments would still grow over the decades
to come.36 The intense debates by colonial experts on the issue of ‘free
labour’ should be considered against the backdrop of the rapid advance
of wage labour relations.

Partial Proletarianisation and the Penal Sanction on Java

Wage labour is not necessarily synonymous to free labour and vice versa.
In the eyes of the Javanese country man or woman the free labourer was
not the proletarian but the gogol, the peasant who owned his or her land
and did not need to rent out his labour to other employers.37 The free
labourer was someone who owned sufficient land for his subsistence, and
he or she was in declining order followed by the tenant and the share
cropper, whereas the landless dependent worker was tailing this ranking.
The proletarian, the quintessential ‘free labourer’ in the Marxist sense,
hardly existed in Javanese rural society, since most labour was absorbed
in some kind of dependent relationship. Apart from a small but growing

34 See Sollewijn Gelpke, Naar aanleiding van Staatsblad.

35 About the advance of wage labour in the Cultivation System see Elson, Village Java,
179–226.
36 Suermondt and Van den Berg, Nota over de suikerindustrie, 11.
37 See Sollewijn Gelpke, Naar aanleiding van Staatsblad, 1878 No. 110.
 dutch imperial anxieties 75

stratum of itinerant workers and urban proletarians landless rural workers

were usually dependents on the desa elites.38 In the course of the nine-
teenth century the class of dependent sharecroppers and entirely landless
Javanese would increase. The strategy of the Dutch colonial government
and private enterprise was to tap on this labour reservoir and reduce the
claims of the Javanese officials and notables on tribute labour. Over time,
as Knight has pointed out, a reduction of their rights to corvée labour had
taken place, which was very much in the interest of the factories who
complained that it were these rights of local chiefs that stood in the way
of the advance of wage labour, which they assimilated with ‘free labour’.39
It all fitted the transformation process of the Javanese from a landowning
taxpayer, which he knew he was for centuries, to a coolie.
Sugar factories were in favour of wage labour but not at all wary if this
had been obtained under coerced conditions. For them free labour just
meant that the labouring masses of Java were liberated as much as possi-
ble from corvée, tribute and extractions they considered to be parasitical.
For the same reason, the increasingly powerful sugar factories developed
a preference for individual labour contracting rather than recruitment
and payment via the desa elites. It is no coincidence that the regulation
about collective labour recruitment in the desa was withdrawn in 1863, by
a Minister of Colonies who had been administrator of a sugar factory him-
self and who had been keen to pay directly to factory coolies rather than
to the desa elites. Otherwise the coolie would get hardly anything, he had
observed.40 On the other hand, the lurah could not be missed by the sugar
factories. And thus he continued his job of recruiting dependent labour in
the village for the sugar factories who provided him with a premium for
every worker he could deliver.41 What sugar factories understood by free
labour came down to a transformation of the lurahs from fiscal officials
into a kind of mandurs (recruiter/foreman) who had to deliver a certain a
number of workers. The agenda of the factories was to reshape the village
into a labour reservoir for the factory.42
Meanwhile, the sugar factories were busy to wrestle the sawahs (irri-
gated rice fields), the means of production of a Javanese cultivator, out
of his hands against sub-economic prices. The Javanese peasant usually

38 See Carey, The power of prophecy.

39 Knight, “Peasant Labour and Capitalist Production,” 259.
40 Fransen van de Putte, De regeling en uitbesteding, 30 ff.
41 Knight, “Peasant Labour and Capitalist Production,” 255–6.
42 Margana, “Hybridity,” 101.
76 ulbe bosma

leased his land to the sugar factory not because he expected to gain from
it, but under the compulsion of indebtedness, the pending payment of his
land rent, or outright coercion by the village head. These facts were well-
known and stated again by a government investigation in 1894.43 More-
over, the price the Javanese peasant could ask for his land was lower at
the time of puasa (ramadhan) and at the time the landrent had to be paid.
This gave the factory an enormous advantage, because with a good timing
it could negotiate with people who were in dire need of money.44 This was
not changed by the emergence of credit cooperatives (volkskredietbanken
or desa lumbungs), which had been actively promoted by the Indische
government as part of its welfare policies that had been inaugurated in
1901 under the name of Ethical policies. Around 1920 it was still primarily
indebtedness that made a peasant lease out his land to a sugar factory.45
It also still happened that impoverished Javanese rented their land out
against a pittance to rich countrymen, who in turn gave this land in lease
to sugar factories.46 This partly undid the protection which the colonial
government had given to Javanese peasants by not allowing Europeans or
ethnic Chinese to have land in property. Even worse had been the effects
of the Agrarian Law of 1870 that allowed for long term leases of 75 years
to establish plantations in the middle of Java on so-called waste land but
that often enclosed villages.
Meanwhile, the government of the Netherlands Indies did its part to
facilitate the transition from tribute to tax. After 1854 it embarked upon
a scaling down of the corvée labour for the maintenance of public works,
a process that was completed only in 1916. Those conscriptions services
were however replaced by monetized taxes. This idea was wholeheart-
edly endorsed by colonial business circles as yet another step towards free
labour for plantation estates, and compelled Javanese to do even more
wage work in order to pay their taxes. It was difficult in the 1860s to get
sufficient coolie labour particularly for the sugar industry, because often
the planting of paddy, the preparation of the fields for the new cane and
the grinding season coincided. Again, the colonial government came to
assist the plantations and the sugar factories in particular. In the early
1860s one could get remission from the corvée imposed by the Indies gov-
ernment for public works, if one would perform coolie services. Wealthier

43 Huender, Overzicht, 90–92.

44 Kohlbrugge, Is grondverhuur, 64.
45 Huender, Overzicht, 91.
46 Verslag van de suiker-enquête commissie, 167.
 dutch imperial anxieties 77

farmers pressured their dependents, often landless, into coolie work to

exempt themselves from these duties. This was the preferred way, because
if factory mandurs had to recruit labour they had to advance consider-
able sums, and it was not always easy to make the coolies honour their
­commitment.47 That labourers who had been attracted with advances—
traditionally sealing a verbal labour contract—did not always honour
their commitment was a reason for employers to appeal for penal sanc-
tions to be enclosed in labour contracts to ensure that coolies would not
simply walk away with the money. The advance was also known under
metropolitan Dutch labour law as handgeld (earnest money) or godspen-
ning (queen’s shilling) sealing the labour contract.
In Dutch law there was no such thing as a ‘masters and servants’ act,
and although the servant or hired worker was not equal to his master
before the law, penal sanctions did not exist.48 As said before, in the days
of the Dutch Republic a distinction existed between regulations on ser-
vants who were part of the household and subcontracting, a distinction
which survived Napoleonic times and continued into the nineteenth cen-
tury. Gradually a more paternalistic relationship emerged in which the
subcontracted workers came in more or less the same position as the
domestic servants. Interestingly enough, and quite similar to the situation
in Great Britain, in the Netherlands the legal definition of a servant was
extended from the quotidian ‘domestic servant’ to hired worker in the
course of the nineteenth century.49
In the nineteenth century regulations between masters and domestic
servants could still be subject to provincial or even municipal regulations
within the Dutch Kingdom and its colonies. Such a local regulation was
the Surabaya police ordinance of 1829 that determined that every domes-
tic servant on Java who quitted from his temporary contract could be
sentenced to time in prison or fined to 25 guilders, and if he was a native
or foreign oriental with 8 days block or 30 beatings with a rattan.50 The

47 Elson, “Sugar factory workers,” 153, 166–7.

48 Heerma van Voss e.a., Asser’s handleiding, Vol. 2, no. 2. As a matter of fact, only from
1 February 1909 did the Netherlands have a full-fledged law on labour contracts. Lourens
and Lucassen, Lipsker, 38–41. For the Law on the Labour Contract (Wet op het arbeidscon-
tract, ter vervanging van BW art 1637–1639) see Diepenhorst, De Nederlandsche arbeidswet-
geving II, chapter I and see Smissaert, De arbeidsovereenkomst.
49 See for the multiple meanings of servant in England Steinfeld, “The Invention of
Free Labor”, 17. For the Netherlands see Lourens and Lucassen, “Lipsker op de Groninger
tichelwerken,” 40; Smissaert, “De arbeidsovereenkomst der dienstboden”.
50 Indisch Staatsblad 1851, no. 26. Van Geuns, De suikeronderneming Djatiroto, 37, 39.
78 ulbe bosma

­Surabaya ordinance was adopted for the entire island in 1851, which in
itself was not a problem as long as it had been confined to domestic ser-
vants. In 1872 it was however extended to Javanese workers at all kinds of
enterprises, which led to an indiscriminate application of the penal sanc-
tion to all indigenous wage workers of Java. This went way too far for the
Dutch Parliament, which in this very year 1872, had just abolished the pro-
hibition of collective action (e.g. unions and strikes) that had been part of
the Code Napoléon. This abolition was attended by the explicit statement
in the draft of the Bill that there was no reason to attach penal sanctions
on either labourers’s or employers’s breach of contract.51 In the same vein
Dutch Parliament decided in 1879 that labour regulations concerning the
position of indigenous (inlandsche) servants and labourers should not be
part of the Penal Code for the indigenous people, but that the position of
indigenous servants and workers, as well as sailors, should be arranged
in accordance with the Dutch Civil Code.52 In 1879 The Hague overruled
Batavia by determining that only walking away with an advance while not
having any intention to perform labour was punishable.53 This was hard
to prove of course. On Java the penal sanction did not resurface, but in the
Outer Provinces it would as we will see in the next section.
The issue of the penal sanction would become less pressing for the
sugar factories, because after 1880 labour shortages became less acute.
Since then coolie wages on Java hardly rose nominally, and even began
to decline in real terms after 1900, to slide down to the level of hunger
wages in the Depression years.54 In the early twentieth century the sugar
industry came to rely on a growing contingent of (migrant) workers who
did probably most of the cutting, haulage from the field and coolie work
at the sugar factory. But an important share of the work in the field, the
planting and tending of the cane, was done by female labourers, most
of whom came from smallholder households.55 In the early ­twentieth

51 Van Houten, Regtstoestand, 7; Brugmans, Arbeidende klasse, 184–9, 249–2.

52 The Penal Code for indigenous was based upon the European Penal Code, but tra-
ditional and religious right could be appointed as advisers to the judge. Kleintjes, Het
Staatsrecht II, 51, 57.
53 Van Geuns, De suikeronderneming Djatiroto, 37, 39; Levert, Inheemsche arbeid in de
Java-suikerindustrie, 92; Kalma, Het arbeidscontract met den inlander 11, 73; Wertheim,
“Planters tegen koelies.”; See also Staatsblad van Nederlandsch-Indië 1872, no. 111, art. 2,
no. 27. The “Soerabaja Politiereglement”, Staatsblad van Nederlandsch Indië 1829, no. 8
and Staatsblad van Nederlandsch-Indië 1851, no. 26.
54 Coolie Budget Commission, Living conditions of plantation workers.; Scheltema, The
food production.
55 Knight, Colonial production, 33–4.
 dutch imperial anxieties 79

c­ entury ­probably more than half of the cane planters were landless peas-
ants (including female) employed by the ‘big peasants’ who held an
increasing share in the ‘communal’ sawahs or increased their individually
owned sawahs.56
Nonetheless, as a concomitant of emerging nationalism and Indonesian
labour unions after 1910, labour issues got a new twist. Coolie wages that
were clearly below subsistence and sub-economic leasing of land became
sources of political contention and collective action. While the coolies of
Java had escaped from the penal sanction because of a lack of basis in
Dutch law, the situation was less positive as far as freedom of collective
action was concerned. Freedom of association was only tentatively estab-
lished on Java, and collective action had been drastically curtailed in the
Netherlands for crucial economic sectors. After the big railway strike of
1903 the right to strike had been withheld from civil servants workers in
economically crucial sectors in the Netherlands. This would set a prec-
edent for the colony. In order to ensure the orderly conduct of the sugar
campaign of 1923 the Indies government outlawed the Union of Indone-
sian Sugar Workers (Personeel Fabriek Bond) and practically annihilated
the Railway Union (VSTP). It was done with a regulation prohibiting ‘all
strikes that could lead to disorganisation of public life or severe economic
consequence’.57 This time objections in Dutch Parliament were not, and
since 1903 could not be, strong enough to repeal such a clause that seri-
ously violated the principle of free labour.
The suppression of Indonesian labour unions was part of a general
repressive course on which the government in The Hague had embarked
after it had disavowed the more forthcoming attitude of the internation-
ally oriented Governor-General Count Johan Paul van Limburg Stirum
(1916–1921). As a diplomat by profession Van Limburg Stirum had been
susceptible to a the new international atmosphere with regard to colonial
self rule and colonial labour issues, epitomized by the League of Nations.
He had announced in November 1918 that the penal sanction for the Out-
lying Provinces would be abolished by 1 January 1926.58 In 1919 he installed
a committee to investigate the labour unrest in the sugar industry and
another one to investigate the desirability to establish a minimum wage
for coolies on Java.59 Again, with regard to political emancipation and

56 Ibidem, 22.
57 See Ingleson, “ ‘Bound Hand and Foot’.”; Bosma Karel Zaalberg, 369.
58 Middendorp, Twee achterlijke arbeidssystemen, 35.
59 See Verslag van de arbeidscommissie.
80 ulbe bosma

labour issues of the people of the Netherlands-Indies his administration

had steered a more progressive course than the metropolitan government.
When this Governor-General left office the announced abolition of the
penal sanction was in danger.

Penal Sanctions and Coolie Scandals on Sumatra

Since the mid-nineteenth century a growing army of Javanese migrant

labourers found their way to the sugar plantations, the harbour towns and
finally to the plantations of Sumatra. Breman’s and Stoler’s well known
publications on the coolie scandals on Sumatra and conditions on the
East Sumatra plantation belt have attracted attention far beyond the circle
of Indonesianists.60 Colonial enterprise defended the penal sanction by
referring to the high costs for recruiting labour and the need to guarantee
a steady supply to this crucial economic sector, in fact elevating their spe-
cific interests to a national cause. They also claimed that there was room
for exceptions to the principle of free labour, since in the Netherlands civil
servants in crucial economic sectors were not allowed to strike and sailors
were not permitted to leave their employer during their engagement.
Like other colonial governments the government of the Netherlands
Indies considered labour contracts with penal sanctions the lesser of two
evils. In order to combat recruitment malpractices, they had taken on the
task of regulating a phenomenon which was at variance with the principle
of free wage labour. In order to ‘regulate’ migratory labour recruitment,
the colonial state had now given legal status to labour contracts issued
by official recruiters and put a penal sanction on leaving work without
the employer’s consent.61 But the Dutch East Indies Coolie Ordinance did
not protect the coolie against mistreatment and continuing indebtedness.
Moreover, recruitment excesses were widespread. Lurahs and other local
Javanese officials were paid for every coolie whom they could deliver to
the recruiters.62 The penal sanction in practice gave the plantation man-
agement the authority of the police; if coolies fled the dreadful circum-
stances at the plantations plantation guards went after them and brought
them back without any interference of the police. If coolies were sent to

60 See Breman, Koelies, planters en koloniale politiek; Stoler, Capitalism and ­confrontation.
61 See for example Jan Breman, Koelies, planters en koloniale politiek; Baak, “Enslaved
Ex-Slaves”; McKeown, Melancholy Order; Robb, “Labour in India”, 48; Roy, “Sardars”, 972.
62 Tideman, “De koelieordonnantie,” 43.
 dutch imperial anxieties 81

the police because of alleged breached of their contract, however, they

were routinely imprisoned otherwise it would be a loss of face for the
plantation management en therefore detrimental to ‘colonial prestige’.63
In its actual working the Coolie Ordinance was heavily if not almost
completely biased against the coolies. For every conceivable transgression
on their part it was explicitly mentioned that severe punishments could
follow, but plantation staff that was known for beating up the coolies regu-
larly had little to fear, as any real measures against them allegedly tainted
‘colonial prestige’. And thus the Coolie Ordinance had become a travesty
of earlier regulations aiming at preventing coolie recruitment from falling
back into systems of slavery. Since 1868, just eight years after the abolition
of slavery in the Netherlands Indies, the first regulations aiming to protect
Foreign Orientals from slavery and deception were issued. In 1872 these
were extended to native Indonesians and particularly targeted labour con-
ditions in Sumatra. The Coolie Ordinance of 1880 was plainly a system of
indenture, as it stipulated penal sanctions on leaving the plantation within
the fixed term of employment of three years.64 In spite of the announce-
ment made in 1918 by Governor-General Van Limburg Stirum about the
out-phasing of the penal sanction it remained in place until 1931 and when
it was abolished it was not because of humanitarian concerns or respect
for international conventions. No, the reason were purely economic. By
that time the Depression had set in and the employers were massively
laying off their contract coolies. Indeed, as Stoler observes, the abolition
of the penal sanction helped the employers to get rid of their labour with-
out paying them anything for their remaining contractual time.65 More-
over, the US Senate had adopted the Blaine Amendment prohibiting the
import of goods produced under forced conditions, which would be put
into effect from 1 January 1932 onwards. The initiative for the Amendment
was not rooted in humanity but a protective measure for American Vir-
ginia tobacco that suffered from competition by the Sumatra leaves.66
This as far as the actual facts of the matter are concerned. What con-
cerns us here is how the penal sanction could be reconciled with the fact
that under Dutch Civil law a labour contract was a contractual arrange-
ment between two equal partners. A fact that had become explicitly part

63 Ibidem, 58–9.
64 Koelie-ordonnantie van 13 juli 1880, Staatsblad van Nederlandsch-Indië 1880–113. See
also Bosma, Karel Zaalberg, 100.
65 See Houben and Lindblad, Coolie Labour; Stoler, Capitalism and Confrontation, 43, 88.
66 See E. Kupers, Het vraagstuk van de poenale sanctie.
82 ulbe bosma

of the Dutch law only in 1909, but which was nonetheless deeply rooted
in Dutch history. Moreover, when the Dutch Parliament had adopted the
law that legalised trade unions and strikes in 1872, it had also decided that
penal sanctions should have no place in an arrangement where the gov-
ernment was not a party. The penal sanction for the Sumatra plantations
had come in through the backdoor of the regulation of recruitment prac-
tices and was maintained on the grounds of exceptional circumstances
and paramount colonial interests. The fact that it was exceptional made
it already a legal anomaly, which became even more pronounced when
the ‘paramount colonial interests’ were themselves were redefined after
World War I. ‘Home Rule’ as adopted as the formal objective not only of
British but also of Dutch colonial policies in Asia. It is no coincidence that
the revisiting of the penal sanction was announced in the context of the
general reform of status of Netherlands Indies inspired by the Chelmsford
Montague reforms for India. The announcement of the Governor-General
van Limburg Stirum in 1918 about the abolition of the penal sanction was
part of a broader constitutional reform that would bring ‘Home Rule’ for
the Netherlands Indies, the so-called November Promise. The East Suma-
tra planters understood that they had to compromise and they came with
their own proposal for a new ordinance. They were ready to acknowl-
edge that the ‘coolie’ was a free subject and contracting partner, they also
admitted that Coolie Ordinance should be better calibrated, but their pro-
posal fell short of parting from the penal sanction.67
In principle, their position might have been accepted 15 years earlier,
as the Dutch Social Democrat and specialist on colonial issues J.E. Stokvis
wrote in a pamphlet against the continuation of the Coolie Ordinance. In
the early years of the twentieth century there had been Labour members
of Parliament who had been in favour of penal sanctions in Dutch labour
law, as long as these were applicable to employers too.68 But, Stokvis went
on, in colonial society where the (European) employers and (native) work-
ers were never equals, a penal sanction only engendered underpayment
and racialised violence. In fact, the labour migrants from Java to the Out-
lying Provinces were recruited to work against a payment just sufficient to
reproduce him or herself, which in Stokvis’s view came down to slavery.69

67 See Ontwerp eener Arbeidswetgeving.

68 Stokvis, Koloniale Schande, 4.
69 Ibidem, 5.
 dutch imperial anxieties 83

E. Kupers, Chair of the Social Democratic Labour Union and Member of

Parliament70 as well as one of the Dutch representatives at the meetings
of the ILO pointed out that it was the combination of the penal sanction,
the racial inequality and the fact that the recruited Javanese labourer had
no countervailing force, no labour union to stand for his rights, which
made the conditions in East Sumatra unacceptable. Reading the Coolie
Ordinance superficially, one might even have the impression that it would
protect the coolie, but the situation at the plantations was violent if not
murderous. It has been noted that around 1918 about 8 percent of the
200,000 contract coolies ended up in prison each year. This percentage may
have just slightly decreased to 5 per cent in 1926, thus in the final years of
the penal sanction. The Coolie Ordinance created a labour regime based
upon fear, punishment and violence.71 For this reason Albert Thomas, the
first director of the ILO condemned the penal sanction as a remnant of
servitude. Moreover, another critic pointed out, contracting labour for the
Sumatra plantations under the penal sanction was utterly uneconomical.
The plantation had to pay 100 to 150 guilders per coolie, while the coolie
wage per day was 55 cents. So, why not encourage free migration against
higher wages?72 The question has never received a serious answer and
deserves further research. I would suggest as a hypothesis that the very
fact that labour recruitment was a highly profitable business may have
played a role in the continuation of the penal sanction.
Indentured conditions to work of the fares were not at all uncommon
in the history of labour migration and was not confined to Asian labour.
Over the course of the nineteenth century white labour continued to work
under various degrees of bondedness or coercion in the so-called settler
colonies.73 Constraints and non-pecuniary sanctions were not absent in
European labour contracts either, particularly not in the maritime sec-
tor. The legal position of the sailors, as a matter of fact, had been subject
of international legislation since the 12th century Rolls of Oléron came
into existence as the germinal of a corpus of international maritime law.
But in the new context of international labour legislation by the ILO the
legal position of seafarers became connected to issues of forced plantation
labour and coerced conditions. Labour conditions of the seafarer became

70 See “E. Kupers.”.

71 Tideman, “De koelieordonnantie en hare toepassing,” 55; Middendorp, Twee achter-
lijke arbeidssystemen.
72 Middendorp, Twee achterlijke arbeidssystemen, 49.
73 Bosma, “European colonial soldiers,” 318.
84 ulbe bosma

a focal point of the ILO legislation, not only because these issues were
almost by definition global in character but also because of the existence
of penal sanctions in maritime law, which were considered to be detri-
mental to the fight against indentured conditions on plantations. Indeed,
advocates of the continuation of the penal sanction on the Sumatra plan-
tations used the fact that desertion of sailors and some categories of civil
servants at railways were still under the penal sanction to shore up their
cause. But the argument lost its final bits of credibility in the 1920s, when
the ILO developed international legislation to ensure honest recruiting
practices in the maritime sector and in 1926 to regulate the indentured
condition of the seafarer.74

Conclusion

In the nineteenth century labour contracts that comprised a penal sanc-

tion were not necessarily conceived as in contradiction with free labour,
as we would today, but rather as constrained freedom which in fact every
type of wage labour was. The Cultivation System in the Dutch East Indies
was implemented as an upshot of ideas about the edifying role of work that
had first been tested in the Netherlands. Van den Bosch had been a vision-
ary whose intention had been to give the Javanese peasant the fruits of
his work and land. That he wanted to make the Javanese rural population
work harder sounds colonial, but it is not necessarily so, and in his case it
was derived from the Netherlands context, where he had established beg-
gar colonies. It could hardly have come as a surprise that the introduction
of the Cultivation System in social conditions that Van den Bosch himself
had qualified as despotic, in practice aggravated the already precarious
position of the majority of Java’s rural population. The colonial govern-
ment not for nothing had to scale down the Cultivation System in the
1840s, which did however not diminish the inequalities at the village level.
Van den Bosch’s Cultivation System was not able to break through despo-
tism for the simple fact that it had to rely on it. The coalition between the
village elites and the sugar factories, which I focused on in this article, con-
tinued even into postcolonial times. Until this very day village patriarchy

74 ILO Convention No. 9, Placing of Seaman Convention 1920 and Convention No. 22,
Seaman’s Articles of Agreement Convention, 1926.
 dutch imperial anxieties 85

supply labour and moneylenders siphon off the resistance of the sawah
owners to lease their land out against sub-economic prices.75
It has been noted that the step from bondage to contract with penal
sanctions may constitute a feeble step towards freedom. For women for
example it could mean an escape from patriarchy.76 In Java large segments
of the population had to enter the wage labour market, simply because
their access to land was encroached upon by plantation enterprise. Free
labour for the Java sugar industry meant free availability of labour. As
soon as the sugar manufacturers were confronted with a tightening labour
market they asked for a penal sanction and they got one, although it was
repealed after an intervention by Dutch Parliament. At that time it was
felt that within the Dutch empire a penal sanction had no role in ordinary
labour relations. The penal sanction for Sumatra came via the backdoor
of regulating recruitment practices. The abolition of the penal sanction
that was in force outside Java was stalled on the basis of extraordinary cir-
cumstances at the Sumatra plantation belt and its paramount economic
importance for the Dutch empire. The repression of collective action in
the Netherlands Indies in 1923 was motivated on behalf of the same para-
mount interest of preserving the colonial order. The way in which the
colonial government of the Netherlands Indies invoked ‘extraordinary
circumstances’ to legitimize penal sanctions caramboled however via
the international arena to the Netherlands. The indentured conditions of
sailors on Dutch ships were softened in 1937, when with some delay the
relevant ILO Convention of 1926 came into force in the Netherlands.77

75 See Ingleson, “Life and Work”; Chandra, “The Role of Female Labor”; Mather,
“­Industrialization”.
76 Sen, “ ‘Without His Consent?’ ”, 77–104.
77 Schuman, Tussen vlag en voorschip, 203.
 Children and Forced Labour in the Indian Ocean World,
Circa 1750–1900

Gwyn Campbell

This paper examines the structure and significance of servile child labour
in the Indian Ocean world (IOW) in the period c.1750–1900. To do so, it
needs first to define the terms “IOW” and “child,” and delineate the main
features of traditional child servitude in the region defined.

The Indian Ocean World (IOW)

The IOW refers to a vast region running from eastern Africa (from the
Cape to Cairo) through the Middle East, South and Southeast Asia to the
Far East. This region was witness to the first “global economy,” defined as
a sophisticated, durable structure of long-distance exchange of commodi-
ties, ideas, technology and people―as distinct from the modern “inter-
national economy” which began to take shape in the nineteenth century.
The economic, social and political foundations of the IOW date back two
millennia and are related to the monsoons, a system of regularly alternat-
ing winds and currents unique to the Indian Ocean, and Indonesian, and
South and East China Seas. The monsoons exerted a huge influence over
the lands and societies around these inter-connected bodies of water:
Monsoon rains underpinned agricultural production, while monsoon
winds created the possibility of direct trans-oceanic sail which in turn
facilitated the rise of a structure of long-distance maritime exchange. The
IOW global economy, which linked the three major production centres
of China, South Asia and the Middle East to each other, and to other
IOW regions such as Indonesia and eastern Africa, started to develop
from about 300 BCE, had become a fully fledged exchange system by at
least 1000 CE, and was only seriously challenged by Europeans from the
late eighteenth century. Traditional IOW patterns of exchange continued
to be vibrant up to the immediate post-1945 era, since when they have
become steadily more attenuated.
88 gwyn campbell

Definitions of Children

The current definition of a child is generally someone under the age of

eighteen years.1 In the IOW, traditional definitions of children varied. In
much of the IOW, as in pre-modern Europe, children and females were
considered inferior to, and under the authority of, adult males. In China,
where a female could never carry on the ancestral line, an unmarried
girl was submissive to her father; a married female to her husband; and
a widow to her son. Females were never accorded full adult status.2 A
male was similarly submissive to his father up to the time of the latter’s
death. Only then did he assume full adult status.3 In China, as in nine-
teenth century Thailand and Madagascar, a husband could sell his wife,
or children.4 On the French Mascarenes in the eighteenth and early nine-
teenth century, child slaves were deemed to be those under the age of
fourteen.5 However, European enclaves in the IOW were rare before the
late eighteenth centuries, and as, historically, most societies indigenous to
the IOW did not issue birth certificates, it was often difficult to accurately
divine the age of a child. Thus in Egypt in 1834, a European doctor put the
age of one young boy slave at three, while two colleagues considered him
to be twice that age.6
Traditionally, servile adults were differentiated from servile children
by the onset of puberty for females and a combination of age and height
for males. For example, in Imperial Madagascar in the early 1830s, slave-
raiding armies executed male captives over a certain height (variously
estimated at between 107 cm and 122 cm) while retaining those under that
height―generally considered to have been ten years old and younger―
as “boy” slaves.7 A similar definition seems to have applied in the 1830s
by Egyptian slave-raiding armies who only enslaved captive females, and
boys.8 The reason older boys and adult males were killed was because
captors feared that they would resist, the costs of surveillance were much

1 Campbell, Miers and Miller, “Introduction” to idem (eds), Children in Slavery Through
the Ages, 3.
2 Hirata. “Free, Indentured, Enslaved;” Dull. Ed. Han Social Structure, 4.
3 Dull. Ed. Han Social Structure, 21–2.
4 Bowring, “Historical Forms of Bondage in Siam,” 286; Dull. Ed. Han Social Structure,
110; Campbell, “Slavery and Fanompoana,” 476.
5 Allen, “Children and European Slave Trading,” 38.
6 La Rue, “Brief Life of ‘Ali,” 73.
7 Freeman and Johns, A Narrative, 39–40; Griffiths, Hanes Madagascar, 62; Callet, His-
toire des Rois, 322; Ellis, History of Madagascar. vol. I, 138; Campbell, “Role of the London
Missionary Society,” 119; Hastie, “Journal” (1817–18), 184, 250–1.
8 La Rue, “Brief Life of ‘Ali,” 77.
 children and forced labour in the indian ocean world 89

higher, and they were less adaptable. Indeed, there is growing consensus
that in traditional IOW slave systems, the majority of humans trafficked
were young females and children.9

The Rise of the International Economy and Bonded Labour in the IOW,
circa 1750–1900

The period 1750 to 1900 witnessed the inter-related development of the

international economy, abolitionist movement, and modern colonial
order―events that shaped the changing nature of forced child labour in
the IOW. The international economy emanated from the Industrial Revo-
lution, associated technological innovations, and an international finan-
cial order centred on the City of London.10 It expanded rapidly, and by the
first decade of the twentieth century had drawn all bar the most remote
societies into its orbit.
In the capitalist heartland of Western Europe and North America, these
forces, occurring in a context of rapid demographic expansion, undermined
forms of bonded labour and promoted the rise of contract wage labour.
This process took longer than is traditionally thought―slave labour per-
sisted in the United States until the 1860s―but by the late nineteenth
century wage labour predominated. However, the structure of the inter-
national economy remained highly unbalanced, with on the one side an
industrialising core of largely western countries, and on the other the extra-
European world the economies of which remained largely pre-modern. In
the IOW, population growth rates were comparatively lower and uneven,
capital was comparatively more expensive, and more readily available
from Indian and other indigenous creditors than from Western sources,11
and attempts at industrialisation―with the exception of Japan―failed (as
in Madagascar)12 or remained small-scale with limited linkages.13 Hence,
the economy of the IOW remained overwhelmingly agricultural and arti-
sanal. Capital investment was directed predominantly into the cash crop
sector, mineral extraction, and exploitation of forest resources, as well as

9 See e.g. Campbell, Structure of Slavery.

10 Cain and Hopkins, British Imperialism: Innovation and Expansion, 1688–1914.
11 See e.g. Campbell, “Indians and Commerce in Madagascar.”
12 Campbell, “An Industrial Experiment.”
13 See e.g. Pomeranz, The Great Divergence; Tomlinson, Economy of Modern India;
Fahmy, All the Pasha’s men, 13–18, 42; Toledano, State and society in mid-nineteenth-century
Egypt.
90 gwyn campbell

into the transport and communications infrastructure required to facili-

tate exports to regional and Western markets.14
Given the limited nature of capital investment, such developments
greatly enlarged demand for labour, not only in extractive and produc-
tion processes, but also in transport (e.g. human porters; dockers; sailors).
However, major troop movements and armed conflicts (notably in Euro-
pean and indigenous “secondary” imperialism), increased trans-IOW com-
munications, and natural disasters, resulted in increased mortality.15 For
example, cholera, endemic in Bengal, was from 1820–22 carried by Brit-
ish troops and ships throughout the IOW.16 Cholera killed approximately
13 per cent of Cairo’s population in the 1831 outbreak,17 and almost 30 mil-
lion Indians in epidemics that swept Indian in the second half of the
century.18 The evidence from China illustrates what could happen when
man-made disasters coincided with natural catastrophes. One of the few
IOW regions initially characterised by very fast demographic growth,
China’s population leapt from approximately 300 million to 420 million
between 1800 and 1850. However, due a combination of exceptionally cold
weather in South China from 1876 to 1895,19 epidemics experienced in 36
of the 61 years from 1811 to 1872,20 almost constant warfare from 1850 to
1878, and famine, population growth stagnated and by 1900 the Chinese
population stood at only 450 million.21

The Abolitionist Movement in the IOW

As most IOW labour was tied in traditional servile relations to indig-

enous courts and elite households, there existed a very restricted free

14 Heydenrych, “Railway Development in Natal”; Maestri, “Naissance et premiers dével-

oppements d’un outil économique: le chemin de fer de la Réunion”; Porter, Victorian Ship-
ping; Cain and Hopkins, British Imperialism, 333; Kenwood and Lougheed, Growth of the
International Economy, 30.
15 For Southern and eastern Africa see Chrétien, “Demography and ecology in East
Africa”; Paillard, “Les recherches démographiques sur Madagascar”; See also Austen, Afri-
can Economic History, 67; Campbell, “State and Pre-colonial Demographic History,” 415–45;
Kjekshus, Ecology Control, esp. ch. 1; Campbell, “Disease, Cattle and Slaves.”
16 McNeill, Plagues and Peoples, 231–4.
17 Ibid., 231–4.
18 Chandra, “Colonial Legacy,” 11.
19 Landsberg, “Past Climates from Unexploited Written Sources,” 61–2.
20 McNeill, Plagues and Peoples, 268–9.
21 Ponting, Green History of the World, 241; Barraclough, Times Atlas of World History,
174, 232.
 children and forced labour in the indian ocean world 91

wage labour sector. The commercial boom associated with the expanding
international economy thus greatly stimulated the regional demand for
servile labour. In this context, abolition was viewed as a means of releas-
ing labour resources onto the free labour market where the forces of sup-
ply and demand would operate. However, the abolitionist movement was
spearheaded by western powers, notably Britain. It had an initially muted
influence in the IOW except in the Cape where, well before the 1834 aboli-
tion of slavery, the 1826 removal of measures protecting wine had diverted
investment into the far less labour intensive wool-producing sector―a
measure that effectively undermined slavery.22 Elsewhere, the labour
intensive nature of local economies and lack of free wage labour rendered
largely academic any arguments that slave labour might be inefficient.23
Until the abolition of slavery in European territory (in British colonies in
1833; French colonies in 1848; the Dutch East Indies in 1860), European-
held territory in the IOW initially depended largely on slaves to meet their
labour demands. For instance, Robert Farquhar, the first British governor
of Mauritius, delayed anti-slave import measures in acknowledgement of
the cheap labour requirements of local sugar planters.24 The British, in a
variant adopted by other colonising powers, declared newly-conquered
territories to be “protectorates” and thus avoided enforcing some aboli-
tionist measures compulsory in “colonies.” India, under East India Com-
pany rule until 1859, was considered a separate case.25 Overall, abolitionist
measures failed to significantly augment the free labour market in the
IOW where both European and indigenous authorities resisted complete
emancipation of slaves who, when genuinely freed, as on Mauritius from
1839, sought livelihoods independent of their former slave owners.26
In the late-nineteenth-century imperialist surge in the IOW, abolition
formed a central justification for the imposition of European colonial rule.
Moreover, colonial regimes, governed by precepts of self-financing, viewed
“liberated” slaves as a vital source of both taxation and manpower. How-
ever, another priority of colonial authorities was to retain the goodwill of
local slave-owning elites whose collaboration was required to administer
the colony. Thus, while moving quickly to hinder slave trading, colonial
authorities were reluctant to enforce rapid abolition lest it spark revolt, as

22 Worden, “Indian Ocean slavery and its demise.”

23 Clarence-Smith, Economics of the Indian Ocean Slave Trade, 4–5.
24 Anderson, “Bel Ombre Rebellion”; see also Oliver, “Sir Robert Townsend Farquhar.”
25 Chatterjee, “Abolition by Denial.”
26 Allen, Slaves, Freedmen, and Indentured Laborers, 105–35.
92 gwyn campbell

occurred in the Muslim province of the Southern Philippines after aboli-

tion was enforced in 1904.27 In Somalia, the colonial regime initially per-
mitted European settler farmers access to slave labour, and even returned
fugitive slaves to their owners,28 while in German East Africa, European
planters were permitted to “ransom” slaves who were obliged to work for
their “liberators” until the ransom had been paid off.29 A widespread vari-
ant of this was the practice of Christian missions in Africa of redeeming
slave children whom they subsequently employed working for them.30
Anti-slavery measures in European controlled territories occurred fit-
fully well into the twentieth century. In Africa, the internal slave traffic
remained buoyant for some 50 years after the banning of the external
slave trade. In the Sudan, effective measures to curtail slavery were taken
only in the late 1920s.31 On the eastern side of the IOW, the French first
seriously applied anti-slavery measures in Indochina in 1897, while the
British abolished slavery in Hulsawng valley in eastern Burma only in
1926. Slavery was outlawed in the Netherlands Indies in 1860, but the
Dutch then possessed only one quarter of the Indonesian territory that
was to pass under their control by 1910―in much of which they tolerated
slavery. Slavery endured in remoter regions of French Indochina and the
Dutch Indies into the 1940s.32 In the Middle East, drawn into the British
informal empire after the First World War, abolitionist pressure remained
muted until the post-1945 era.33

Types of Servile Labour in the Period 1750–1900

The failure to abolish traditional forms of servitude, and the search for
new forms of coerced labour, resulted in a particularly wide range of types

27 Clarence-Smith, “Islam and the abolition of the slave trade”; Salman, “meaning of
slavery.”
28 Eno, “abolition of slavery and the aftermath stigma,” 83–9.
29 Miers and Klein, Slavery in Colonial Africa, 6; see also Kopytoff and Miers, “African
‘Slavery’,” 73–4.
30 See e.g. Clarence-Smith, “Redemption of Child Slaves by Christian Missionaries.”
31 Miers, “Slavery and the Slave Trade in Saudi Arabia”; see also Kopytoff and Miers,
“African ‘Slavery,’ ” 72; Miers and Klein, Slavery in Colonial Africa, 1–2, 4–5.
32 Klein, “Emancipation of Slaves in the Indian Ocean”; Boomgaard, “Human Capital”;
Delaye, “Slavery and Colonial Representations in Indochina”; see also Reid, Slavery, Bond-
age and Dependency, 34; Klein, “Introduction: Modern European Expansion and Tradi-
tional Servitude in Africa and Asia,” 24.
33 Miers, “Slavery and the Slave Trade”; Mirzai, “The 1848 abolitionist farmān”; Klein,
“Emancipation of Slaves in the Indian Ocean”; Clarence-Smith, “Islam and the abolition
of the slave trade.”
 children and forced labour in the indian ocean world 93

of servile labour from 1750 to 1900. Traditional IOW systems of servile or

unfree labour that were perpetuated into the post 1750 period varied
from chattel slavery to the client slavery that characterised the prazos of
Mozambique, to pawnship, to agrestic servitude in India, to debt bondage.34
Forms of enslavement varied, but the five types dominated: enslavement
through warfare (captives); justice (criminal penalties); raids/kidnapping;
sale of dependants; and indebtedness.35
Traditional slavery continued to be important in the IOW. Moreover,
children were in greater demand than adults as slaves because they could
be more easily moulded, learn the host language, and imbibe local reli-
gious values and traditions. They also more easily forgot their own lan-
guage, customs and memory of homeland, and forged relationships of
total dependence upon their master.36 For this reason, children born
into slavery were highly valued,37 but demand was not met locally, partly
because of a generally low rate of slave reproduction, partly because of a
relatively elevated rate of manumission. Certainly a steady trickle of slaves
was assimilated into many African communities, depleting local slave
stocks and encouraging further slave imports. A similar process was evi-
dent in Muslim communities, where the sharia extolled manumission as
meritorious, stipulated that children born to her owner by a slave women
would be free, and that a concubine who bore a child to a free Muslim
would, upon his death, be manumitted.38 The rate of manumission could
theoretically be high; whereas a rich Muslim was legally restricted to four
wives, the number of concubines he might possess was unlimited.39
These factors ensured a high demand for imported slave children. Such
was demand that the IOW slave trade peaked in the nineteenth century
despite growing abolitionist scrutiny that induced slavers to adopt indi-
rect routes and pass slaves off as non-slave porters, sailors, domestics,
and even as children or other kin.40 Large numbers of East Africans were

34 For Africa see Isaacman, Mozambique: The Africanization of a European Institution,

The Zambesi Prazos; Kopytoff and Miers (eds.), Slavery in Africa; Lovejoy, Transformations
in Slavery; Meillassoux, Anthropology of Slavery; Manning, Slavery and African Life. For Asia
and the IOW see Campbell, Structure of Slavery; Reid, Slavery, Bondage and Dependency;
Watson, Asian and African Systems of Slavery; Prakash, Bonded Histories; Klein, Breaking
the Chains; Clarence-Smith, Economics of the Indian Ocean Slave Trade.
35 Lasker, Human Bondage in Southeast Asia, 16–17.
36 Law, “Slavery and Debt Bondage in Sarawak,” 291.
37 Manning, Slavery and African Life, 114.
38 Sheriff, “Slave trade and its fallout”; Clarence-Smith, “Islam and the abolition of the
slave trade.”
39 Miers, “Slavery and the Slave Trade.”
40 Campbell, “Introduction: abolition and its aftermath in the Indian Ocean world.”
94 gwyn campbell

shipped to Zanzibar, Pemba, Somalia, Madagascar, the Mascarenes, and

some to Cape Town. They were also exported to India and the Ameri-
cas. Malagasy slaves were sent in considerable numbers to Réunion and
Mauritius. Indian slaves were shipped to Indonesia, Mauritius, Cape Town
and the Middle East. Most slaves to the Middle East initially originated
from the Caucasus, Eastern Europe and Africa but were joined in the
nineteenth and early twentieth century by slaves from the Makran coast
of Iran, some from Western India and a few from Indonesia and China.41
Indonesians were dispatched mainly to Southeast Asia (but also to mar-
kets such as Mauritius), Indochinese and Koreans to China, and Chinese
to Singapore and San Francisco. In all of these trades, sources, markets,
routes, and slave functions, varied considerably.42
It is currently impossible to estimate with any precision the number
of slaves traded in the IOW given the duration of the slave trade there,
and the limited nature of extant records. Even in European-dominated
enclaves, statistics are at the very best, patchy. Research is only just begin-
ning into slave-ship journals and records in the region,43 but, in contrast
to the Atlantic system, IOW slaves rarely constituted a specialist cargo.
This was especially the case on indigenous ships, but also on many Euro-
pean-captained slaving vessels.44 Thus scholars’ attempts to divine quan-
tities of slaves traded in the IOW are at best “guesstimates” that inevitably
have considerable range. Moreover, they rarely take into account variance
in the number of slaves embarked and those reaching their destination:
Mortality rates aboard ship decreased significantly over the centuries as
slave traders paid increasing attention to improving slave hygiene, nutri-
tion and security aboard ship,45 so that historians have emphasized that
by the nineteenth century distance was the main factor influencing mor-
tality rates: Thus there appears to be consensus that the slave mortality

41 Miers, “Slavery and the slave trade in Saudi Arabia”; Sheriff, “slave trade and its fall-
out”; Klein, “emancipation of slaves in the Indian Ocean,” 198–218.
42 Allen, “Mascarene Slave-Trade”; Warren, “Structure of Slavery in the Sulu Zone”;
Delaye, “Slavery and Colonial Representations”; Alpers, “Flight to Freedom”; Schotten-
hammer, “Slaves and Forms of Slavery in Late Imperial China”; Machado, “Forgotten Cor-
ner of the Indian Ocean”; Boomgaard, “Human Capital”; Worden, “Indian Ocean slavery
and its demise,” 29–49; Campbell, “Unfree labour,” 66–82; Sheriff, “The slave trade and its
­fallout.”
43 See e.g., Ross, “Dutch on the Swahili Coast”; Allen, “Constant Demand of the French”;
Westra and Armstrong, Slave Trade with Madagascar.
44 Campbell, Economic History of Imperial Madagascar, 228; Allen, “Constant Demand,”
64–5.
45 Westra and Armstrong, Slave Trade with Madagascar, 13, 15, 25, 33, 35.
 children and forced labour in the indian ocean world 95

rate aboard ships to the Mascarenes from other regions was 20 to 25 per
cent (from India), 25 to 30 per cent (West Africa) and, in the late eigh-
teenth century, 12 per cent (Madagascar) and 21 per cent (East Africa).46
However, this takes little account of disease, which played a major role in
the nineteenth century, or mounting pressure by anti-slave trade patrols
in the closing decades of the nineteenth century which forced the cap-
tains of slave ships to take more risks, overcrowd their ships, and jettison
slaves should their ships be threatened with capture.
Because of the conventional focus on Black slaves, East Africa and Mad-
agascar are the only IOW regions with concentrated slave export guessti-
mates―respectively ranging from between 800,000 to over two million,47
and between about 72,000 to double that figure.48 In total terms, factoring
in other regions and non-African slaves, it is probable that the cumula-
tive number of slaves traded across the maritime space of the IOW over
the centuries well exceeded the 10 to 12 million landed in the Americas.
However, it is highly probable that the greatest IOW slave traffic was over-
land, notably within Africa, Hindu India and the Confucian Far East: in
1841, Bartle Frere estimated that there existed eight million to nine million
indigenous slaves in India alone49―double the number of black slaves in
the United States in 1865. This again speaks to the major difficulty faced
by Europeans, but largely ignored in the literature―the tying up of the
IOW workforce in indigenous systems of bonded labour.50
Again, while adult males were the most valued in some markets, such
as Cape Town in the late eighteenth century,51 and the plantations on
Zanzibar and the French Mascarenes,52 children continued to form a

46 Allen, “Mascarene Slave-Trade,” 39.

47 Austen, “The 19th Century Islamic Slave Trade from East Africa”; Sheriff, Slaves, Spices
& Ivory, 226; Campbell, Economic History of Imperial Madagascar, 238; Alpers, Ivory and
Slaves, 15 1, 185–7; Allen, “Mascarene Slave-Trade,” 38–9; Capela and Medeiros, O Tráfico de
Escravos, 24–5, 41; Liesegang, “First Look at the Import and Export Trade,” 463; Kjekshus,
Ecology Control, 14–16.
48 Filliot, La traite des esclaves, 157–9; Campbell, Economic History of Imperial Madagas-
car, 55–6, 238; Allen, “Mascarene Slave-Trade,” 35, 38.
49 Balfour, Cyclopædia of India, 674.
50 Watson, “Transactions in People,” 235.
51 In 1777 at Zanzibar, a Dutch slaving ship from Cape Town took aboard 328 slaves, 70
percent of whom were males, mostly adults, and 27 percent of whom were adult females―
Sheriff, “Localisation and Social Composition of the east African Slave Trade,” 138. See also
Ross, “The Last Years of the Slave Trade.”
52 This is reflected in the values placed on slaves in Mozambique coast markets where
slaves were split into three categories: the pesca, comprising healthy young male and
female adults aged between 18 and 25 years, followed by the pote d’agua—healthy slaves
96 gwyn campbell

majority of slaves exported from East Africa. For example, in 1856, of the
69 African slaves freed by the British from ships travelling to the Persian
Gulf, 29 percent were aged less than 10 years old, and 36 percent aged
between 10 and 19 years.53 None of the males was older than 25 years
(38 percent were aged under 10 years); whereas the average age for females
was 19 years—17 percent being over the age of 25.54
In Southeast Asia, children formed the major target of slave raiders and
kidnappers. Thus Hugh Low commented of Sarawak in the early nine-
teenth century:
The slaves (ulun-ulun) in Borneo are generally Dyaks and their descendants
who have been captured by the rulers of the country . . . it was no uncom-
mon thing for these tyrannical chiefs . . . to send up parties to the Dyaks to
bring down all the young girls and boys they could catch. I have been told
by Dyaks, and by dependants of the Sereib, that 300 girls and boys have fre-
quently been brought down at one time; such of these as the chief selected
as likely to suit his purposes were reserved, the remainder were sold to
whoever would buy them, the chief taking a considerable share, though his
servants who had kidnapped the poor children generally contrived to cheat
him of a considerable portion of the produce.55
“Primitive” hill peoples were particular targets of kidnappers who seized
and sold children, notably girls.56 Sometimes such kidnappings were so
extensive―as by the Rawa of Sumatra on the Mantra of Malacca―that
it forced the communities subject to the raids to abandon their ancestral
lands.57
Children who entered slavery were seldom sent directly to their final
market. For example, in interior East Africa, the passage to the coast of
those younger slaves destined for coastal and export markets was frequently
delayed, or made in stages, possibly in order that they learned to be doc-
ile, but also to ensure that remained healthy and thus retained their value.
During the often lengthy interval between enslavement and shipment,

aged between 14 and 18, and finally a group comprising less desirable slaves―Campbell,
“Madagascar and Mozambique in the Slave Trade of the Western Indian Ocean 1800–1861,”
178. See also Géraud, “Les esclaves à l’épreuve de l’industrie,” 297–8.
53 Sheriff, “Localisation and Social Composition of the east African Slave Trade,” 139.
54 Ibid., 139.
55 Law, “Slavery and Debt Bondage in Sarawak,” 289.
56 See e.g. the enslavement of Negrito children by urbanized Christian Filipinos; and
of hill girls captured by the Muslim Sulu―Lasker, Human Bondage in Southeast Asia, 37,
40–1.
57 Lasker, Human Bondage in Southeast Asia, 48.
 children and forced labour in the indian ocean world 97

they often changed hands a number of times, and were employed by their
owners in a variety of ways―as domestic servants; tenders of animals,
gardens, or even of children younger than themselves; and fetchers of fire-
wood and water.58 Similarly, the children kidnapped and sold by the Red
Karen in nineteenth-century Burma were frequently sold several times
before finally being retained by a master.59
Child slaves also constituted a commodity money, “small capital suited
for negotiating deals, settling debts, or establishing credit”60 in East Africa;
and in Central Africa were commonly used in payment for ivory and
transport.61 In much the same way, merchants visiting Mecca on the Haj,
frequently took slaves who they used as a kind of “travellers cheque,” sell-
ing upon arrival in order to finance their return trip.62
The traffic in children reflected both supply and demand factors. In
times of dearth, and notably famine, families throughout the IOW com-
monly sold their children in order to give them, and the remaining mem-
bers of the household, a better chance of survival.63 However, slavers
did not take very young children, who they considered a burden, likely
to slow down a caravan and prone to high mortality.64 On the demand
side, elite males purchased slaves, notably children and young women,
chiefly as items of conspicuous consumption to reflect their power, pres-
tige and wealth.65 Indeed, in some cases, the costs of maintaining a slave
exceeded the benefits accruing from his/her services which, in exceptional
circumstances, could lead to the financial ruin of the owner.66 Neverthe-
less, a false dichotomy has often been assumed between “productive” and
“unproductive” slave activities67 many of which had considerable eco-
nomic significance. Children and wives acquired for kinship groups were
both status symbols and important additions to the group’s productive

58 Morton, “Small Change,” 60–1.

59 Lasker, Human Bondage in Southeast Asia, 44.
60 Morton, “Small Change,” 55.
61 Renault, “Structures of the Slave Trade in Central Africa,” 155–6.
62 Ahmad, “Ethiopian Slave Exports at Matamma, Massawa and Tajura,” 99.
63 Lasker, Human Bondage in Southeast Asia, 53; Morton, “Small Change,” 65; Ferny-
hough, “Slavery and the Slave Trade in Southern Ethiopia,” 107; Deutsch, “Notes on the
Rise of Slavery,” 84.
64 Manning, Slavery and African Life, 114.
65 Ahmad, “Ethiopian Slave Exports,” 93; Boomgaard, “Human Capital”; Lasker, Human
Bondage in Southeast Asia, 49–50; see also Patnaik and Dingwaney, Chains of Servitude,”
2–4, 26; Reid, Slavery, Bondage and Dependency, 13; Goody, “Slavery in Time and Space,”
36–7; Klein, “Introduction,” 8–13.
66 Boomgaard, “Human Capital.”
67 See Kopytoff and Miers, “African ‘Slavery,’ ” 55–7, 64–6.
98 gwyn campbell

and reproductive capacity, while throughout the IOW slaves acquired

by elite households were often encouraged to engage part-time in profit
bearing activities.68
In the eighteenth and early nineteenth centuries, slaves comprised
between 20 and 30 percent of the population of many IOW societies, ris-
ing to 50 percent and over in parts of Africa and in Indonesian ports.69
Boys were traditionally valued chiefly as household servants, guards, and
soldiers. For example, one of the main destinies of Serb, Greek and Alba-
nian Christian boys aged 10 to 18 years absorbed into the devşirme system
that functioned in the Ottoman Empire until 1826 was the janissary corps
of the army.70 Slave soldiers enforced law and order―a pre-requisite for
economic growth―and safeguarded vital trade routes, supply centres
and markets. Most were “remunerated” with battle spoils or expected to
engage part-time in economic activities. Indeed, the capture and exchange
of slaves (chiefly women and children) was a principal objective of most
pre-colonial armies and navies. Slave armies thus often generated slaves,
sometimes―as in the Sulu case―on their own initiative, and for their
own as well as their owner’s material interests.71
Like slave girls, slave boys were also valued for sexual purposes by mas-
ters, but also―clandestinely―by the master’s wife.72 Sometimes, sexual
exploitation occurred haphazardly, at other times sexual services formed
an integral part of a boy slave’s functions. For example, in societies with
an established tradition of homosexuality, as in parts of the Ottoman
Empire, Pakistan, Muslim India, Afghanistan and Turkestan, there was
considerable demand for slave boys (including eunuchs) used as sexual
companions,73 prostitutes,74 and as bacchá―boys who, chosen for their
height, size and beauty, were trained to sing and dance for elite male
audiences amongst whom they were traded for sexual favours.75 In Persia,

68 See e.g. Reid, “Introduction,” 14.

69 Kim, “Nobi: A Korean System of Slavery”; Boomgaard, “Human Capital”; Kopytoff
and Miers, “African ‘Slavery,’ ” 60–1; Reid, “Introduction,” 12, 29; Campbell, “Slavery and
Fanompoana,” 474–5.
70 Yilmaz, “Becoming a Devşirme,” 121–8.
71 Warren, “Structure of Slavery in the Sulu Zone,” 111–28; Goody, “Slavery in Time and
Space,” 26–7.
72 Morton, “Small Change,” 65.
73 Murray, “Homosexuality among Slave Elites in Ottoman Turkey.”
74 Neill, Origins and Role of Same-Sex Relations, 169.
75 Mathee, Pursuit Of Pleasure, 49, 285.
 children and forced labour in the indian ocean world 99

most slave boy prostitutes and dancers were aged between 10 and 16 years,
and were of Circassian, Georgian and Armenian origin.76
The highest valued child slave was the eunuch, commonly sought by
sovereigns throughout the IOW from the Ottoman Empire to China, nota-
bly as personal bodyguards, and keepers of the royal harem. There were
different types of castration: In Korea, these were listed as “involuntary,”
“accidental,” “congenital” and “self castration”―meaning “voluntary.”77 In
the case of slaves, it was mostly “involuntary,” enforced on enslaved boys;
but some submitted voluntarily in an attempt to achieve the higher status
and corresponding privileges accorded to eunuchs.78 However, the risks
were high: It has been estimated, despite the care taken because of the
value of a eunuch, that in Korea between 20 to 50 percent 79 and in Africa
up to 90 percent80 of boys died whilst undergoing surgical castration. Of
those who survived, some of the “self-castrated” had only their testicles
removed, which meant that for some, sexual intercourse remained pos-
sible―there are, for example, cases of love affairs developing between
court eunuchs and ladies, and of highly-placed Korean eunuchs taking
wives and concubines.81
The commercial boom from the late eighteenth century led to rising
demand for male slaves for activities associated with the production, pack-
ing and transport of export commodities, such as pearls, dates, cotton,
wool and opium in the Persian Gulf,82 cloves, coconuts and grain on the
Swahili coast,83 sugar on Mauritius, and hides in imperial Madagascar,84
and the return traffic comprising staples such as cloth, iron goods, firearms,
gunpowder, and alcohol.85 Most of this demand was for adult slaves, but
there was also significant demand for boys. Thus Richard Allen ­considers
that even in the case of Mauritius, a plantation economy, the trade was as
much in boys as in young adult males.86

76 There is also reference to Russians―Mathee, Pursuit of Pleasure, 169, incl. fn. 139, 170.
77 Kim, “The Third Gender. Palace Eunuchs,” 138.
78 Ibid., 138.
79 Ibid., 147.
80 Lovejoy, Transformations in Slavery (Cambridge: Cambridge University Press, 2000), 35.
81 Kim, “Third gender,” 130.
82 Ricks, “Slaves and Slave Traders in the Persian Gulf,” 60, 65; Clarence-Smith, Islam
and the Abolition of Slavery, 10.
83 Cooper, Plantation Slavery, 3.
84 Campbell, Economic History of Imperial Madagascar, 166–88.
85 See e.g., for Madagascar―ibid., 249–59.
86 Allen, “Children and European Slave Trading,” 36.
100 gwyn campbell

Nevertheless, most of the children traded as slaves were girls who con-
sistently fetched higher prices than non-eunuch boy slaves. For exam-
ple, in the mid nineteenth century, the average price for a male slave at
Gondar, in Ethiopia, was from 15 to 18 thalers, and for females from 20
to 25 thalers (the price for an exceptionally beautiful girl could rise to
80 thalers); while the price of a eunuch was 80 to 120 thalers.87 The price
paid on the coast was invariably higher than at source—15 times more in
the case of female slaves; and on the Arabian coast doubled again (while
prices paid for beautiful Ethiopian girls rose even more dramatically in
Persian Gulf markets).88 Thus the kidnapping of young girls was common
across Ethiopia―although traffickers often used captive girls as concu-
bines during the trip to market, despite the fact that the loss of virginity
lowered a girl’s sale price.89
Traditionally, at destination, some girl slaves were employed as water
carriers, and in agriculture, textile production and mining,90 but most
were absorbed by wealthy households that employed them predomi-
nantly in personal domestic and sexual services (for masters and their
wives)91 and entertainment.92 Female slaves who were secondary wives,
concubines, entertainers and domestic servants of the wealthy, enjoyed
a lifestyle and―as Anthony Reid argues for Southeast Asia―a respect,
superior to that of female peasants.93 There are instances of concubines
in the Middle East sending for family members to join them, albeit as non-
slaves.94 Female slaves were also less likely to be sold.95

Indebtedness

Possibly the majority of people entering slavery in the IOW did so through
indebteedness. Impoverishment frequently forced families into debt
which was normally expressed in monetary terms, although it was often
incurred in non-cash forms such as food or tools. Commonly, household
heads attempted to meet debt payments through the sale or mortgage

87 Ahmad, “Ethiopian Slave Exports,” 96, 98.

88 Fernyhough, “Slavery and the Slave Trade in Southern Ethiopia,” 113.
89 Ibid., 107, 112.
90 Goody, “Slavery in Time and Space,” 21, 32.
91 See the extraordinary case cited by Anwar, “Attitudes towards Homosexuality,” 404.
92 See Miller, “A Theme in Variations,” 169–94.
93 Reid, “Introduction,” 25–6.
94 Miers, “Slavery and the slave trade in Saudi Arabia,” 120–36.
95 Reid, “Introduction,” 25–6.
 children and forced labour in the indian ocean world 101

of a household member, notably a girl, but if in dire straits, also boys

and wives. This was a common pattern across the IOW, from China96 to
East Africa.97 Moreover, in many IOW regions, enslavement was legally
enforced for debtors and their relatives. In Imperial Madagascar, for
instance, creditors could, through recourse to justice, enslave a debtor, his
wife and ­children.98 This was also the customary practice in Thailand and
Malaya.99 In addition, the punishment for certain crimes was exacted in
fines, which often led to indebtedness and subsequent enslavement.100 If
the debt was paid off, an enslaved debtor could regain non-slave status.
Here, slavery needs to be distinguished from debt bondage with which
however, it could overlap. Enslavement for indebtedness was involuntary,
whereas most people entered debt bondage voluntarily, as a credit secur-
ing strategy. Mortgaging a child, or wife, to raise a loan was common prac-
tice in the IOW from early times.101 Certainly by the nineteenth century,
debt bondage embraced a vast range of people in the IOW, from farmers
mortgaging future harvests and potential grooms borrowing a bride price,
to small traders living off credit from larger merchants, the ubiquitous
rural gambler of Southeast and East Asia and opium addicts in nine-
teenth-century China.102 During catastrophes, people often entered debt
bondage or slavery in return for subsistence as a survival strategy, either
voluntarily, as was the case of many dvija caste members in India, or pro-
pelled by their kin group.103 They did so whether they lived in regions of
relatively low population density, such as Cambodia, Laos and Indonesia,
the Middle East and Africa, or of relatively high population density, such

96 Hirata, “Free, Indentured, Enslaved,” 4; Dull. Ed. Han Social Structure, 110.
97 Morton, “Small Change,” 59.
98 Campbell, Economic History of Imperial Madagascar, 295–6.
99 Lasker, Human Bondage in Southeast Asia, 147, 150.
100 Reid, “Introduction,” 10.
101 Thus the Code of Hammurabi (1795–1750 BC) in Ancient Mesopotamia stated: ‘If
an obligation is outstanding against a man and he sells or gives into debt-service his wife,
his son or his daughter, they shall perform service in the house of their buyer or of the
one who holds them in debt-service for three years; their release shall be secured in the
fourth year”―cited in Harris, “Did Solon Abolish Debt-Bondage?” 418. Different provisions
applied to slaves: ‘If he should give a male or female slave into debt-service, the merchant
may extend the term (beyond three years), he may sell him; there are no grounds for a
claim.’—ibid.
102 Boomgaard, “Human Capital”; Delaye, “Slavery and Colonial Representations”;
Schottenhammer, “Slaves and Forms of Slavery in Late Imperial China”; see also Watson,
“Transactions in People,” 228–36.
103 Klein, “Introduction,” 11; Patnaik and Dingwaney, “Chains of Servitude,” 25–6.
102 gwyn campbell

as Bengal, Vietnam, Korea and South China.104 Moreover, most victims

appear to have been pushed into debt bondage as children―in early
twentieth century Thailand they were rarely aged over ten years.105
Those subject to debt bondage could outnumber slaves. For example,
they were possibly the most numerous social category in Majapahit, in
Java, while in central Thailand in the eighteenth and nineteenth cen-
turies, they formed up to 50 percent of the total population. Again, in
Burma in the mid nineteenth-century, when debt bondage was far more
widespread than hereditary slavery, the male head of an impoverished
household frequently sold his wife and children to meet tax impositions.106
The servitude to which those in debt bondage were subject was generally
taken as paying off interest on the loan they had contracted, to which was
added the cost of lodging, feeding and clothing the debtor. Consequently
the debt in most cases increased and servitude could become permanent,
or even hereditary―at which point there was little to distinguish debt
bondage from slavery.107 However, in nineteenth-century Thailand, if the
interest owed by a debt slave grew to a level exceeding the original loan,
the master deemed the slave a bad investment and customarily sold the
slave―albeit on unfavourable terms.108
Pawnship, common in Eastern Africa, was a variant of debt bondage
wherein a pawn, frequently a girl, was given as security for a loan. The
creditor did not have the right to sell the pawn, whose labour he could
use until the moment that the family paid off the debt. However, in cases
where an inordinate amount of time passed with little or no sign or repay-
ment, the creditor’s ownership of the pawn became established.109
In some groups, the sale of children became a traditional, accepted
form of generating revenue. This was the case amongst many kasama
(tenant farmers) in nineteenth century Philippines of whom John Fore-
man (1899) commented:
Under the pretext of guaranteeing a loan, parents readily sell their children
(male or female) into bondage; the child is handed over to work until the
loan is repaid, but as the day of restitution of the advance never arrives,

104 Campbell, “Introduction,” 15.

105 Lasker, Human Bondage in Southeast Asia, 151.
106 Ibid., 138.
107 Kim, “Nobi”; Reid, “Introduction,” 12; for debt bondage from another angle, see
Miller, “A Theme in Variations.”
108 Lasker, Human Bondage in Southeast Asia, 151.
109 Lovejoy, Transformations in Slavery (2000), 13–14; Deutsch, “Notes on the Rise of
Slavery & Social Change in Unyamwezi,” 89.
 children and forced labour in the indian ocean world 103

neither does the liberty of the youthful victim. Among themselves it is the
law, and is still a practiced custom, for the debts of the parents to pass on to
the children, and . . . debts are never repudiated by them.110
This appears to have also been the case amongst the Aetas, a non-Christian
Negrito group, traditionally hunter-gatherers, from the isolated mountains
of Luzon, who in the nineteenth century and into the early twentieth cen-
tury were “adopted” by Christian Filipinos in return for a small payment
to their parents or guardians. Filipino masters justified such purchases as
being a means of Christianizing pagans.111 This rationale was similar to
that expressed by Christian missionaries in Africa who redeemed slaves,112
and stemmed from the widespread belief in Enlightenment Europe that
children from “primitive” pagan societies required “domestication,” and
that slavery was the protected status best suited to that end. From the
Enlightenment era, religious leaders propounded parallel concerns for
the moral progress of “children” that was translated into the provision
of Christian ministry to “childlike” slaves. Slave-owners needed to disci-
pline their slave wards into accepting their “public” duties, but also had
a responsibility to save their souls and teach them the moral virtues of
adulthood in the personal realm of “marriage” and “family.” By this line
of reasoning, bondage could be viewed as a state wherein undeveloped
humans might receive the physical and moral discipline and training nec-
essary for the attainment of civilised virtues―which alone could justify
“freedom” in the sense of political participation or economic autonomy.113
European powers that established colonies in the IOW facilitated a
growth in indebtedness, through imposing monetary taxes, promoting
commercialisation, and enforcing credit contracts. At the same time,
colonial authorities both maintained tight budgetary regimes that avoided
funding public welfare programmes, and distinguished debt bondspeople
from “true” slaves, whose condition they attributed solely to violent cap-
ture. As a result, debt bondage and enslavement through debt expanded
considerably across the IOW in the 1800s.114 India is a prime example. In a
century characterised by rising taxation and years of famine, “freedom” for
members of the former slave outcastes, who had deliberately been kept

110 John Foreman (1899), cited in Lasker, Human Bondage in Southeast Asia, 131.
111 Lasker, Human Bondage in Southeast Asia, 38.
112 Clarence-Smith, “Redemption of Child Slaves,” 252–79.
113 Campbell, “Children and slavery in the new world,” 261–85; Ariès, Centuries of Child-
hood, 128–33.
114 Reid, “Introduction,” 11.
104 gwyn campbell

destitute and debarred from land ownership, translated into the liberty
to starve. Some adopted sharecropping, but with two-thirds of the crop
paid to the landlord, the risk of failure was high. In order to survive, many
entered debt bondage that was from 1859 reinforced by the Breach of Con-
tract Act.115 In some areas of India, members of the most depressed castes
formed the overwhelming bulk of those in debt bondage. The situation
closely resembled slavery in that bondage could be inherited, and the vast
majority of bonded people had their geographical mobility restricted.116
Nevertheless, colonial powers across the IOW considered debt bond-
age to be a benign form of private welfare, and generally condoned its
continuation well into the twentieth century.117 In some areas, such as
Thailand, Burma and Indochina, this encouraged a revival of covert slave
raiding.118 In Africa where debt bondage was represented by the “pawn-
ship” of a person, usually a young girl, to a creditor in return for a loan, the
system weakened only during the post Second World War boom.119 In all
cases, the debtor had a clear market value, expressed in more monetized
Asian economies in terms of cash, and in less monetized economies, as in
most of Africa, in terms of “human” capital.

Apprenticeship

In the hope of avoiding manpower shortages on plantations after abolition,

the British introduced an apprenticeship scheme whereby emancipated
slaves would be forced to continue work for their ex-masters for a limited
period. For example, when on 1 February 1835, slaves were accorded non-
slave status on Mauritius, only those aged under six years were granted
total freedom. The 66,613 former slaves above that age were obliged
to serve their former masters as “apprentices”—for four years for non-
­praedial and six years for praedial former slaves (in the event non-praedial
apprentices were liberated on 1 February 1839, and praedial apprentices

115 Patnaik and Dingwaney, Chains of Servitude,” 29–30.

116 Ibid., 30–1.
117 Chatterjee, “Abolition by Denial”; Klein, “emancipation of slaves in the Indian
Ocean”; Salman, “Meaning of Slavery.”
118 Delaye, “Slavery and Colonial Representations”; Turton, “Violent Capture of People,”
69–82; Klein, “Introduction: Modern European Expansion and Traditional Servitude in
Africa and Asia,” 23.
119 Lovejoy, Transformations in Slavery (2000), 13–14; Miers and Klein, Slavery in Colonial
Africa, 12.
 children and forced labour in the indian ocean world 105

two months later).120 Apprentices were obliged to work 45 hours a week

(Sundays and national holidays excepted), and could, in addition, take on
paid work for a maximum of three additional hours a day. They could not
be sold individually, but could be transferred along with the property on
which they worked.121 Many female child apprentices—some as young
as eleven—were hired out by their “owners” as prostitutes in Port Louis.122
Apprentices did all in their power to escape the system. Many appren-
ticed mothers resorted to theft and prostitution to avoid the necessity of
contracting their children as apprentices,123 while many others ran away:
an average of 4,300 refugee apprentices a year being captured by police.124

Indentureship

Another temporary solution to labour shortages was the placement of

“prize negroes,” “liberated” from slave ships, under contract to Euro-
pean settlers; those granted to Mauritian planters were “leased” to the
government for four days a year to perform public works.125 However,
both the Prize Negro and Apprenticeship systems were overshadowed by
the indenture system. Indentured labour was an old institution used in
European settlements, either formalised by contracts, or ad hoc, as with
Khoi and San boys captured by Dutch farmers in the Cape interior dur-
ing the eighteenth century. The captives were forced to work until the
age of 25, by which time they were often married, with sons who were
subject to similar obligations. Many parents refused to abandon their chil-
dren, and so remained tied to the farm for life. Following abolition of the
slave trade in 1807, the Caledon Code of 1809 formalised indenture in the
Cape through a “pass” system, which restricted San and Khoi to farms in
a system of “virtual slavery.” Thus, Nigel Worden argues, the first effective
formal abolition in the Cape was not official Abolition in 1834 but the
1828 repeal of the Caledon Code.126 Upon abolition ex-slaves in the Cape
(mostly Khoikhoi and San)127 and Mauritius (chiefly African, ­Malagasy,

120 Peerthum, “Le système d’apprentissage à L’île Maurice 1835–1839,” 285, 293.

121 Ibid., 285–6.
122 Ibid., 290.
123 Ibid., 289–90.
124 Ibid., 291.
125 Anderson, “Bel Ombre Rebellion.”
126 Worden, “Indian Ocean slavery and its demise.”
127 Lovejoy, Transformations in Slavery (2000), 239.
106 gwyn campbell

Malay, Indian and Creole)128 who received no financial assistance, were

declared “apprentices” and obliged to continue working for a fixed period
for their old masters. In the Cape, the system ended in 1838, but on Mauri-
tius it continued until mid-century. Many apprentices fell into debt bond-
age to their old employer or chose to continue working for him in order
to remain with their children.129
A new indenture system arose due to the failure of abolitionist mea-
sures to transform ex-slaves into pliant wage labourers. Generally involv-
ing five-year contracts, it channeled manpower resources to enterprises
both within and external to the IOW―such as the sugar plantations of the
Fiji Islands. The geographical range of indentureship within the IOW was
enormous, running from the rubber plantations of Sumatra and Malaya
from the 1860s,130 the tin mines and rubber plantations of Malaya, 131 the
tobacco and rubber plantations of the Dutch East Indies,132 to the Assam
tea gardens from 1859–77,133 to Burma (Myanmar), Ceylon (Sri Lanka),
Queensland, Australia, where over 60,000 Pacific Islanders served as con-
tract labour on the sugar plantations from 1863 to 1904;134 New Guinea;135
and the South African sugar plantations from 1860 and goldfields from
1903–8.136 The nominal duration of indenture contracts varied. At the out-
set in Natal, contracts were of 10 years duration; 137 the Assam tea gardens
(from 1859) of between 3 and 5 years.138 In Queensland in the 1890s, con-
tracts were mostly of less than a year’s duration,139 and by 1904 in New
Guinea, contracts were of only 12 months.140
The historiography has concentrated on recruitment for European
enterprises, particularly of Indian indentured labourers, some one million
of whom were employed in India by the close of the nineteenth century;
two million others were shipped to overseas plantations between 1834 and

128 Allen, Slaves, Freedmen, and Indentured Laborers, 43.

129 Worden, “Indian Ocean slavery and its demise”; see also Carter, Servants, Sidars and
Settlers.
130 Gordon, “Contract Labour in Rubber Plantations.”
131 Jain, “Tamilian Labour and Malayan Plantations,” 2364.
132 Das Gupta, “Structure of the Labour Market in Colonial India,” 1801.
133 Ibid., 1784–6.
134 Fitzpatrick, “Indentured Labour in Australia.”
135 Shlomowitz, “Mortality and Indentured Labour.”
136 Richardson, “Recruiting of Chinese Indentured Labour.”
137 Brain, “Indentured Indians.”
138 Das Gupta, “Structure of the Labour Market,” 1784.
139 Shlomowitz, “Markets for Indentured and Time-Expired Melanesian Labour,” 88.
140 Shlomowitz, “Mortality and Indentured Labour,” 73.
 children and forced labour in the indian ocean world 107

1920.141 However, labour from many other sources were involved, including
the Pacific Islands and China. Indeed, Chinese sources suggest that Chi-
nese emigrant indentured labour was far greater than previously thought.
Some 7.7 million people emigrated from South China in the period 1851
to 1901, of whom 86.8 percent travelled to Southeast Asia, and most of
this serviced indigenous IOW rather than European demands for labour.142
Recruitment, transport, and living and working conditions of indentured
labourers were often similar to those of slaves.143 However, recruitment
was overwhelmingly of young adult males. Thus in Queensland from
1884–1901, females comprised only 200 of the 3,867 Pacific Island inden-
tured workforce;144 while the last shipment of Chinese indentured labour
for the Witwatersrand goldfields in November 1906 comprise 2,129 adult
males, three women, and four children.145
Although regulations forbade the recruitment of children, it is clear that
both recruitment agents and the recruits commonly practiced deception
as to age, so that probably a significant proportion of the predominantly
young males who entered indentureship were under the age of eighteen.146
The same applied to the females who in some instances, as on Mauritius,147
formed a more significant percentage of indentured recruits. In the case
of the Assam tea plantations from the late 1850s, professional recruiters
(arkattis) working for licensed contractors based in Calcutta, recruited
and abducted many women and children from rural villages to serve as
indentured workers.148 Moreover, the ranks of the children were added to
between the time the destination was reached and the end of the period
of indenture. As children of indentured parents grew up, they too came to
form part of the Indian indentured labour force. Thus on Reynolds Broth-
ers’ sugar plantations, alongside Tongaat the largest in Natal in the 1890s, a

141 Chandra, “Colonial Legacy,” 11; see also Allen, “Mascarene Slave-Trade”; Patnaik and
Dingwaney, “Chains of Servitude,” 5–7, 27; Klein, “Introduction: Modern European Expan-
sion and Traditional Servitude in Africa and Asia,” 20–1.
142 My thanks to Jesse Sayles for this calculation based on information from McKeown,
“Global Chinese Migration, 1850–1940.” For the traditional view, see e.g. Blue, “Chinese
Emigration and the Deck Passenger Trade”; Campbell, Chinese Coolie Emigration.
143 See e.g. on Indian indentured labour, Pineo, Lured Away; Mookherrji, Indenture
System in Mauritius; Tinker, New System of Slavery; Carter, Servants, Sirdars and Settlers;
Gerbeau, “Engagees and coolies on Réunion Island”; Carter and Gerbeau, “Covert Slaves”;
Dasgupta, “Plantation labour.”
144 Shlomowitz, “Markets for Indentured and Time-Expired Melanesian Labour,” 88.
145 Richardson, “Recruiting of Chinese Indentured Labour,” 86.
146 Northrup, Indentured Labor, 57–8, 70.
147 Allen, Slaves, Freedmen, and Indentured Laborers, 58.
148 Das Gupta, “Structure of the Labour Market,” 1785.
108 gwyn campbell

highly exploitative labour regime was established in which work lasted 17

to 18 hours a day for which adult males were weekly given extra one-and-
a-half pounds of rice, but women and children only if they worked―thus
contributing to high infant mortality rates.149 In Natal, upon the comple-
tion of their indentures, Indian workers were encouraged to re-indenture
to the same or another employer, Indeed, in 1895, a law was passed, effec-
tive from 1901, imposing a tax of £3 on those time-expired workers who
did not sign up for a further period of at least two years; and in 1903 this
measure was extended to cover the children (boys over 16 and girls over
13 years of age) of formerly indentured men.150

Forced Labour Regimes

Penal Labour
Europeans also adopted penal labour. Tens of thousands of indigenous
convicts, predominantly male, were shipped to labour in European settle-
ments, as for example to most British colonies,151 from Goa to Mozam-
bique, and from Batavia to the Cape.152 This practice increased in the
transition years of abolition as slave labour progressively dwindled, penal
labour sometimes being leased out for use by European individuals, as on
Mauritius until 1851.153 Penal labour was also widely used by indigenous
authorities for the harshest types of work. For instance, the Merina regime
in Madagascar used convicts in road construction, mining and foundries.154
However, it is unknown whether, as in indentured labour schemes, signifi-
cant numbers of children were involved.

Corvée Labour
Western abolitionist pressure prompted indigenous regimes as far apart
as the Ottoman Empire, Thailand, Zanzibar, Imerina and Ethiopia to
make at least official proclamations against the slave trade and slavery.155

149 Halpern, “Solving the ‘Labour Problem’,” 33.

150 Brain, “Indentured Indians.”
151 Anderson, Legible Bodies.
152 Worden, “Indian Ocean slavery and its demise”; Shirodkar, “India and Mozam-
bique,” 51.
153 Anderson, “Bel Ombre Rebellion.”
154 Campbell, “Unfree labour.”
155 Delaye, “Slavery and Colonial Representations”; Klein, “Introduction,” 25; Miers,
“Britain and the Suppression of Slavery in Ethiopia.”
 children and forced labour in the indian ocean world 109

In Thailand such measures largely ended slavery by 1900.156 In general,

however, indigenous anti-slavery measures were rarely effective, or were
manipulated to divert slaves from private ownership into state labour
pools, as occurred in late eighteenth-century Korea,157 and in nineteenth-
century Thailand158 and Madagascar.159 In these countries and in Java160
most productive labour was performed not by slave but by the nomi-
nally “free” population subject to state corvées, while in many states, as
in Egypt under Muhammad Ali,161 and in Imperial Madagascar,162 indig-
enous regimes used the forced labour of nominally “free” subjects to drive
programmes of economic modernization. Often inextricably linked with
such programmes were attempts, from Korea, Thailand and Burma to
Iran, Zanzibar and Imerina, to create and economically exploit “secondary
empires.” Critical to their success were armies comprising mostly forced
labour. Indeed, in order to escape state corvée, which in Korea, Burma
and Thailand often claimed up to 50 per cent and in Imperial Madagascar
up to 100 per cent of the labour of “free” draftees, some slaves rejected
opportunities to gain “free” status, while some non-slaves voluntarily
entered slavery.163
In some cases, a significant number of children were drafted into forced
labour units. Because of their small size, children were particularly valued
in mines. Thus bonded children were used in Chinese mines from early
times until at least the mid-twentieth century;164 and in the mica mines of
India from the late nineteenth century.165 Again, with the opening up of
goldfields (mines and alluvial) in imperial Madagascar from 1883, Merina
authorities summoned child labour from mission chapels and schools―
where attendance was made compulsory and registers formed the basis
for fanompoana (unremunerated forced labour) quotas. Missionaries and

156 Klein, “emancipation of slaves in the Indian Ocean.”

157 Kim, “Nobi.”
158 Terwiel, “Bondage and Slavery”; Turton, “Thai institutions of slavery”; Feeny, “Decline
of Property Rights”; Feeny, “Demise of Corvée.”
159 Campbell, Economic History of Imperial Madagascar, 112–19, 131, 213–8.
160 Reid, “Decline of Slavery in Nineteenth-Century Indonesia”; Boomgaard, “Human
Capital.”
161 Toledano, State and society, 6–7; Batou, “Attitudes of State and Society towards
Industrialization in the Nineteenth-Century Third World,” 14; Fahmy, All the Pasha’s men,
10, 11, 86–93, 96.
162 Campbell, “Slavery and Fanompoana.”
163 Campbell, “Unfree labour”; Kim, “Nobi”; Reid, “Introduction,” 18–19.
164 Lasker, Human Bondage in Southeast Asia, 337 fn. 19.
165 Ibid., 357 fn. 5.
110 gwyn campbell

other state-church agents played their role in drawing up the quotas and,
increasingly, in overseeing the fanompoana units.166 For example, as early
as 1887, Norwegian Missionary Society churches in the northern Betsileo
regions of Mananadona and Fisakana were interrupted by the recruitment
of 400 gold fanompoana workers comprising both sexes. Child labour in
the gold fields became so widespread that in December 1888 the Mada-
gascar Times declared: “by forced labor, into which even evangelists and
school children are pressed, the Prime Minister is washing for gold on
his own account. It is said that more than a thousand laborers are daily
employed on this work.”167 Child exploitation in the gold fields intensi-
fied from the late 1880s as the threat of French intervention grew. In gen-
eral, the predominantly Protestant Merina officials obliged the pupils of
Roman Catholic schools to undertake the heaviest gold fanompoana, but
in 1889 the same London Missionary Society (LMS) schools in the Ambo-
hibeloma district, where military drill was instituted by the local mission-
ary, were being called on to supply “gold” labour, and that same year the
entire “free” population of Ambositra region, stretching from Behenjy to
Ambositra and excepting only Antankaratra, were summoned.168
Resistance to fanompoana was met by force. For example, in 1889 mem-
bers of the church and school at Ambanimaso were seized and militar-
ily escorted to the goldfields.169 From 1890, all Malagasy entering Indian
or European shops in Mahajanga were searched lest they be carriers of
stolen gold, and it was commented of the neighbouring “gold” town of
Maevatanana:
People working steal the gold, sell it, and, if caught, their heads are cut off
and stuck on poles. It is reported that there are dozens of heads of gold-
stealers and so-called robbers stuck up like this at Maevatanarivo [Maeva-
tanana] . . . Lately I have heard that a young girl of twelve or thirteen years,
along with a young man, were caught with a small quantity of gold on them
at Maeratanarivo [Maevatanana] and their heads were sawn off by a spear-
head, taking hours over it.170
Finally, colonial regimes, burdened by metropolitan exigencies that
insisted on self sufficiency, and local realities wherein there were short-
ages of labour, also fell back on forced labour regimes. This was, for exam-

166 Campbell, “Gold Mining,” 113.

167 The Madagascar Times (8 December 1888).
168 Campbell, “Gold Mining,” 113–4.
169 Ibid., 116.
170 Anon, “Gold in Madagascar.”
 children and forced labour in the indian ocean world 111

ple, evident from at least the 1880s in India where European collieries
established zamindary rights over villages, and thus over the forced labour
of villagers.171 Again, the French in Madagascar after 1895 quickly resorted
to a forced labour regime that resembled, and was justified in reference to,
the pre-colonial fanompoana practiced by the Merina.172

Sexual Slavery

In the nineteenth century, demand grew for girls and young women due
to the rise of modern form of sexual slavery in the IOW. This, in turn, was
due to rising gender imbalance in key locations as a result of the commer-
cial boom associated with the international economy, and military action
associated mainly with European imperialism (indigenous IOW army
camp followers included large numbers of wives and concubines). This was
reflected, for example, in the increase in the number of traditional female
dancer-singers in Egypt turning also to prostitution by the mid-nineteenth
century.173 On the other side of the Indian Ocean, the migration of mil-
lions of Chinese male labourers to IOW centres such as Singapore led to a
huge demand for females for sexual purposes that was met by a traffic in
mostly involuntary prostitutes, many of whom were girls engaged through
deceit, or sold by their parents.174 A variant system was that of the mui
tsai, in which in exchange for payment a girl was transferred from one
family to another, as an adoptee, nominally to become a domestic servant.
The mui tsai system, which continued well into the twentieth century,
involved very young girls―in Hong Kong in 1921, 69 per cent of mui tsai
were under the age of 14175―many of whom were by the late nineteenth
century used as prostitutes by their new owners.176

171 Das Gupta, “Structure of the Labour Market,” 1788.

172 Jennings, “Forced labour in Madagascar.”
173 Nieuwkerk, ‘A Trade like Any Other,’ 34–5.
174 Jaschok and Miers, “Women in the Chinese Patriarchal System,” 19–20; Warren,
“Chinese Prostitution in Singapore.” Between 1860 and 1880, most girls recruited for pros-
titution in America varied were aged between 16 and 25 years old: in 1870, almost 46 per-
cent were aged under 20. Many of the female children of such prostitutes were similarly
raised to be prostitutes by the brothel owners, raising the number of child prostitutes: of
the total number of Chinese prostitutes in San Francisco between 1860 and 1880, 5 percent
were aged 15 and under―Hirata, “Free, Indentured, Enslaved,” 21–2.
175 Poon, “The Well-Being of Purchased Female Domestic Servants (Mui Tsai),” 153.
176 Lasker, Human Bondage in Southeast Asia, 53–5.
112 gwyn campbell

Large concentrations of soldiers led to a similar demand for females to

provide sexual services,177 that in European colonies was often facilitated
by colonial officials who feared unrest should the sexual needs of soldiers
not be met. Indrani Chatterjee notes that British garrisons in India openly
employed “slave” prostitutes who in the 1860s were officially taxed―a
measure which, alongside fines, only served to increase the debt bondage
that more often than not had “enslaved” such girls.178 Again, in Natal in
the late nineteenth-century some white planters controlled male workers’
access to women―sometimes their wives and at other times prostitutes―
to reward or punish indentured Indian males.179

Conclusion

There is growing consensus that, unlike the trans-Atlantic slave trade, most
victims of human trafficking in the IOW were children. The traditional sys-
tems of bondage remained vital during the nineteenth-century when the
growth of a truly international economy created an unprecedented com-
mercial boom in the IOW that established a large and growing demand
for menial labour. However, there was difficulty meeting that demand, for
two main reasons: comparatively low and uneven demographic growth,
and the tying up of the potential workforce in largely indigenous systems
of bonded labour. Abolitionist pressure in the IOW was muted as the focus
of abolitionists was on the Atlantic slave system, and European officials in
the IOW were reluctant to attack a ubiquitous system that most regional
elites held dear. They therefore perpetuated some traditional systems of
bondage, and introduced new ones. It is clear that, while the historiogra-
phy has concentrated on the role of adults, children continued to play a
highly significant part in these new forms of bondage.

177 See e.g. the impact of French troops in the 1798–1801 French occupation of Egypt―
van Nieuwkerk, ‘A Trade like Any Other,’ 30–1.
178 Chatterjee, “Abolition by denial,” 150–68; see also Patnaik and Dingwaney, “Chains
of Servitude,” 32.
179 Halpern, “Solving the ‘Labour Problem’,” 34.
 Part two

Dependance and Servitude at work:

Local Customs and Global Dynamics
 Factors that Shaped the Organization of Labor
and the Labor Market in Tokugawa Japan:
Kyoto and Central Japan

Mary Louise Nagata

Introduction

The organization of labor and the labor market in Tokugawa Japan, 1600–
1868, was shaped in major ways by the political policies and laws designed
to establish and maintain political stability and social order, consequently
changing the social, political and economic framework of society. One con-
sequence of these policies was urbanization and the expansion of urban
markets with a population dependent upon the market to supply their
daily needs. At the same time, the land under cultivation was also expand-
ing, demanding expansion of the agricultural labor force. Labor migra-
tion from village to town and the expansion of the urban labor market
was one result. Rural handicraft production for urban and other markets
was another. Labor and the labor market in Tokugawa Japan was largely
shaped by these competing needs. The first part of this article discusses
these constraints and how they shaped labor in Tokugawa Japan.
The Japanese family and demographic system also interacted with
these needs to shape and be shaped by labor and the labor market. Japa-
nese family businesses tended to follow the stem, branch and sub branch
framework of the stem family which could link villages and cities in a
complex business network. Japanese family practice allowed for the estab-
lishment of non-kin as well as kin branches and used adoption, marriage,
and divorce to recruit individuals into or expel them from non-kin, dis-
tant kin or close kin relations. These practices also provided important
opportunities for the ambitious young person to prosper through his or
her ability and work. The second section discusses various labor relations
and the organization of labor in the family business as well as the role of
the family system that shaped and was shaped by the labor market.
Then the focus will narrow to the traditional capital city Kyoto and the
organization of labor and the urban labor market. Kyoto suffered from a
severe economic crisis partly due to the opening of treaty ports to inter-
national trade in 1857 with sharp price inflation in grain prices and raw
116 mary louise nagata

silk 1858–1863. The political crisis that ultimately overthrew the Tokugawa
regime also included random murders and violence in Kyoto and the city
was under martial law 1863–8. The final part of this article uses popula-
tion registers compiled by Kyoto neighborhoods to investigate the effect
of these political and economic crises upon the Kyoto labor market. The
results show that people reacted to the economic crisis by leaving the
city if they could or by not coming to the city in the first place. However,
after these initial actions, once the political crisis became severe, residents
stayed in place and quietly continued their lives seeming almost to ignore
the crisis. The final sections will address this analysis before the conclud-
ing discussion.

The Pax Tokugawa and the Organization of Labor

The Tokugawa regime gained official recognition from the Japanese impe-
rial court in 1603 culminating a process of reunification begun in the mid-
dle of the sixteenth century by the warlord Oda Nobunaga, continued by
Toyotomi Hideyoshi and then stabilized by Tokugawa Ieyasu. This process
included several policies and practices, also begun under Oda Nobunaga
around 1581 and more or less completed by Ieyasu’s grandson Tokugawa
Iemitsu in 1635, that established the framework for the new political,
social and economic order and shaped or constrained the organization of
labor and the labor market. Although any major reorganization of society
is bound to effect the organization of labor in innumerable ways, I limit
the discussion here to several major components of the peace that were
essential to shaping the labor market.
Before the Tokugawa regime, the Muromachi government, a hegemony
supported by a majority of regional warlord magnates, established control
of much, but not all, of the Japanese archipelago in the late fourteenth
century. By the middle of the fifteenth century this hegemony was fall-
ing apart and losing what nominal support or control of religious institu-
tions and other political rivals it had been able to claim. One part of the
decline of the Muromachi government included vassals and lieutenants
of individual warlords overthrowing them to take control of the territory
the warlords had claimed. Of course, if a lieutenant could overthrow his
overlord, he could also be overthrown by his subordinates or others. The
social order and public security broke down as local strong men or even
village bullies fought to claim as much territory as they could take and
control.
 factors that shaped the organization of labor 117

At the same time, villages, towns and some provinces organized for
mutual self defense and local autonomy in egalitarian, or semi-egalitarian
leagues called ikki. The process of reunification begun by Oda Nobunaga
entailed bringing all of the warlords under the authority of one overlord—
Nobunaga, later Hideyoshi and ultimately Tokugawa Ieyasu—by conquest
or alliance. Policies designed to guarantee warlords security in their status
were critical for establishing the framework of the new social order.1
Two policies originally begun by Oda Nobunaga in his own domains
and later extended nationally by Toyotomi Hideyoshi were cadastral sur-
veys and the separation of warriors from the land. With these two linked
policies, men who considered themselves warriors and vassals of some
daimyo warlord lost their claim to land and moved into castle towns or
cities. At the same time, all agricultural land was registered with the cul-
tivators of that land and attached to rural villages. Villages gained confir-
mation of local autonomy and independence from warrior interference in
exchange for paying a harvest tax on rice based on the annual estimated
potential for producing rice on the land attached to the village. Daimyo
were granted domains defined by the amount of harvest tax they could
collect from the territory under their administrative control and were to
use this income to pay stipends to their vassals, who were prohibited from
most commercial activity, and fund any other duties or services necessary
for the administration of their domains.2 Separation of warriors from the
land they had controlled and forcing them to live on stipends from their
overlords in town separated hope-to-become warlords from the resources
to overthrow their overlords, providing political stability and security to
the overlords.
Separating warriors from the land also resulted in a rapid urbanization.
Since warriors, clergy and the nobility were also restricted from most com-
mercial activity, this sudden urbanization also demanded a rapid expan-
sion of production for urban markets and increasing demands for labor
There were 260 domains during the Tokugawa period at any given time,
each with a castle town. By 1650 there were 250 castle towns of various
sizes with 25 of them cities with populations of 30,000 or more.3 In addition
to the castle towns were three metropolises. Edo was the Tokugawa politi-
cal capital. All of the daimyo were required to maintain residences in Edo

1 See for example Souyri, The World, 121–135.

2 Wakita Osamu, “The social”.
3 Nakai Nobuhiko, “Commercial change”.
118 mary louise nagata

where their wives and children were required to live as hostages while the
daimyo were required to alternate attendance in the shogun’s court with
time back in their domains.4 Thus Edo grew rapidly to a major metropolis
just from the residences, family, vassals and servants of the daimyo, not
to mention the merchants, artisans and entertainers necessary to supply
their needs. The population of Edo reached one million by 1700. Osaka
was the financial center where a portion of the tax rice from each domain
was stored or converted to cash. As the financial center, Osaka was also a
major commercial center with many large merchant houses and a center
for the cotton textile industry. Osaka’s population was stable at around
600,000 by the eighteenth century. The third metropolis of Tokugawa
society was Kyoto, the traditional capital and home to the emperor as
well as a center of handicraft industry and silk textile production with a
population of around 200,000. The warrior class alone comprised about
eight percent of the population at that time. When the populations of
the castle towns are added to the three metropolises—Edo, Osaka, and
Kyoto—and other commercial cities in Central Japan, the urban popula-
tion in 1650 was around fifteen percent of the total population, estimated
at 12.3 million and approached twenty percent of the population in the
eighteenth century.5
Before the early seventeenth century labor was usually supplied from
within the household, defined broadly to include individuals or family
units that were dependent upon the main household, or supplied from
casual labor. The dependents included several groups. Some were working
to pay off a debt. Others had been sold or even sold themselves into ser-
vice to gain the basic support of room, board and clothing. In these cases,
they frequently were rewarded with parcels of land from the main house-
hold after ten to fifteen years in service, but continued to barter labor for
other needs. Yet others joined the main, or master, household during the
period of political disorder and war in the fifteenth and sixteenth centu-
ries. While many of these dependent individuals and households were not
blood kin of the master household, others were kin or became kin through
intermarriage with the master’s family or through adoption. In the case
of marriage, girls inherited their mother’s status and boys inherited their
father’s status. These dependent relations could continue across genera-
tions or be limited in time. Those who remained as dependents across

4 Hall, “The bakuhan system,” 158.

5 Shinbo Hiroshi and Hasegawa Akira, “Shōhin seisan”.
 factors that shaped the organization of labor 119

generations, called nago, hikan, or fudai in various regions of Japan, could

be regarded as hereditary servants or slaves.6
The ability to sell oneself or a family member into service was consid-
ered an important survival mechanism and potential opportunity for the
poor or others of low status that could also provide training as well as
protection. For example, the new daimyo warlords created by the Toku-
gawa regime rewarded with domains from the territory held by warlords
defeated in the Battle of Sekigahara in 1600 were fudai from within the
Tokugawa house. At the same time, this system could certainly be harsh
and was open to abuse. One long time problem was bandit gangs who
kidnapped people off the road and sold them into service. Political and
legal authority was so fragmented and limited during the fourteenth to
sixteenth centuries, even during periods of relative strength and stability,
that there seemed little hope for controlling the practice.7
The cadastral surveys and political stability changed these relations as
dependent households gained independent claim to the land they farmed.
This began the process of emancipation as well as breaking up overly
large holdings and denying absentee landlords—warriors, clergy or reli-
gious institutions, or nobility—ownership or control of the land. The new
regime was also much stronger and claimed far more legitimate central
authority than the Muromachi regime. So, even though the political order
remained relatively decentralized, it was far stronger and more central-
ized than previously. As political order was established, new laws in 1616
and 1619 prohibited the permanent sale of persons. Although individu-
als selling themselves or members of their households into service were
not considered a major problem, bandits had kidnapped members of the
nobility and others of high status to sell into service as well. According
to the new laws, both the seller and the buyer would be imprisoned and
fined the amount of the sale thus removing the profit of the sale from the
seller and doubly fining the buyer, whereas the putative slave went free.
In some cases the seller was executed.
These laws were the last of a series of steps taken to end the practice
of bandits kidnapping people and selling them, and ultimately the mar-
ket for persons.8 The sale of persons devolved into two types of service

6 Nelson, “Slavery in Medieval Japan”. See also Farris, Japan’s Medieval Population,
79–83, 235–45, 247–52.
7 Nelson, “Slavery”. See also Karen Brazell (trans.), The Confessions and Ishii Ryōsuke,
Shinpen Edo, 106–7.
8 Ishii Ryōsuke, Shinpen Edo, 106–7.
120 mary louise nagata

agreements. One was called “pawn service”9 in which a head of household

borrowed money on the collateral of the labor of a member of his house-
hold. Sometimes the labor of the servant was considered the interest for
the debt, but the servant often received a wage for his service after the
debt was repaid. Since the 1619 law prohibited the permanent sale of per-
sons, there were also time limits set for how long the servant could remain
in service with maximum periods of up to ten years depending on the
region. Ten years was the maximum allowed in the domains of the Toku-
gawa House and most other domains followed suit.10 When a family that
still owned some slaves went into debt, they typically pawned their slaves,
who then usually went free at the end of the service contract. These and
other opportunities for wage labor and commerce led to the near total dis-
appearance of “hereditary servants” as slaves by the eighteenth century in
Central Japan, where two of the metropolises were located that provided
many economic opportunities, and in all but the most isolated regions by
the early nineteenth century.
The second type of service agreement was understood as a temporary
sale and this developed into contract wage employment. As a temporary
sale wages were originally paid in advance to the family or owner of the
servant. Labor had originally been supplied from family and slave labor
within the household and a network of personal connections such as the
children of kin, friends and neighbors, but rapid urbanization and expan-
sion of the market demanded more labor than available through per-
sonal connections. This increased the demand for contracts beyond those
offered for debt service. Since this labor force was also supported within
the household of the employer, the new conditions left the system open
to abuse. Employers in need of workers would pay employment agents a
commission to introduce prospective employees for a five day trial period
and then negotiate a contract. The agent would receive the wages, paid in
advance, and remit them to the family of the worker less another commis-
sion. Problems occurred as the agents persuaded the worker to abscond,
stealing whatever s/he could, and introduced the worker to a new employer
to go through the process again. With the growing demand for labor, this
scam could be quite lucrative for the agents. Efforts to control these abuses
resulted in the development of written labor contracts and wages paid, or
credited, regularly to the worker rather than paid in advance to his or

9 質入れ奉公.
10 Ishii, Shinpen Edo 108–9.
 factors that shaped the organization of labor 121

her family by the employment agent. The state also organized the agents
into self-policing and licensing organizations and set a flat rate paid once
by the worker, so the agent could not profit so easily from this scam.11
Although wages came to be paid directly to the worker over the course of
employment, in reality they were credited to the worker’s account with
the balance paid in cash when the worker left service.12
The political, legal and economic relations between village and domain
shaped labor in many other ways as well. With 260–70 semiautonomous
domains under the Tokugawa regime, and because local autonomy meant
that each village was free to develop its own accommodation to the new
environment of political stability and growing commerce, there is not
room here to address all of the variations in detail, nor has any study been
able to accomplish this enormous task. Here I will simply note several
important trends and highlight some of the ways villages developed new
organizations of labor for market production.
In general terms, daimyo income and domain administration depended
upon the harvest tax on the estimated rice crop. This originally resulted
in domain demands and restrictions to maximize the land cultivating rice
and the agricultural labor force. On the other hand, the tax was commonly
paid in some combination of cash and kind, usually rice which also served
as another form of cash. While the rice was expected to come from the
harvest, the cash could be earned from other sources. When weather or
natural disaster caused the harvest to fall short of the estimated amounts,
cash from other sources became even more important so peasants could
buy the rice necessary for that part of the tax.13
The need for cash income from other sources encouraged the develop-
ment of commercial ties to the market. For many, this meant labor migra-
tion to other villages, the castle town, or opportunities outside the domain.
The distance and direction could be constrained by domain policies trying
to keep agricultural labor local, but long lists of people leaving villages
without reporting where they were going where domain restrictions were
strong demonstrate that domain policies could constrain, but not prevent
labor migration outside the domain.14 Who tended to leave the village

11 Ishii, Shinpen Edo, 91–105.

12 Nagata, Labor Contracts, 85–93.
13 The rice tax and adjustments to the system have been studied and discussed in detail
by other scholars. Our focus is on how the system constrained the organization of labor
and the labor market. For more information on the rice tax kokudaka system, see Chie
Nakane, “Tokugawa Society,” or Smith, Native Sources, 50–70.
14 Nagata, “Leaving the Village”.
122 mary louise nagata

for labor migration depended upon the family system and regional differ-
ences in access to national commercial networks. Analysis of population
registers in Central Japan has shown a life course pattern of labor migra-
tion during a period of life-cycle service during the ages 11–25 followed by
marriage. Population registers in Northeastern Japan have shown a longer
period of labor migration after marriage during the ages 20–60.15
One reason for these differences was the reliance upon rice agriculture
for the economic base. Northeastern Japan did not have a climate suit-
able to the strain of rice common at the time, so there was more trouble
and poverty from harvest failures. Southwestern Japan, on the other hand,
did so well with rice that there was less incentive for investment in other
market production. The region was also less urbanized than central Japan
resulting in yet other patterns of labor organization and migration, but lit-
tle has been published on this topic as yet. Domain authorities sometimes
encouraged local industries and commercial production to provide wage
labor and economic opportunities within the domain to encourage the
agricultural labor force to remain closer to home. This was often with the
understanding that labor migration would take place during the agricul-
tural off season of the winter months, as was often the case for sake and
soy sauce breweries.16 Breweries also employed a large casual labor force
that worked by the day, or even the half day, providing more flexibility
for labor amid various constraints as commerce, industry and agriculture
competed for labor force.17
Yet another solution was for villages or wealthy peasants within villages
to invest in production for the market. This could include cash crops,
including industrial crops such as paper mulberry, cotton or lacquer
trees, or producing primary products such as raw silk from raising silk
worms or dye cakes from safflower or indigo. Some villages also invested
in textile manufacture with younger girls spinning or winding the cotton
thread or silk floss, older women weaving the cloth and men dyeing the
cloth or the thread in proto-industrial organization of household labor
by-employments. In other villages, one or more of the wealthier house-
holds invested in the production of paper, sake, soy sauce or other prod-
uct employing men and women from the village and surrounding villages.
The organization of production labor including the use of casual labor in a

15 Nagata, “Labor Migration”, 60–84.

16 Nagata, Labor Contracts, 81–82. See also Roberts, Mercantilism.
17 Nakayama Shōtarō, “Shōyu jōzō”, 175–98. See also Yunoki Manabu, Sake zukuri and
Nagata, Labor Contracts, 48–50.
 factors that shaped the organization of labor 123

manufactory, or using multiple households as household by-employment

depended upon the production needs of the product.18 Ironically, rural
investment in manufacturing as by-employments or using casual labor
was so successful that its success delayed or prevented the formation of an
urban proletariat and wage labor, whether contract or casual, was closely
tied to agricultural income.19
Larger and more successful of such businesses, some of which are still
in operation today, expanded in a network following the stem and branch
organization of the Japanese stem family to have retail and manufactur-
ing branches of the business in the major cities of Kyoto, Osaka and Edo
as well as shipping their products to other markets through the national
commercial networks. The next section addresses the family system and
lineage network as constraint and factor shaping the organization of labor
and the labor market beyond the local community.

Family As Business and the Organization of Labor

Labor in Tokugawa Japan fell into five broad categories:

1. Casual labor—also known as daily wage labor hired without contract

by the half day, the day or longer periods. For example, the sake brew-
ing industry employed 100-day laborers. Casual labor was employed by
all sectors of the economy as it made use of the free time from agricul-
tural labor, either to work as a farm hand for someone else, providing
transport or in market production.
2. Live-in contract labor—servants, farm hands or business employees
living in the employer’s household. The age period and character of
the labor force for this type of arrangement varied between northeast-
ern and central Japan. As business employees they were employed as
apprentices, clerks and managers.
3. Commuting contract labor—management level employees in com-
mercial enterprises, semi-independent artisans without independent
business licenses, some rural village labor. Live-in business employees
moved out and continued working as commuters when they reached

18 Nagata, Labor Contracts. This book is a study of labor contracts and labor relations
from rural businesses taking each of these forms as well as urban businesses from Kyoto.
See also Pratt, Japan’s Protoindustrial Elite.
19 Osamu Saito, “Wages”.
124 mary louise nagata

management or were ready to marry and establish their own house-

holds. From the late eighteenth century, villagers who lived within
commuting distance of employment opportunities tended to commute
instead of living in. This became more common as more labor oppor-
tunities, both agricultural and manufacturing, became available in the
countryside as well as in the city.
4. Pawn service—domestic service and farm labor. These contracts rarely
appear where commercial opportunities were common. Most of the
pawn service agreements have been found in agricultural villages.
5. Hereditary servants/(former?) slaves—could no longer be sold and
often had opportunities for freedom or independence. These could
become integrated as kin or almost kin members of the master’s house-
hold in the business context, but were likely to continue in dependent
relations in isolated villages. Where the employer/master owned a fam-
ily business, they became part of the elite management of the business
because they were considered more reliable.

Labor in Japan was not the purview of any guild-like organization control-
ling apprenticeship or the technical standards of any industry. Individual
employers trained their own apprentices and relied on the loyalty of the
live-in contract employees trained in the business to guard their techni-
cal production secrets as best they could. Businesses competed with each
other on quality and innovation. There were no particular rules for the
above labor relations, but employers were generally expected to provide
room, food, clothing and basic medical care to the live-in employees and
food to the daily wage laborers and the commuters. Study of labor con-
tracts reveals no obvious difference between proto-industrial handicraft,
manufactures, or agriculture in the contract conditions, likely because the
purpose of the contracts was to prevent conflict and provide information
on who would settle any conflict between employer and employee.20
Pawn service contracts tended to disappear where there were other
commercial opportunities. For example, records from the Sugiyama fam-
ily document collection of Tondabayashi village south of Osaka include
some pawn service contracts from the early eighteenth century, the most
recent dated 1738.21 Other records, however, demonstrate how a poor

20 Nagata, Labor Contracts.

21 Osakaya Sanzaemon and Suzuya Shōbei, [to Sugiyama Zenzaemon], “Uke tori
shōmon no koto,” Document of receipt, 11/1738, Sugiyama collection no. 62B, Kyoto Uni-
versity Museum, Kyoto. Nagata, Labor Contracts, 86.
 factors that shaped the organization of labor 125

t­ enant farmer could borrow money to open a shop in town and pay off
his debt to his landlord and employer by working in the landlord’s sake
brewery.22 Business records rarely include pawn service contracts and
most of the pawn service contracts have been for domestic service or agri-
cultural labor, so I exclude this form from consideration in this part of the
study. Evidence from northeastern Japan where pawn service remained
relatively common into the nineteenth century suggests that a farmer was
more likely to pawn his own labor rather than that of another member of
his household. One possible reason was that the servant received a wage
independent of the debt transaction. A farmer knew he would receive that
wage if he pawned his own labor and left his family to figure out how
to repay the loan, whereas another family member might take the wage
and leave.23 In addition, although children do occasionally appear in the
data as servants at ages younger than ten, the occurrence was rare. The
probability that a servant would be less than ten years of age was only
five per thousand in late Tokugawa Kyoto.24 This is not to say that child
labor was not used, but children were not a substantial portion of the
labor force. Adoption was the main means of transferring guardianship
and any potential exploitation of a child from one family or individual to
another. Yet, adopted children under the age of ten were also rare, or at
least rarely identified as such.25
Tokugawa enterprises formed on the stem and branch framework of
the Japanese stem family. In this framework, one married child remained
home with the parents and, in the case of a business, apprenticed to the
parents and succeeded to management as head of the family business.
The other children all left to manage another branch of the business if it
were large enough, marry or otherwise join another family, or establish
their own alternative occupations. A small business was only one house-
hold. A large business was formed by multiple households built on this
framework and the organization of labor in these businesses was con-
strained by this family framework. Even smaller businesses might have

22 Haya et al., [to Sugiyama Chōzaemon], “Issatsu no koto,” Letter of request, 8/1776,
Sugiyama collection, no. S97, Kyoto University Museum Archives, Kyoto. Nagata, Labor
Contracts, 66–7.
23 Nagata, “Leaving the Village”.
24 Analysis of the Kyoto population registers explained and used in the next two sec-
tions of this article reveal twenty-eight servants listed under the age of ten: 1 age two, 1 age
five, 2 age six, 2 age seven, 4 age eight, and 18 age nine out of a total 4,794 observations for
servants for a probability of 0.0058.
25 Nagata, “L’adoption à Kyôto” .
126 mary louise nagata

one collateral branch managed by a younger son, the retired head or some
other member of the family. I begin with the organization of labor in a
small business and expand the explanation to larger and more complex
businesses with multiple branches.
Most business households were comprised of family members, primarily
nuclear or stem kin, and live-in employees of both sexes. These employees
appear in four categories: apprentices, skilled or management workers,
maids and (rarely) other domestic servants. I use the term apprenticeship
loosely. There were no official rules for apprenticeship-journeymen as a
system. A boy or girl entered employment at ages 10–14 with the under-
standing that they would learn to read, write, calculate, and the basic
skills of that business or industry. Their training at this point from early
teens until around 17–20 was under the supervision of the wife of the head
of household and business and assisting her was part of their duties. Labor
turnover tended to be quite high during this period, although the details
are still under investigation.
Apprentices who finished their apprenticeship without being let go
generally moved on to more intense technical training becoming skilled
workers, who would move out after a few years to become commuters.
These workers also tended to change their personal names at this stage
of employment, sometimes receiving a name from the employer, and
were socially identified as members of the employer’s house. Many young
women also received new names while in service.26 Although the major-
ity of apprentices following this career path were young men beginning
their apprenticeship at around the age of twelve, records from the Kyoto
silk textile industry show young women following this career path too, at
least to skilled artisan if not management. Clerks, management employees
and skilled male workers were called tedai. Although women were listed
as student/apprentices, they do not appear listed as tedai, suggesting they
could become skilled weavers or artisans, but not management employ-
ees. If the business involved production that could use unskilled or semi-
skilled casual labor, as with brewing or dyeing, then there would be meals
provided with room on the floor for the day laborers to sleep, often in the
manufactory or the attic. This casual labor was employed in addition to
the live-in management and domestic labor.
Nearly everyone married in the family and demographic system of
Tokugawa Japan, usually before the age of 30 for men and around age 23

26 Nagata, “Names”.
 factors that shaped the organization of labor 127

for women in central Japan. Live-in servants were expected to be single,

so the live-in period of employment usually ended for men and women
around the time they married. The transition from a live-in skilled or
management employee to a commuter who lived out often accompanied
marriage. In the extant population registers of Kyoto, for instance, adult
women born outside of the city were either in service or married. Women
only appear living independently and single if they were natives to the
city.27 The reasons for this are not clear, but the analyses in the final sec-
tion of this paper suggest that the economic and political climate of the
final decades of the Tokugawa period in Kyoto were not conducive to
young women living alone. Unfortunately, there is no usable population
register data for the other cities or for Kyoto before 1843 that recorded age
and birth province, so I cannot say if this was a general trend or the result
of that particular economic and political context.
Larger businesses expanded by establishing a branch household of fam-
ily and business, a regular mechanism of the Japanese stem family system.
Kyoto businesses often had a collateral branch where younger siblings of
the heir to the business gained training or where the former head of a
business could retire to. Skilled or management employees would be sent
from the stem household of the business and apprentices might circulate
between the branches. Larger businesses also often subsidized skilled or
management employees to establish their branches of the business in the
same way, and the business could continue expanding with both kin and
non-kin branches and sub-branches.
For example, the Konishi Breweries had their main or stem business
and brewery in Itami, a rural village outside of Osaka, from the mid six-
teenth century. They established a kin branch in Osaka in the seventeenth
century, a non-kin branch in Edo in the early eighteenth century and a
non-kin sub-branch of the Edo branch also in Edo in the nineteenth cen-
tury. Apprentices were recruited by all branches for training in the stem
business household and were sent to the various branches from the stem
household after completing their apprenticeships.28
These larger businesses were typically controlled by a senior manage-
ment council made up of the head of each branch, the head couple of
the stem business and the retired head of the stem business. Clearly this
council included non-kin employees. At least one larger business that

27 Nagata and Hamano, “Marriage Market”.

28 Nagata, Labor Contracts, 28–30. Yunoki Manabu, “Kinsei Itami”, 1–23.
128 mary louise nagata

retained hereditary servants employed them in these top positions. These

were the most trusted employees often serving as the agent or representa-
tive of the head of the stem business as well as managing branch shops
in Edo or major transportation hubs.29 Since the large family businesses
were controlled by such councils, disaggregating certain management
decisions can be difficult. Certainly the head of the business made most
everyday decisions, but the council could and did overrule him. Moreover,
the family could, and sometimes did, disown and divorce the head from
the family and business.30
The practices of adoption, divorce and disowning in the Japanese family
system were also important for shaping the opportunities for labor as well
as privileging ability over heredity in the management of family businesses,
including disowning family members. Skilled employees could, of course,
marry into the employer’s family. In addition, they could be adopted with-
out marriage to a daughter of the employer’s family, although they usually
married some member of the family, such as a cousin or a niece, after
adoption. Moreover, members of the family, including the head of the
family and business, could be forced out of the position and disowned
from the family, even in favor of an adoptive son from outside the family
such as an employee. The overall head of the Mitsui conglomerate was
disowned from the family in the mid eighteenth century.31 Because of
such actions, the family as business appeared to act as a corporate unit.
Certainly different members of the family were more involved in particu-
lar decisions than others, but there has been less research and there is less
information on the internal politics of family and business management.
Tokugawa family practice in central Japan shaped the labor market in
other ways as well. Nearly everyone married, but there often was little
room for establishing new households in village society. Moreover, some
time in service was a common way to gain training and experience in agri-
cultural techniques as well as in proto-industrial skills. Pre-marital service
was a useful way for young people to gain the training and skills, bring
extra income into the village and send extra youths out of the village to

29 Kobata Atsushi, Okamoto, 350–1. Nagata, Labor Contracts, 87.

30 Nagata, Labor Contracts, 27.
31 Kagawa Takayuki, “Kinsei shōnin”, 173–80. For discussion of an adoptive heir who
disowned his adoptive father and head of the business see also Mary Louise Nagata, “Mis-
tress or Wife?”
 factors that shaped the organization of labor 129

find other places to settle thus circulating labor and population to where
there was need for it.32
So the Japanese family provided the framework for business and there-
fore the organization of labor. In addition, Japanese family practice also
constrained and shaped the labor market and the opportunities provided
for labor. Certainly this is true everywhere. In Tokugawa Japan, how-
ever, the family was a large corporate unit that could be limited to one
household and nuclear unit, but often included many branches and sub-
branches and many non-kin employees and managers as well. Individu-
als were rotated in and out of various levels of kin relations through the
mechanisms of adoption, marriage, and divorce for the convenience and
longevity of the family business. Maintaining the profitability and longev-
ity of the business was important because the business provided the eco-
nomic base for many people, both kin and non-kin, and the welfare of the
group was more important than the individual wishes or success of any
individual member including the head or the heir.
The previous section showed how the political system and political,
legal and economic policies were important constraints and factors shap-
ing the organization of labor and the labor market. The next sections
examine the interaction and effect of economic and political crisis upon
urban labor in the final decades of the Tokugawa period through analysis
of the population registers of Kyoto neighborhoods.

Labor in Late Tokugawa Kyoto—Data, Background and Methodology

In the 1630s Tokugawa Iemitsu, the third shogun of the regime, banned
Christianity and took strict monopoly control of foreign trade, limiting
trade to China, Korea, the Netherlands and later Siam. Foreign ships from
these countries were allowed to come only to Nagasaki, although Japan also
maintained a closer trade relationship with Korea through the daimyo of
Tsushima Island.33 These events also marked the beginning of an annual
compilation of religious investigation registers designed to ensure that no
Japanese residents were Christian. These registers also proved useful for
keeping track of the population and vital events such as marriage, adop-
tion and divorce. Many domains also used the registers to keep track of
the agricultural labor force and many villages and domains further used

32 Nagata, “Leaving the Village”.

33 For a detailed discussion of Tokugawa foreign policy, see Toby, State and ­Diplomacy.
130 mary louise nagata

them to keep track of how much land, or agricultural production, each

peasant household was responsible for.34
Kyoto was the smallest of the three metropolises of Tokugawa society
after the political capital Edo and the commercial center Osaka. Kyoto
was the official capital as the home of the emperor and the imperial court
nobility as well as the capital of traditional culture. It was also a manufac-
turing center for silk and cotton textiles and other handicraft industries
such as ceramics, paper and lacquerware. The best data for understand-
ing the general nature of labor and the labor market is the religion and
population investigation registers. Although Edo and Osaka were larger
cities and Osaka was the larger commercial center, few of the population
registers of Edo are extant and those of Osaka did not record age until the
final year of the registration.
In the city of Kyoto these religious and population investigation reg-
isters (shūmon ninbetsu aratame chō)35 were compiled annually in the
ninth month of the lunar calendar by each neighborhood community, the
urban unit of local autonomy. The registers list each resident of the neigh-
borhood by religious sect (all Buddhist), and household of residence by
name and relation to the head of household. From 1843 the listings also
recorded the age and birth province of each individual and made greater
effort to keep track of movement into or out of the community through
notations on slips of paper pasted on the registers. In addition, the ninth
month listing each year was followed by an update listing of newcomers
to the neighborhood in the second lunar month of the following year.
These annual listings continued until 1868, the year the Tokugawa regime
fell, although some neighborhoods also compiled a final listing in 1869.
Since age and birth province are important information for indentifying
and linking individuals from one listing to the next and age is essential
for any demographic analysis, this study uses only the listings compiled
1843–1869.36 The recording of individual ages in the Kyoto registers also
means that they are more useful than those of Edo or Osaka where this
critical information was not recorded until later.
A neighborhood consisted of all households on both sides of one block
of a city street from one corner to the next. As autonomous units, each
neighborhood had its own headman and administrative council, collected

34 Laurel Cornell and Akira Hayami, “The Shumon Aratame”.

35 宗門人別改帳.
36 One neighborhood, Shimizu, included this information from 1842, so that listing is
also included in the data.
 factors that shaped the organization of labor 131

its own dues and had its own ordinances. There were 1500 such neighbor-
hoods in Kyoto during these final decades of the Tokugawa period, but
only twenty-eight neighborhoods have extant listings available for analysis
at this time. The availability of the listings is also spotty. Ten neighbor-
hoods only have one extant listing from the data period while others may
have multiple extant listings, but few or no consecutive listings. Only two
neighborhood collections have more than twenty of the twenty-six possi-
ble listings of the population registers providing a nearly unbroken series.
The data period of 1843–1868 is very good for investigating the effects
of economic and political crisis upon labor in the urban context. The data
series begins in the final year of the Tempō reforms (1841–1843) enacted
by the Grand Councilor Mizuno Tadakuni to address the economic crisis
after the Tempō famine of 1833–1837. Two reforms had negative effect for
Kyoto and particularly the Nishijin silk textile industry. The first was a
number of edicts that dissolved trade and business associations or any
organization that looked like a guild or a cartel on the assumption that
these associations were keeping prices high. However, these associations
functioned to regulate the market, manage credit, manage distribution
and enforce contracts. Moreover, many of these associations had been
formed by order of the state in the eighteenth century with the purpose
to control prices and keep them down. The effect of the edicts dissolv-
ing and prohibiting these associations was an upset of the distribution
system, lack of credit in a market that operated primarily on credit, and
price inflation.37
In addition to the negative effect upon the market and the resulting
price inflation, the trade associations also had insurance functions that
provided assistance to members. When they were dissolved, this safety
net for businesses in times of economic trouble disappeared. This was
particularly a problem for the Nishijin textile industry because another
part of the Tempō reforms restricted production and sale of luxury items
such as silks and introduced new sumptuary laws. So the Nishijin silk tex-
tile industry took a direct hit with these reforms, which also removed one
of the safety nets for businesses in the industry.38 Even so, the popula-
tion trends in Kyoto neighborhoods outside of the Nishijin district show
some evidence of economic growth or stability. Nevertheless there was a

37 Ishii Ryōsuke, Shōnin, 78–85. Miyamoto Mataji, Kabu nakama, 330, 337–343. See also
Nagata, “Brotherhoods”.
38 Yagi Tetsuo, “Kinsei”, 169–172. Hamano Kiyoshi, “Kinsei toshi”, 211.
132 mary louise nagata

further crisis in 1853 when the price of rice jumped fourteen percent due
to another poor harvest. The effect of this crisis was particularly heavy in
Nishijin because it came at a time when weaving workshops were shut-
ting down leaving weavers unemployed due to the effect of the reforms.
Ironically, records show family members of these weavers returning to
live-in service positions during this time of hardship. The city of Kyoto
also provided emergency welfare distributions of rice to households in
distress in that year.39
1853 was also the year that the American Commodore Matthew Perry
forced Japan to open wider to international trade. By 1857 Japan had trea-
ties with five Western nations and had opened several treaty ports. The
opening of the treaty ports contributed to the economic growth of some
parts of Japan, especially regions in the northeast that produced raw silk.
However, raw silk that was exported abroad was also raw silk that did not
reach the Kyoto market. This caused the price of raw silk to rise in Japan
bringing a supply crisis to the silk textile industry that now found itself in
competition to buy raw silk. At the same time, brokers who dealt in raw
silk took advantage of the rising prices and stockpiled the silk. In 1863
one such broker’s warehouse was attacked revealing a stock of 500,000
ryō of raw silk at a time when finding raw silk for sale in the market was
difficult.40
Two years earlier, in the middle of this crisis, another neighborhood
in Nishijin compiled a listing of households in need of assistance. The
price of rice also began a continuous rapid rise at that time. Whereas the
price of rice had been hovering at around eighty monme (300 grams) of
silver per koku (180 liters) of rice, it more than doubled during the Bunkyū
era 1861–64 to around 200 monme (750 grams) of silver per koku of rice.
Poor though the economy was in 1861, the price of rice had only begun
to rise. During the Keio era (1865–8) the price of one koku (180 liters)
of rice rose to 400–1,000 monme (15–37 kilograms) of silver. The magis-
trate’s office could not ignore this rise and solicited money for rice and
porridge directly from the large businesses, business associations and
neighborhoods, establishing various centers for rice and porridge to be
distributed.41 When rice became too expensive, they gave out money. The
economic situation for the city of Kyoto did not improve until after the

39 “Nanjū sha kazoku tori shirabe sho shita gaki,” [Draft of an investigation of families in
distress], 1853, Hanaguruma-chō collection, Kyoto City Library for Historical Documents.
40 Hamano Kiyoshi, “Kinsei toshi”.
41 Kobayashi Takehiro, “Bakumatsu ishin”, 3–15.
 factors that shaped the organization of labor 133

regime fell in 1868. During the 1860s, various outside political groups such
as leaders who opposed the regime, daimyo who came to Kyoto to assist
the Kyoto warden and, of course, the central government all competed to
provide assistance to the people of Kyoto. The impossibly high and rising
price of rice, however, suggests that little could be done until economic
controls were re-established under the new regime.
The treaties and the opening of the ports after 1857 also precipitated a
domestic political crisis. As the official capital and home of the emperor
and the imperial court, Kyoto was also at the center of the political crisis.
Political legitimacy in Japan has traditionally depended upon recognition
or support by the imperial court, usually in exchange for economic sup-
port of the imperial court. Young warriors opposing the Tokugawa regime
came to Kyoto in secret to make contacts and alliances with the imperial
court. Tokugawa representatives were also sent to Kyoto to prevent this
and strengthen and affirm regime ties to the court. Each grouped attacked
the other with assassinations, counter-assassinations and random attacks
on people suspected of supporting one side or the other in the city. The
Tokugawa regime sent a group of young loyalists to defend their position
and maintain order placing the city under martial law from 1863 until the
regime fell in 1868.42
So, the data period 1843–1868 begins with the final year of the Tempō
reforms and a brief period of growth followed by another major economic
crisis ultimately leading to a political crisis that finally brought the end of
the Tokugawa regime. The problems were aggravated even further when
much of the city burned down in the seventh month of 1864. Did these
crises reduce immigration to the city or encourage increased emigration?
Did they limit the labor market or did live-in employment provide refuge
in time of need for people who otherwise would have been finished with
the life-cycle service portion of their lives?
The next section will use analysis of the population registers of Kyoto
to address these questions in several ways. First I examine the probabil-
ity of being a servant as it changed over time and the probability that a
servant would be a native of Kyoto or an immigrant from another prov-
ince, as well as the average ages of servants that appear in the registers.
Second I use consecutive year data to investigate the probability of enter-
ing or leaving the neighborhood communities, comparing again natives
and immigrants. More pertinently, I then limit this analysis to servants.

42 Totman, The Collapse.

134 mary louise nagata

Finally, I examine the results according to the background of economic

and political crises to address the above questions.

Labor in Political and Economic Crisis

The above story of economic and political crises during the data period
1843–1868 suggests that the Nishijin silk textile district was the hardest
hit. Although there are records of hardship distributions in two Nishijin
neighborhoods, the population trend does not entirely support this. Fig-
ure 1 shows the annual population trend for the two neighborhoods that
have near continuous series of extant population registers during the data
period. Seido is a neighborhood near the commercial center of the city
containing a number of family businesses. Kankiji is a neighborhood in
Nishijin also containing a number of family businesses including weav-
ing workshops that employed young women as apprentices. The popula-
tion trend for Kankiji neighborhood shows increase 1852–6, followed by
a decline until 1861, and then rapid growth until 1865. Thus, the popula-
tion was growing at the time of the 1853 hardship distribution, but had
declined to the lowest point at the time of the 1861 hardship distribution.
On the other hand, the period of population decline 1857–61 is the period
when the opening of the ports began to affect the silk trade. The popula-
tion trend for Seido, less dependent upon the silk trade, is relatively flat.
The dip in population in 1864 reflects the effect of the fire that burned
down much of the city that year including houses in Seido neighborhood
(see figure 1).
Figure 2 shows relations to head of household by age for the full data
set graphically demonstrating the life-cycle nature of live-in employment
(hereafter called service). The age period of life-cycle service was 11–28 for
young men and 15–25 for young women. Figure 3 shows the probabilities
that natives or immigrants would be in service at ages 11–30 over time.
This figure shows that natives and immigrants show the same employ-
ment trend, but immigrants at these ages were far more likely to be in
service than natives. In other words, most non-natives who came to the
city came for labor migration. Moreover, 1853 was a year when ninety per-
cent of non-natives and fifty percent of natives ages 11–30 were in service,
an increase of roughly ten percent from previous levels. In addition, 1861
shows a dip in the proportion of both natives and immigrants in service
at these ages, reflecting the economic hardship for employers, but 1864
shows a sudden peak in the year of the great fire. This suggests that people
 factors that shaped the organization of labor 135

300

250

200

150 Kankiji
Seido

100

50

0
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868

Figure 1 Neighborhood population trends.

not in service left for safer refuge elsewhere if they had other places to
go, and this could also be the reason for the peak in 1853 as well. While
both the native and immigrant trends show a general decline, this is less
obvious among natives as periods of crisis seem to have either sent immi-
grants not in service outside the city or sent them back into service.
Figures 4–5 provide more explanation for these trends. Figure 4 merely
shows immigrant and native servants as proportions of the total popu-
lations of immigrants or natives. In the crisis years 1853, 1861 and 1864
the probability that an immigrant of any age would be in service rose
dramatically, especially in 1853 and 1864 when sixty percent of the entire
immigrant population was in service, suggesting both that immigrants
not in service left or re-entered service. Figure 5 shows the proportion of
immigrants among all servants and in the total population, confirming the
explanation that times of crisis sent people looking for live-in service posi-
tions, likely for the provisions of room board and clothing that came with
136 mary louise nagata

900

800
sibling
700 stem
spouse
600 sib
servant
Observations

500 servants parent

spouse
400 other
nonstem
300 nonkin
200 head
parent
child child
100 head

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91
Age

Figure 2 Observations by age and relation to head.

0.9

0.8

0.7

0.6

0.5 Immigrants

0.4 Natives

0.3

0.2

0.1

0
43

45

47

49

51

53

55

57

59

61

63

65

67
18

18
18

18

18
18

18

18

18

18

18

18
18

Figure 3 Probability that natives and immigrants ages 11–30 were servants.
 factors that shaped the organization of labor 137

0.7

0.6

0.5

0.4 Immigrants

Natives
0.3
Proportion
0.2 Immigrant

0.1

0
1843
1845
1847
1849
1851
1853
1855
1857
1859
1861
1863
1865
1867

Figure 4 Probability that natives or immigrants were servants.

0.4

0. 35

0. 3

0. 25

0. 2 Servants

0.15 Total

0.1

0. 05

0
1851

1861
1843

1853
1855
1857
1845
1847

1859

1863
1849

1865
1867

Figure 5 Proportion of immigrants in the population.

138 mary louise nagata

the live-in position.43 The proportion of immigrants in the total popula-

tion declined from fifteen to ten percent after 1853, but the proportion
of immigrants in the live-in service population increased 1860–64. After
1864, during the political crisis and martial law, the proportion of the
population in service fell dramatically for both natives and immigrants as
the people of Kyoto stayed in place. Possibly, those who had places to go,
whether leaving the city or entering service, went during the economic
crisis and then settled in to wait out the political crisis.
One sign of people re-entering service for refuge could be a rise in the
average age of live-in employees. Figure 6 compares the average ages of
native and immigrant young men in service over time. These young men
followed the career path of apprentice to skilled worker and management
employee with the successful ones becoming branch managers or mov-
ing out to establish their own branches or businesses. The young average
age of these employees represents the high turnover as many left service
before finishing the apprenticeship period or soon after. The average age
for completing apprenticeship was seventeen.44 At the same time, some
firms also employed older young men in their twenties in domestic posi-
tions. Figure 6 shows the average age of immigrant men in service rising
higher than native men in times of economic and political crisis. The aver-
age ages of native men in service fell during the economic crisis, possibly
reflecting difficulty in advancing as skilled worker, or reflecting young
men who took their skills home or outside of the city.
Life-cycle service for young women was generally understood as a way
to gain skills in managing a business household in preparation for mar-
riage, sometimes including the craft skills to assist their future marital
households in the textile industry. The age that young women left service
was closely linked to the age at marriage. I have already noted that immi-
grant women in Kyoto appear either in service, married or widowed.45
Thus the average age of young women in service should be around nine-
teen or twenty since they entered service at around sixteen and left at
around twenty-three. Figure 7 shows that the average ages for immigrant
women were higher until 1853, but the age of native women rose close
to that of immigrant women in service in 1853. Then the average age of
women in service, both immigrant and native, fell until 1860. The average

43 Nagata, Labor Contracts.

44 Ibid.
45 See for example, Nagata and Hamano, “Marriage Market”.
 factors that shaped the organization of labor 139

25
24
23
22
21
Age

20
Immigrants
19
Natives
18
17
16
15
1851

1861
1853
1855
1857
1843
1845
1847

1859

1863
1865
1867
1849

1869

Figure 6 Average ages of immigrants and native male servants.

27

25

23
Age

21
Immigrants

19 Natives

17

15
1843

1846

1849

1852

1855

1858

1861

1864

1867

Figure 7 Average ages of immigrant and native female servants.

140 mary louise nagata

age for native women remained lower, declining to around eighteen until
1867 when it returned to the original level. The average age for immigrant
women, however, rose dramatically to twenty-five and generally remained
rather high until the end of the period. This analysis supports the explana-
tion of service as a refuge in time of political and economic crisis, espe-
cially in time of political crisis.

Concluding Discussion

This discussion of the constraints and factors that shaped the organiza-
tion of labor and the labor market in Tokugawa Japan has taken us from
the sixteenth century and the founding of the Tokugawa regime in 1603
to the fall of the regime in 1868. The constraints and factors that have
played major roles in shaping labor during this period were political, legal,
economic, social and cultural. In other words, they involved all aspects of
Tokugawa society.
Political policies during reunification and the beginning of the Toku-
gawa period changed the social and economic framework of society and
the labor market. The separation of warriors from the land, the cadas-
tral registration, laws prohibiting the sale of persons and other policies
designed to establish peace and security formed the framework of Toku-
gawa society creating an urban market and a population dependent upon
supplying its needs from the market as well as changing the source of
labor for that market production. On the other hand, political policies
often had very little effect upon labor even when they were intended to
address the labor market. For example, attempts to prevent or limit labor
out-migration from northeastern domains were ineffective as shown by
long lists of people who left without reporting where they went. Perhaps
because of local autonomy, political authorities did not play a major role
in shaping village industries or the rural-urban labor market. Instead,
the timing and direction of labor migration depended much upon the
convenience of village society as well as networks of kin, friends, region
and business. The main concern of political authorities was maximiz-
ing the stability of the rice harvest and therefore their tax income. Agri-
cultural production and the land under cultivation was also expanding,
and there was no large population forced off the land or a large landless
population. Village communities and their member families balanced the
need for extra income from remittances, useful for paying taxes or other
investment, with the need to retain sufficient agricultural labor force. If
 factors that shaped the organization of labor 141

a­ nything, political authorities and village communities in some parts of

Japan worried about land left abandoned or underutilized and looked for
ways to retain rural agricultural labor force in the face of demand for labor
in other places. Encouraging investment in manufacturing in the villages
could keep labor available for agriculture.
Village economic needs and the labor market in turn shaped the demo-
graphic system. Everyone married in the Japanese family system, but only
one married child inherited management of the family holdings. If the
land under cultivation or the village economy expanded so there was
room for more households, the village would try to retain more people
and encourage those who left for labor migration to later return and settle
down. If not, then labor migration was a way to send young people out to
find other places to settle and other people to marry as well as circulate
labor to where it was most needed. Labor within villages, both agricul-
tural and proto-industrial or handicraft labor, and labor migration to and
from villages depended upon local solutions balancing these needs as well
as whether and how many families in the village or nearby villages had
invested in market production. This complex mixture of concerns also
shaped regional differences in the labor market and the demographic sys-
tem. Villagers in Northeastern Japan where agriculture was not so produc-
tive needed outside income for survival, but also needed agricultural labor
force. Marriage in this region was early and universal and labor migration
took place after marriage. Villages in central Japan were close to many
economic opportunities and villages could need to shed population. Labor
migration took a life-cycle form and marriage was delayed, although still
universal, especially for younger sons and young people of less affluent
families in the village community.
Through business investment and expansion following the stem, branch
and sub-branch framework of the Japanese stem family, corporate family
and business organizations formed a network of households that func-
tioned as shops, manufactories and offices, often with manufacturing in
the home and neighboring villages, retail shops in cities and towns with
some branches also serving as warehouses, and transportation hubs that
all served as residences where family members, managers and apprentices
lived. Traditional techniques and the training of labor in Europe were
often under the control of guilds and the apprenticeship system. Japa-
nese trade associations or other guild like organizations were not con-
cerned with training labor or controlling apprenticeship. Each employer
trained its own employees in its own methods “in house”. Thus family,
business and regional networks became very important as each employer
142 mary louise nagata

developed training methods and tried to control their techniques without

formal constraints. This was still apprenticeship in that young employees
worked with or for a skilled worker and, in the process of following direc-
tions and helping out, learned the techniques developed and used by the
employer’s business. But this was not subject to outside testing or valida-
tion beyond success in the market.
Labor contracts and the various categories of relations between
employer and employee were about economic relations, prevention of
conflict and mechanisms to settle conflict privately, so there was little dif-
ference between industries or sectors of the economy. The different cat-
egories reflect the degree the worker was integrated into the employer’s
business. Casual labor was not integrated at all. Pawn servants were not
much integrated into family or business. Live-in workers, however, became
non-kin members of the household, family and business. The higher they
reached in the business hierarchy, the stronger the social identification
with the employer. Since one of the major issues of the Japanese labor
market was trust, the hereditary servants one would normally consider at
the bottom of the labor hierarchy were often at the top and considered
almost family members in Japanese businesses.
Japanese family practices of adoption, filiacentric marriage where an
heiress is used to recruit a capable son-in-law, divorce and the integration
of non-kin into the family also offered many opportunities for the capable
and ambitious employee to gain a sponsor to establish his own business
or a branch of the employer’s business, rise to top management or even
inherit management of the employer’s business. They also provided the
family business with methods to ensure capable management and leader-
ship while remaining a family business.
Analysis at the more micro level of the labor response to political and
economic crisis in Kyoto, still the official capital and home of the emperor
at the end of the Tokugawa period, revealed that live-in service continued
to provide refuge from hardship. Unemployment was certainly a problem
during the crisis. At the same time, people affected by the crisis tried to
find more secure situations with some, particularly immigrants to the city,
leaving the city and others returning to service. Especially in the crisis of
the final years of the regime, the majority of immigrants were in service
suggesting that others left the city or avoided coming. In spite of these
effects, political and economic crises did not much change the system.
In the end, labor in Tokugawa Japan was shaped and constrained by
a complex package of factors. Political policies, but also the political
system formed the social, legal and political framework for economic
 factors that shaped the organization of labor 143

d­ evelopment and labor. The economic needs of village as well as families

and individuals were also important for the timing, direction and volume
of labor migration and labor opportunities. Urbanization also played a
role, but village production and the need to retain agricultural labor were
also important factors. The family system also provided a business frame-
work that could connect rural and urban manufacturing and markets as
well as labor and migration opportunities through the family and business
network.
 Contractual Relations, Tariffs and Customs in the
Lyon Silk Industry in the Nineteenth Century

Pierre Vernus

In the nineteenth century, industrialization in France combined the

dynamics of several types of industrial systems. Among them, alternatives
to more or less standardized mass production in factories, in which the
Lyon silk industry can be included, played a preponderant role for a long
time.1 The Lyon silk industry, a unique community of this kind, called the
‘Fabrique’ contributed greatly to making Lyon an important industrial and
commercial center on a global scale, and for a long time it employed the
majority of Lyon’s working population. Throughout the nineteenth cen-
tury it increased its production and made a great contribution to French
exports. The Lyon silk industry is also one of the most representative
examples of a collective industry;2 a system of localized production made
up of specialized productive entities of various shapes and sizes, linked by
the process of manufacturing goods, in this case silk fabric. The elements
of this economic ‘meso-system’3 were linked by a network of various
relationships which could stretch from cooperation to competition and
included the pay system. The stability of the organizational set up of this
system has often been emphasized. It was marked by the domination of
mercantile capital over the workshop and from the eighteenth century the
vast majority of workshop heads had lost their access to the market which
was controlled by merchants who organized production. But behind the
organizational continuity of the industry, (at least until the early 1880s),
the first decades of the nineteenth century were marked by great changes;
institutional, administrative, legal and political, by a shift in the markets
and consequently in production, as well as changes in personnel.
We will briefly recall the main structural characteristics of the Lyon silk
industry (the ‘Fabrique’) and explain the new institutional and legal frame-
work in which the contractual relations between the merchants and the

1 Dewerpe, Le monde; Sabel and Zeitlin. “Historical”; Woronoff, Histoire.

2 This notion borrowed from Le Play was revived by Cottereau, “The Distinctiveness”.
3 On the notion of economic ‘meso-system’, see De Bandt, “La filière.” In Traité.
146 pierre vernus

workshop heads from the beginning of the nineteenth century took place.
Then our aim will be to try to understand the effect of these changes on
the relations between merchants and workshop heads and in particular
the contractual relations between them.
We will show that the legal equality in the contractual relations between
merchants and workshop heads established by the French Revolution
did not put an end to the latters’ economic dependence on the formers
that had grown gradually during the seventeenth and eighteenth centu-
ries when the old corporations polarized increasingly. This dependence
placed mainly on the workshops heads’ shoulders the consequences of the
incompleteness of the contacts due to the uncertainties characteristic of
the manufacture of silk fabrics but also to changes in the personnel and the
markets.
The workshop heads tried to reduce the social imbalance between
them and merchants in negotiations which could generate major con-
straints on their working conditions in a competitive framework such as
intensification of work and threat to their freedom to organize autonomy,
by demanding the setting up of a tariff, but also by trying to improve the
organization of information between weavers, and more widely inside the
industry.
Moreover the traditional practices of the industry, one of whose func-
tions was to reduce the uncertainty over certain aspects of the contractual
relationship often worried the weavers. The changes the industry had to
make encouraged attempts to make the old customs obsolete or to set up
new ones. So it will be useful to explain the strategies aimed at maintain-
ing the practices that the different parties judged fair, or modifying them
in a way that seemed more favourable.

The Apparent Permanence of the Organizational Pattern

At the beginning of the Restoration the productive structures of the ‘Fab-

rique’ seemed little different from those it used at the end of the eigh-
teenth century. It was a luxury industry depending on the extremely
random fluctuations of a demand which was sensitive to the capricious
nature of fashion and to political and diplomatic crises. The organization
seemed to be flexible, allowing rapid adaptation of the volume of its pro-
duction and the range of products.
The system was run by merchants, also called merchant-manufacturers,
or even ‘merchants ordering products’, but who did not produce the goods
 contractual relations, tariffs and customs 147

themselves. They prepared their collections of samples (several hundred

for the largest houses) before the two main seasons, then, after booking in
their orders, bought the raw silk fiber which they paid weavers to trans-
form into fabric and dye, before handing them over to the workshop heads
who wove the prepared silk, following the instructions they were given.
The weaving was done by the ‘silk workers’ (the ‘canuts’). But this body,
far from being homogeneous, was made up of two large groups, the work-
shop heads and the journeymen. The workshop head wove in his own
shop and owned his own looms. He was in a middle ground between the
merchant, a mercantile capitalist, and the journeymen or women and
the apprentices, who he employed. This ternary structure was the result
of a process which, inside the community during the old regime (before
1789), had led to growing social differences among the group of ‘masters’
which ended up as a split between the master-merchants and the master-
workers (usually workshop heads). This development was accompanied
by the adoption of various regulations which led to the traders usurp-
ing the responsibilities and management of the community. Right up
to the beginning of the Revolution the master workers challenged this
development, at first violently, then more peacefully. Towards the end
of the eighteenth century they finally saw it as inevitable. From then on
the major social division was between the merchants and the large group
made up of the master-workers and the journeyman workers. There was
a long standing convergence of their interests and conditions. Most of the
journeymen ended up by setting up as master workers, so finally, the only
clear difference was that of age.
Moreover the rule, maintained throughout the nineteenth century,
which attributed to the journeymen a part (usually 50%), of the wages
given to the master-worker, linked their pay to that of the workshop
heads.4

A New Institutional Framework

The stability of this global scheme of organization of the ‘Fabrique’ in the

first decades of the nineteenth century contrasted with the changes in
the legislative and institutional frameworks which appeared during the
Revolution and the Empire. The relations of provision of work and pay

4 Garden, Lyon et les Lyonnais; Godart, L’ouvrier en soie.

148 pierre vernus

between the merchants and the workshop heads had to fit into a new
order which constituted a break with the old corporate order and was
based on new principles, in particular civil equality and the freedom of
work and trade.
Merchants and workshop heads agreed to recognize that the work
relations which they had established were of a contractual nature and
resulted from interpersonal negotiations, even if they disagreed on the
content they gave to this notion.5 In this respect, in the legal framework
defined by the Civil Code, these relations were dependent on contract law.
The workshop heads were considered as work contractors and the con-
tract made with the trader was seen as a contract for work. However, the
intermediate position of the Lyonnais workshop heads did not fit into the
categories lawmakers had used in the texts concerning the administration
of work. The law of 12 Germinal year 11 (12 April 1803) and in particular
its third section dealing with the workers’ accounts booklets, completed
by the decree of 19 Frimaire, year 12, which applied “only to workers
employed as journeymen or simple workers” did not apply to the Lyon-
nais shop heads because they had benefitted from the status of ‘masters’
in the former corporate structure and were the employers of journeymen
and workers. It was the third section of the law of 18 March 1806 which
created the ‘conseil de prud’hommes’, (originally a conciliation board, but
with wider legal prerogatives), which dealt specifically with their relations
with the traders. It obliged the workshop heads to keep a double accounts
booklet, delivered by the tribunal for each of the looms they possessed, to
keep a copy and hand it over to the merchant-manufacturer for whom he
worked on his loom. When a workshop head stopped working a loom for
a trader, the merchant manufacturer had to settle his accounts and, when
necessary, give the details of any debts he had incurred towards him. The
other merchants for whom the workshop head worked also had to inspect
the booklet and indicate the state of their accounts with the weaver. In
the case of debt, the new merchant for whom the workshop foreman set
up the loom committed himself to keeping back an eighth of the price
of transformed silk to pay back his creditor colleagues, in chronological
order of the creditors. The law maintained the practice of accounts book-
lets which had existed before the Revolution It also revived a measure
installed in 1737, but abandoned in 1744, which linked the debt not to the
workshop head but to the loom. Besides, this system constituted a credit

5 Cottereau, “La disincorporation”.

 contractual relations, tariffs and customs 149

mechanism for the workshop heads and offered a collective guarantee of

their debts. In this way the shop heads had fully gained their freedom,
often challenged, to change their clients and use their looms for several
merchants, but also to stop working for a merchant as soon as the job he
had been given was finished, even if he was in debt to him.6
By getting rid of the corporate structures the Revolution had opened
up a period of uncertainty because the institutions in charge of organizing
and regulating production disappeared with them. Even if the institutions
had disappeared, this was not the case for demands for regulation.7 From
now on what kind of regulation should be put in place? In what institu-
tional framework? And what should be done with the old regulations?
The statute of 1744 which had been modified several times and which
was still in force just before the Revolution, was often cited as a reference.
Thus, in 1807, the chairman of the ‘conseil de prud’hommes’, referring to
article 10 of section 2 of the law of 1806 which told the board to note any
infractions of the new laws and statutes or those which had been revived,
considered that all the articles of this statute (of 1744) which conformed
to the law of 18 March 1806 should “be carried out, until replaced by new
ones”.8 There was a more general call for reflection on the ways and means
of reestablishing regulation for looms; several projects were produced, but
were never carried out. Nevertheless, the ‘Fabrique’ benefitted from the
institutional reconstruction carried out under the First Empire9 which
was marked by the propensity of the minister of the interior to consult
the notables rather than the workers, as can be seen by the organization
of the industrial tribunal itself. The law of 18 March 1806, even if it low-
ered the proportions anticipated in the first Lyonnais project10 gave five
of the nine posts as councilor to merchant manufacturers, giving them a

6 The workshop head and the merchant “are linked to each other simply by mutual
convenience; and when this link ceases to exist, both parties settle their accounts and
separate without discussion or reproach” (Montfalcon, Histoire des insurrections); “neither
one nor the other [the workshop head] work exclusively for this merchant or that, but
successively and for several at the same time. The merchants’ agent usually overlooks the
production.” (Villermé, Tableau), 355. On this point see also Pariset, Histoire de la Fabrique
and especially Cottereau, “The fate”.
7 Minard, “Les corporations”.
8 Lyon municipal Archives (AM Lyon) 783 Wp 1, Letter from the President of the Lyon
‘conseil des prud’hommes’ to the mayor, 17 June 1807.
9 Moullier, Le ministère, chapter 14.
10 The report by Regny senior to the Lyon chamber of commerce 7 thermidor year XII,
anticipated giving two thirds of the seats to the merchants (Archives of the Lyon chamber
of commerce (ACC Lyon), minutes).
150 pierre vernus

majority on paper.11 Beyond this, the only workshop foremen who could
vote were those who were duly licensed by the tribunal, which meant the
tiny minority who produced independently, (105 in 1807, only 50–60 in
1831).12 Moreover, even if the relations between the heads of workshops
and journeymen were the responsibility of the tribunal, the journeymen
were not represented. Nevertheless, the rapid increase in the activities of
the tribunal witnesses to the fact that it answered a need on the part of
all the people involved.13

Contractual Inadequacies and Social Imbalance

Contract law left both parties, merchant and workshop head, great leeway
in negotiations about the content of the contract, which gave the oppor-
tunity to really negotiate, but the hazards Inherent in the production of
silk goods gave rise to many uncertainties as to the carrying out of the
contract. On the other hand, despite the affirmation of equality between
the two negotiators in theory, their respective positions were not equal.
Even though long term relations over several years between a merchant
and the workshop head could arise,14 each order from the first to the sec-
ond required its own contract which produced real negotiations about
the price for transforming silk into fabric, delivery dates and the quality
of the raw material, and the fabric itself. Two high points influenced the
negotiation, firstly the reception of the raw materials and the manufactur-
ing instructions given to the weaver, and secondly the delivery of the fab-
ric to the merchant. The contract was written—or should have been—in
the accounts booklet on which were generally written the conditions (the
nature of the article, the price, delivery dates, sharing of the overheads
for the setting up of the loom, etc.) As contract law stipulated, any failure
by either of the parties implied paying damages (article 1142 CC). The pay
was the equivalent of a salary (the price of transforming silk). So there
were neither hiring for a specific period nor, unlike for industrial workers,

11 Real equality in numbers was only established by the law of 7 February 1880.
12 Rude, L’insurrection; L’Écho de la Fabrique (from now on EF), 12 (1832): 3.
13 The number of cases arbitrated tripled between 1807 and 1810, rising from 407 to 1218
(National Archives (AN) F12 934).
14 Worries about the future were very often felt both by the merchant who gave his
orders, when he wished to be sure of keeping a weaver whose ability he recognised, and
by the weaver who needed new orders to keep his/her looms working.
 contractual relations, tariffs and customs 151

a fixed time for starting and finishing the day: it was up to the workshop
foreman to manage his work time to respect the delivery dates.
The contract did not lead to hierarchical relations; the workshop head
often suggested improvements in the designs the merchants gave them15
and were free to organize and share out the work inside their workshop.
What is more, the quality control of the work was not just a single stamp
after the work was finished; contacts between the two parties or their
assistants, were maintained throughout the period of production of the
fabric in visits to the workshops, and the fabric had to undergo two tests,
when it arrived at the merchant’s, and finally when it was delivered to
the customer who had given the initial order to the merchant.16 As well
as this, unlike what could be seen in a factory situation, the contract con-
cerning quality was not in the context of a company, but in this particular
intermediate collective called the ‘Fabrique’, with its own rules and cus-
toms. All these practices were aimed at clearing up, at least partly, some
of the uncertainties that the contract had left vague, uncertainties which
were largely determined by the nature of the products, which were central
to the work obligations.
The silk goods which came from the looms of the Lyonnais workshops
were luxury or semi-luxury items. Their production was not homogeneous.
The big families of woven materials (plain, figured, novelties, velvets)
were themselves subdivided into numerous categories defined according
to their weaves and designs, the characteristics of their woof and weft
frames and their chains. And these categories in turn included a large
number of textures. For example in taffetas there were plain, glossy, shot
silk, changeable taffeta, Marceline, sarsenet, ‘gros de Naples’ and poult-de-
soie, etc.
However, even the materials which were easier to produce, like plain
taffeta demanded “extensive and rational experience”. The know-how of
the weavers, like that of the other professions involved in the production
of materials contributed to their final appearance, by their ability to get
over the difficulties arising from imperfections in the material, difficulties
in adjusting the loom or reproducing the designs. This know-how was an
essential quality in production, inside which new types of material were
constantly being created.

15 Montfalcon, Histoire, 17. Confirmed by Bezon who, among the people his dictionary
of silks is addressed to, mentions “the workshop heads who throw themselves into serious
and deep study of production, which they saw as perfectible”, Bezon, Dictionnaire.
16 Cottereau, Distinctiveness, 91.
152 pierre vernus

According to the goods to be produced the merchants thus had to go to

the weaver he thought was most capable of producing the desired result.
These skills were reinforced by the polyvalent nature of the workshops’
production. Unlike the Spitalfields weavers in London, who were special-
ized in one type of material, their Lyonnais counterparts were skilled in
several types.17 Overall, in the framework of traditional silk manufacture
before the depression of the 1880s, silk goods were made up of goods
which could be qualified as ‘specially ordered’ or ‘interpersonal’, to use the
categories suggested by Robert Salais and Michael Storper.18 From then
on, several gradations around this pole could be seen, on the one hand
towards ‘standard-specially ordered’ quality for the most common goods,
and on the other, towards ‘immaterial quality’, for example when creating
a new kind of material or a new article using combinations of thread, new
patterns or new finishes.
The inherent uncertainties involved in producing such a variety of arti-
cles took different forms for the merchants and the workshop heads. The
first one concerned the mutual understanding of the needs expressed by
the merchant and the suggestions put forward by the workshop head. It
was partly eased by the fact that, because of its longevity, Lyonnais silk
weaving formed a community which benefitted from particular advan-
tages, similar to the localized industries mentioned by Marshall, and in
particular a ‘shared language’ and specialized and diffuse skills, transmit-
ted by the publication of technical journals and by being passed down
from generation to generation.19
A second uncertainty came from the high price of silk. The merchant
who gave it to the weaver suspected that the weaver might steal some
for his own use. Weaving inevitably produced waste so they had to find
out what proportion was acceptable. Another uncertainty concerned the
finished quality of the goods. When the roll of material was delivered,
how could a mutual agreement on whether the product met the require-
ments of the person giving the order be arrived at? The management of
time was also important. For the merchant, if the production deadlines
were not met it meant the risk of losses on the selling price or the can-
cellation of the order by the client, especially for the seasonal novelties
the price of which were high at the beginning of the season but fell later.

17 Cottereau, Distinctiveness, 91–93.

18 Salais and Storper, Les mondes.
19 Marshall, Principles; L’Écho de la Fabrique de 1841, 64 (1844), 1.
 contractual relations, tariffs and customs 153

For the weaver, if the raw materials were delivered late, some time after
the loom had been set up, there was a slack period which meant a loss
of income to be made up. Moreover the setting up of a loom, especially
for the production of novelties and fashioned goods, could take several
days or even weeks which implied high expenses for the workshop head.
But if it was not followed up by an order allowing him to cover the set-
ting up expenses, who would bear the cost? The workshop foreman? The
merchant for whom the loom had been set up? Both, but in what propor-
tion? Since there was no standardization in the process of preparing the
silk threads and because of the diversity of their origins, the quality of the
silks was an element which could greatly slow down the making of fabric
and damage the final quality of the material. On his side the merchant
had to have guarantees that he would get the necessary quality of work,
and flexibility, especially if he aimed to have ‘tasteful’ goods or novelties
woven for him. From this standpoint he was interested in the mainte-
nance and renewal of the skills needed for his productions. The varia-
tions in the economic situation were a widespread risk especially since
luxury goods were the first to be hit by a reduction in spending. The ebb
and flow of fashion were added to these economic swings, without being
totally in synch with them. After being a sudden hit an article could just
as quickly be discarded. The organization of production limited the risks
for the merchants whose investments in production were limited. On the
other hand, they were borne more heavily by the weavers whose work and
wages varied greatly with time, even though the combination of special-
ties mentioned earlier partly compensated for their effects.

Shifts in Markets, Types of Production and Personnel

By unsettling the old practices, the shifts in the markets and the trade
cycles increased the sources of uncertainty. In fact the strong growth
in the British and American markets to the detriment of the traditional
markets of continental Europe brought with them an acceleration of the
rhythm of trade and great changes in the rhythm of work in the silk pro-
duction centers.
During the last few years the way of doing business has completely changed.
In the past an order was given six months in advance. We used to work
for Germany, France and Spain and we knew nothing of the great deals
with America and England. The merchant used exactly the same number
of looms throughout the year. He never varied more than ten or twelve: his
154 pierre vernus

clientele was stable. Today it is the opposite; the price of silks, the designs,
the colors, etc. change twice per season, the orders have to be delivered at
thirty days, six weeks, two months at the most. Among all these constant
changes the producer has to adapt; the number of looms may go up to 300,
or go down to about 150 according to the extent of his capital, but always
in this sort of proportion. What effect do these ups and downs produce? He
can never guarantee a worker a fixed number of days of work. The worker
sometimes works for six to eight producers so that they all spend their lives
running after each other. On the one hand, the worker often delivers his
loom, set up in a certain way, to the first producer he finds, which forces
him to totally change the setup of his loom and finds himself faced with ever
renewed overheads. On the other hand, to finish on time an order needing
forty looms, the manufacturer is obliged to look around for a fortnight, to
knock on two hundred doors and hire workers left, right and center, who
often have never worked on the article they are offered.20
The expansion of the new Anglo-Saxon markets also meant a change of
clientele. The dynamism of the demand from the middle classes overtook
that of the old ecclesiastic hierarchy or aristocrats. This encouraged the
renewal and diversification of products which contributed to shaking up
the methods and practices applied to the older types of production21 and
also led some traders to try to change the rules governing their relations
with the weavers. For example, the rapid changes of materials on the
looms brought the question of the setting up of the looms to the fore.22
Finally, these developments tended towards the disappearance, among
the main merchants, of the desire for regulations which had been expressed
up to the first half of the 1820s. The new generation of merchant-producers
seems to have been much more amenable to the principles of ‘laissez faire’
capitalism than the previous one, both because of the spread of the ideas
of economic liberalism and because the old rules hampered the adapta-
tion to changes in demand.
What is more, the sharp increase in the numbers of merchants (which
had doubled between the Empire and 1827, before settling to between
400 and 450), intensified local competition. Even though the biggest mer-
chants were in a comfortable financial position which helped them to
weather the crises, the financial situation of a good number of the new
arrivals, much tighter, forced them in slack periods to sell off their stock to

20 EF, 46 (1833), 5.
21 Cayez, Métiers.
22 Benoît, Confession, 66.
 contractual relations, tariffs and customs 155

honor their debts and lower their prices to be in favour with the buyers.23
This movement forced itself eventually on their competitors and affected
the price of transforming silk goods. Besides this, the renewal among the
ranks of the merchants undermined the traditional processes of self-
regulation and the internal social constraints in the ‘Fabrique’.
The unease brought about by these changes soured the relations
between the merchant-manufacturers and the workshop heads and this
grew with the crisis at the end of the Restoration. The workshop heads
complained of abuses perpetrated by the merchant-manufacturers and
even though they were in a privileged position compared with other
groups of workers, their position remained precarious and suffered the
consequences of periodic unemployment. They were deeply affected by
the difficult economic situation, especially since the drop in wages also
affected the price of figured products.

Imbalance of Positions

The weavers were all the more sensitive to the mounting uncertainties
concerning their relations with the merchants that the theoretical equal-
ity of contract set by contract law or by economic liberalism became in
fact an imbalance of positions. In fact, in the eyes of the workshop heads,
the argument upheld by many merchant-manufacturers, that the work-
shop heads were free to refuse a salary that was too low, was simply a
trap. From the last decades of the eighteenth century the master weavers
argued that:
We only have the appearance of this freedom; no rule obliges us to accept
work for which we would be underpaid, but the most implacable law, that
of need, imposes on us the cruel reality; our workshops are not fortresses
which it is easy to bring down by starving us out; we have to live every day
and work every day.24
Beyond this, the prices paid for transforming silk greatly influenced the
length and intensity of the weavers’ work because in order to counteract
the fall in prices they had to work longer.
Besides, the merchant had at his disposal various means which allowed
him to transfer part of the production costs to the workshop head. He

23 Montfalcon, Histoire, 35, 48–49.

24 Godart, Ouvrier, 262.
156 pierre vernus

could, for example, provide substandard threads, make the weaver pay
certain costs which he had paid himself before, or extend the payment
date. On the other hand during slack periods the merchant-manufacturer
found himself in a better position to demand earlier delivery, which
forced the weaver to lengthen his workdays if he wanted to avoid late-
ness penalties. But weaving at night increases the risk of bad work due to
fatigue, and it is more difficult to distinguish shades of color or see traces
of smoke or grease.
The workshop heads were affected in various different ways by this
inequality in contractual relations. In fact as a group they were riven with
splits on several levels. The first division was based on the hierarchy of
profits and wealth. At the end of the 1850s L. Reybaud distinguished a tiny
minority of the privileged who owned their house or who had a capital of
20–30,000 francs. Below them were the ‘majority of good workshop heads’
who managed to earn 3–4,000 francs net when work was continuous. The
third level consisted of ‘the less rich heads’ whose takings did not exceed
2 000 to 2 500 francs. The base of this pyramid was made up of ‘the major-
ity of heads who,’ because they owned fewer looms or their equipment
was not up to scratch’, who earned less than 2 000 francs, which could go
down to 1 200 or even 1 000 francs.
Furthermore, although there was a certain diversification in the pro-
duction of the workshops, there was a clear difference between the weav-
ing of plain articles and fashioned goods. The amount paid for the former
was sometimes three or four times lower than that for the latter. However
the production of fashioned goods was very dependent on fluctuations in
fashion. They stopped before the plain fabrics and they often started again
afterwards, and during the setting up, the loom was idle, to the extent that
many foremen who were capable of doing fashioned goods preferred to
produce plain fabrics. Besides, a gender difference in tasks appeared; the
weaving of light plain cloth, the lowest paid, was gradually given over to
the women workers, often daughters or wives of the foremen, and whose
pay was often half that of the men.25
As one of them pointed out, for the weavers this imbalance also fluctu-
ated with the general economic situation:
If there are only a few looms available [the merchant-manufacturer] pays a
salary for the most expensive articles, but on the other hand, if there were a
lot of idle looms he pays less because he can choose from a large number of

25 Reybaud, Etudes, 374–384; Villermé, Tableau, 168–169.

 contractual relations, tariffs and customs 157

looms in different workshops, so the workshops compete to hire out looms

at ever lower rates. It is the total application of supply and demand.26
This situation was very unpopular with the weavers, even more so as
they considered it an attack on their dignity. Their demands for equality
stemmed from the principles of the Revolution, reactivated by the 1830
and 1848 revolutions. But it was also nourished by the memory of the con-
flicts between master workers and merchants in previous centuries and
the fact that in the old community the workshop head had benefitted, on
an equal footing with the merchants, from the title of ‘master’.27 It was a
memory which brought with it nostalgia for a mythical golden age of the
‘Fabrique’, of a community of small independent producers crushed by
the deceit of the merchants.
To give a framework to contractual relations and reduce the social imbal-
ance they suffered from, the weavers formulated two major demands: the
setting up of a ‘minimum tariff’ which would limit the fall in the payment
for fashioned goods during slack periods, and the establishment of regula-
tions to put an end to the injustices they claimed to be victims of.

Reducing the Social Imbalance: Tariffs and Information

The weavers justified the setting up of tariffs by the social inequality they
claimed they suffered from. It was true that this imbalance did not guaran-
tee freedom of negotiation between the parties. The weavers put forward
the idea that conventions between two parties are legal only on condition
that they are free. So it was up to the authorities to step in to restore this
freedom by restoring social balance. The tariff was to stop salaries from
going below the workers’ real needs.28 Through collective organization,
better communication of information, often difficult to access for weavers

26 Benoît, Mémoires, 44.

27 The booklet of Joseph Masson-Sibut, a workshop head and member of the founding
group of the cooperative movement at the origin of the mobilisation of the silk workers
at the beginning of the 1830s, bears witness to the transmission of this memory (Mois-
sonnier, Maurice. La révolte des canuts. Paris: Editions sociales, 1975, 175). See also EF, 40
(1832), 2–3.
28 EF, 39 (1832): 2. This thesis, which had already been developed before the Revolution,
was taken up again in 1860: “The freedom of convention [. . .] only exists under conditions
of equality; it does not exist and will never exist between the worker, waiting for his sal-
ary in order to live, and the merchant, speculating on his capital”. (ACC Lyon, documents
of the ‘Condition des Soies’, petition submitted to the Emperor during his visit to Lyon,
1860).
158 pierre vernus

cut off in their workshops, would also contribute to correcting the imbal-
ance between workshop heads and merchants.29
The demand for tariffs by the Lyon workshop heads since the last
quarter of the eighteenth century remained central throughout the fol-
lowing century. It was closely linked to the payment for transforming
silk, which remained in force both for the urban looms and those in the
country and later in the weaving factories which flourished in the 1880s
and 1890s. In Lyon itself the demands for tariffs were the source of many
work stoppages in other areas of work, particularly in the hat industry and
the building trade.30 We also know that they were the source of numer-
ous conflicts throughout France and continued with the large number of
wage demands in the strikes during the last 30 years of the nineteenth
century.31
In the Lyon ‘Fabrique’ the tariffs which were demanded and sometimes
granted were for a ‘minimum wage’ for the difficult periods and aimed at
avoiding the drop in wages for transforming raw silk, particularly in plain
fabrics where price competition was rife, plain goods and some popular
novelties. These tariffs were a point of reference. They were negotiated
more or less collectively. Interpersonal bargaining between the workshop
head and the merchant could also use it as a reference.32
The tariffs were generally a list of variable length of fabrics grouped
in families (plain, figured, velvets, handkerchiefs, shawls etc.) subdivided
into types (for the plains: (taffetas, satins etc.) and subtypes. Each fabric
was defined by a certain number of technical characteristics. For example,
for the plains, the width of the ells, the number of threads which made
up this width, the type of chain and the quality of the weft were taken
into account. For fashioned goods it was the number of repetitions of the
design in the width of the fabric or the number of wefts which followed
each other regularly (the lats), the nature of the threads, an estimation
of the number of passages of the woof threads (the coups) per ell. The
amount paid per length or per article (e.g. for handkerchiefs).
The aim of these tariffs then was to work out a minimum payment
for the workshop head. From a ‘conventionalist’ point of view, we would
speak of a ‘formal investment’33 aiming at standardising the criteria for

29 Godart, Ouvrier, 247–271.

30 Rude, Insurrection; Festy, Le mouvement.
31 Bourgin and Bourgin, Le Régime; Perrot, Les ouvriers.
32 Cottereau, “Désincorporation”.
33 Thévenot, “Les investissements de forme”, 21–71.
 contractual relations, tariffs and customs 159

paying the weaver and reducing the uncertainties about the definition
of the product. In fact the wage wasn’t paid for a certain length of time
but for a certain length of fabric. Before agreeing on the amount, it was
necessary to agree on the nature of the object. But because Lyon’s textile
production was so varied, and that this meant many degrees of difficulty
in the production of goods, it was essential to fix representative indica-
tors of these difficulties, or at least part of them (Length, width, number
of prints etc.).
The question of tariffs shows how the positions of the groups involved
changed, and indicates that the variety of interpretations of the new post-
revolutionary legal corpus, which could be seen in the central administra-
tion itself, opened up possibilities which were only closed at the beginning
of the 1830s.
Prolonging a series which opened with the Revolution, several tariffs
were applied, at least momentarily, under the Empire and the Restora-
tion.34 However, contrary to the process of wide consultation and collec-
tive formulation carried out by the workers in 1789, the tariffs of the first
20 years of the nineteenth century were rather the fruit of deliberations
by the notables. For example, in 1811, the “tariff additional to the minimum
payment for fashioning of fashioned silk goods”, which followed a tariff
established in 1802 concerning plain fabrics,35 was passed during a meet-
ing of the Chamber of Commerce, sitting as the consultative chamber of
the chamber of arts and manufacture, in the presence of the municipal
authorities, members of the ‘conseil de prud’hommes’, three important
merchants and the ‘préfet’ who chaired the meeting. When the tariff was
passed it immediately became a prefectural decree.
In December 1817 in a petition, rich merchants denounced their col-
leagues who “took advantage of the difficult situation of the workshop
heads” and demanded the reinstatement of the 1811 tariff.36 A municipal

34 Cottereau, “Désincorporation”; Cottereau, “Fate”.

35 It had been adopted on 10 frimaire year XI (1st December 1802) during a session of
the ‘conseil des prud’hommes’ in the presence of ‘préfet’. A petition by merchants of plain
fabrics, undated but probably written at the end of 1806 or the beginning of 1807, asked
the council to “have the bill renewed and publicised so that no one could plead igno-
rance” (AM Lyon, 784, Wp6, tariffs for transformation of rolls of silk). It was the ‘conseil
des prud’hommes’ which had asked that the tariff in use for plain fabrics should be sup-
plemented. (Archives départementales (AD) du Rhône, U 134, conseil des prud’hommes
Lyon. Soieries. Président, Vice-président nomination ou élection 1822–1858, letter from the
chamber of commerce to the ‘préfet’ of, 22 June 1811).
36 A similar approach had been used at the end of 1806 or the beginning of 1807 con-
cerning the 1802 tariff.
160 pierre vernus

bye-law dated 6 February 1818 “modifying the tariff for fashioning silk
goods” of 18 June 1811 “satisfied their wishes.”37
So the influential merchant-manufacturers accepted tariffs. It is true
that the 1817 petitioners underlined that the price for fashioned goods
fixed by the 1811 tariff gave the worker just enough to renew his capac-
ity to work.38 Moreover these prices could be lowered.39 So it was not a
question of stopping the variations in price due to the law of supply and
demand, but giving it a framework. However, beyond any charitable con-
siderations, these measures were also concerned with regulating compe-
tition among the merchant-manufacturers themselves. According to the
signatories of the petition, the payment of salaries below the tariff rep-
resented unfair competition towards those who respected the tariff and,
they wrote, “If these differences in price were not curbed it would have
fatal results for our town, since the prices fixed by these tariffs, which
have not been abrogated, leave the worker below the poverty line, and by
lowering them we would be forcing them to leave the straight and narrow,
to beg or to go abroad.”40
For them, the functioning, the cohesion even the survival of the ‘Fab-
rique’ were at stake. Faced with misery, the workers could be tempted
to steal part of the raw materials they had received to bolster the paral-
lel economy (which was called ‘stealing an ounce’). This would affect the
cost price of materials negatively, or worse; if they set up shop abroad,
they would transfer their know-how and reinforce the foreign silk centers
whose competition was a major source of worry for the local professionals
and the local authorities.
On their side these authorities and the ‘préfet’ guaranteed and ratified
these documents while occasionally wondering about their legality. Thus,
in 1811 the chamber of commerce recognized that no law explicitly author-
ized tariffs, only to underline immediately that a ruling from the Royal
Council on 5 September 1789 had authorized:

37 AM Lyon, 1 I 11, arrêtés, ordonnances, règlements.

38 AD Rhône U 134, ‘conseil des prud’hommes’ Lyon. Silks, president, vice-president
nomination or election 1822–1858, letter from the chamber of commerce to the ‘préfet’,
22 June 1811.
39 In 1818, at the request of merchant-manufacturers in gauzes, crêpes and plain cut
velvets, a cut in prices for the transformation of these articles was proposed by the ‘conseil
des prud’hommes’ “given the fact that they [the merchants] found it impossible to with-
stand the competition from the large number of ‘factories’ of this type which have been set
up in the neighbouring districts where manpower is cheaper” (AM Lyon, 1 I 11, municipal
ordinance, 10 february 1810).
40 AM Lyon, 784 Wp 6, petition from merchant-manufacturers to the president of the
Lyon ‘conseil des prud’hommes’, 11 December 1817.
 contractual relations, tariffs and customs 161

“the setting up of a mixed commission of master workers and master mer-

chants with a brief to provisionally fix the prices paid for fashioned goods”,
and that “as soon as order had been restored in the Lyon silk industry by the
‘conseil des prud’hommes’, the Council wisely renewed its use, to which all
the honest merchants quickly agreed. Today these tariffs can be considered
as family law which constrains, by the pressure of opinion, those who would
be tempted to speculate on the distress of the worker. They can also be used
as a guideline for the arbitrators in ruling on any disputes which may arise
between the master and the worker over the price for fashioned goods.”41
In 1819 the question of the legal status of these tariffs was raised once
more following a conflict not in the silk ‘fabrique’ but in the hat industry.
It seemed at that time that the doctrine of the Ministry of Home Affairs,
which had been consulted several times, was ambiguous and that it could,
under certain conditions, tolerate the proclamation of a tariff.42 However,
the legality of tariffs appeared less and less clear to the administrative
authorities and during the 1820s no new tariff was ratified or modernized
in the silk trade even if, in 1822, the municipality of Lyon passed a statute
“concerning the fixing of a minimum wage in the hat industry.”43
The events of November 1831 clearly marked the end of the authorities’
tolerance of tariffs. Even if the uprising resulted from the indignation pro-
voked by the attitude of a majority of merchants who rejected the signed
agreement, it is useful to recall that the government encouraged them in
their refusal.
The 1830 revolution and the change in the régime reinforced the demand
for equality, freedom and dignity on the part of the workshop heads, who
had taken an active part in the revolt in Lyon and encouraged their collective
mobilization.44 Resuming the process of 1789, the workshop heads engaged
in a process of collectively drawing up a project for tariffs which finally served
as a basis for bipartite negotiations under the aegis of the ‘préfet’.
A large number of merchants refused to agree to the project and balked
at applying the tariff which was finally signed. Some of them laid the
matter before the government which overruled the ‘préfet’ and forced
him to point out that the tariff was only a “declaration on honour”
with no legal status.45 Instead of tariffs the new ‘préfet’ proposed that the

41 AD Rhône, U 134, letter from the chamber of commerce to the ‘préfet’, 22 juin 1811.
42 Rude, Insurection, 78–79, 82–85, 99–101, 167.
43 AM Lyon, 1 I 8, municipal ordinance, 24 october 1822.
44 “Those who, from pride or sensitivity of character [the worker] considers with arro-
gant disdain, will go back on their error and admit that the class they have so long despised
is worthy of their respect and deserves to see its fate improved” (EF, 12 (1832), 2.
45 Rude, Insurrection.
162 pierre vernus

‘conseil des prud’hommes’ should establish a market price list of prices,

in other words a statement giving the average prices for fashioned goods
at a given moment for the different categories of fabric. When it became
clear that to the workers’ minds this statement was the same as a tariff,
the merchants on the ‘conseil des prud’hommes’ manoeuvred to block it
and erase it from memory.46
From then on, in the name of the freedom of legally equal individu-
als to establish agreements, the authorities considered any tariff as a rule
which at best had no legal value and at worst was illegal. What is more,
in 1840 during the drawing up of Lyon’s municipal code, all the bye-laws
and ordinances which had allowed tariffs were declared null and void and
were discarded as “contrary to free trade.”47 This meant that in the case of
collectively negotiated agreements between weavers and merchants, the
local and national authorities refused to guarantee them and did not force
recalcitrant merchants to apply them. The ability of the workshop heads
to have this type of agreement respected depended on their degree of
mobilization and organization. But from 1832, and more so from 1834 until
the last years of the Second Empire, (with the exception of a gap under
the second republic), the workers’ movement suffered from repressive
policies of varying intensity but which overall reduced the silk workers’
ability to organise and protest collectively. It led some of them to become
resigned and led others to adapt their forms of resistance and the types
of their demands.
After 1843, police pressure on associations relaxed a little, so the weav-
ers formed groups corresponding to the specialized goods.
It was soon clear that these groups served as cover for fighting funds
when, the following year, four of them revived the practice of blacklisting
the merchants who paid below a certain price for fashioned articles. But
this kind of action was limited and was virtually repressed.48
The liberalization of legislation on the forming of groups during the
1860s (law of 25 May 1864), then administrative recognition of profes-
sional associations from 1868, prolonged under the third republic (law of
21 March 1884) allowed the weavers to organize collectively once more.
The groupings were once more according to the categories of the articles.
In 1869, when the raising of the prices for transforming the silk lagged

46 EF, 39 (1832), 2–3 and 50 (1832), 3.

47 AM Lyon, 1 I 5 to 9.
48 McDougall, Lynn. The Artisan Republic, 14.
 contractual relations, tariffs and customs 163

behind the renewal of industrial activity there was a vast mobilization in

favour of a tariff. Groups of weavers drew up tariffs by category of article
which were then submitted to the merchants concerned. These prices
were accepted but, as one merchant-manufacturer wrote, it still had to
be put into practice.
Concerning the tariff, I haven’t changed my mind since you gave it to me.
It is in your interest to keep it because if it didn’t exist, competition would
turn to reducing salaries, and in periods of trade crisis we don’t know how
far the reductions could go. The tariff puts all the manufacturers (i.e. the
merchants) on the same footing and no one can do deals which harm wages.
But to put it into practice, you have to be capable of getting it done.
In order to do so the “Provident and Information Society for the work
of the weavers in the Lyon ‘Fabrique’” was founded. It was made up of
14000 members in several hundred cells, grouped by specialty. The work
of the cells was to get work for its members and compensate them if they
moved. In fact, in their big meetings the weavers had refused the strat-
egy of the general strike in favour of that of ‘displacement’ which con-
sisted in refusing the orders of merchants who did not respect the tariff.
If that happened, the Provident society gave the weavers the names of
merchants who were looking for looms to be worked, and who respected
the tariff, or it paid them compensation. In order to do this more effi-
ciently the Providence society opened an information bureau. At the same
time, in April and May 1870, delegations were sent to contact the weavers
in the countryside, who had stopped work, to get a rise in the price of
transforming. After the gap of the Franco-German war this system worked
until the Society was dissolved by the authorities in 1874 during the ‘Moral
Order’ period.49 This strategy could only be efficient if it was set up by an
effective and representative organization with the go-ahead, or at least
the benevolent neutrality, of the authorities and in favourable economic
conditions.
But although the organization reopened in 1876 under the name ‘The
Weavers Federation’ the situation changed profoundly from the follow-
ing year. The crisis of 1877 heralded a period of great change. The shift in
demand towards light fabrics to the detriment of the beautiful rich fabrics
which had made the name of the urban silk industry, and the acceleration
of mechanization, spelt the inevitable and rapid decline of Lyon’s urban
weaving. In 1889 there were only about 16000 looms, half the number in the

49 Mathé, aîné. Les tisseurs: 190; Maritch, Histoire, 226–228.

164 pierre vernus

1855–1875 period.50 The change in the economic situation brought about

a fall in the prices of fashioned goods and the conditions of the weavers
compared with that of the workers in the other local industries; the crisis
of recruitment which had already been apparent, got worse. The average
number of apprentices dropped and virtually disappeared in the 1890s.
The workshop was reduced to the ageing household of the weaver.51 From
then on, as the last great movement in 1885 showed, the workshop heads
were no longer in a position to impose respect for the prices which had
been negotiated, even if, although divided into two rival organizations,
they managed to bring the merchants to discuss a rise in prices (in fact
an arrangement below the tariff of 1869). But their negotiating partners
manoeuvred to stop the list of prices, which had been settled, from looking
like a commitment signed from organization to organization which could
later be considered as ‘local practice’.52 The only thing that remained of
the agreement was a list of ‘recommended prices’. The organization of a
huge demonstration to try to make the forty or so recalcitrant merchants
to give in, then the attempt to have the prices guaranteed by the ‘conseil
des prud’hommes’, and finally, the demand for legislative sanctions, all
failed. These movements revived the memories of the revolts of 1831 and
1834 and accelerated the putting-out of production outside Lyon, and with
this accelerated the inexorable decline of urban workshop weaving.53

‘Mutualism’ and ‘Regeneration’ of the ‘Fabrique’

‘Mutual duty’ or ‘mutualism’ started in 182854 and aimed at being a tool

for moral regeneration, aiming at opening the way for reforming abuses.
At the outset this moral regeneration targeted the workshop heads group:
“Walk with heads high against other abuses.”55 It was a secret society,
divided into lodges of twenty members, with mutual links and was marked
by the restrictions imposed by article 291 of the penal code. It was inspired
by freemasonry and trade guilds, only recruited workshop heads, and had
three aims; ‘information, help and assistance’, which meant encouraging

50 Lequin, Les ouvriers, 76–77, 82–91.

51 Cayez, Crises, 55–62.
52 The Association of the Lyon ‘Fabrique’, an employers’ association of merchant-
manufacturers of silk goods had been founded in 1868.
53 Vernus, “ Pour une histoire”.
54 On the twists and turns of its beginnings, see Rude, Insurrection.
55 EF 1841 49 (1843), 2–3.
 contractual relations, tariffs and customs 165

the spread of information useful to the members, favouring mutual aid

by the loan of materials and in difficult moments, helping members
financially or morally, particularly for funerals.56 It started with just a
few dozen members, but numbers grew rapidly from 1831. Mutualism was
one of the numerous initiatives which the idea of association spawned
from the 1820s.57 Its adherents saw in it a response to the atomization of
the group of workshop heads, and more broadly to the forces of dissolu-
tion, in the front line of which was the selfish pursuit of personal gain,
which threatened the industry as a community whose main victims, it
seemed to them, were the silk workers. The more or less implicit ideal of
this movement of workshop foremen was of a ‘Fabrique’ inside which the
varied interest of the constituent groups could find a balance thanks to
the action of joint commissions.
Besides, the notion of ‘information’ which was fundamental for the
founders of mutualism, was a real part of their project for regenerating
the ‘Fabrique’.58 For them, the association should favour the circulation of
information in the workshop heads’ milieu.59 It was one of the functions
which the ‘Echo de la Fabrique’ gave itself. The journals which followed it
carried on the same line. This is clearly visible in the offices of the ‘Echo de
la Fabrique’ which set up a ‘special information bureau for the silk materi-
als industry’ in Lyon.
[. . .] The artisans and those who provide them with work are completely
isolated from each other. The social position of one group and often the
destitution of the other, impedes the contact without which the industry
cannot prosper. The workshop heads and the workers experience the same
isolation. From this innumerable problems arise when it comes to mount-
ing new articles, and the enormous overheads which often overwhelm the
workshop heads. They can be avoided or at least partly diminished, if every-
one could turn to a common centre.
The benefits that the ‘special information bureau for the production of
silk fabrics’ would provide are innumerable; there, the workshop heads
will find either workers and apprentices, or all the tools, harnesses and

56 Blanc, Histoire, 516–523.

57 Rosanvallon, Le modèle, 84–189; Sewell. Gens de métier, 274–281.
58 28 June 1828, the date of the durablefounding of Mutualism, was celebrated as the
beginning of “Year 1 of regeneration” (Rude, Insurrection, 141).
59 The preamble of the rules for mutualism declared: “The aim of mutualism is thus
between the founders and all those who will be accepted as brothers: 1° to inform each
other, frankly and loyally, mutually and generally, of everything that may be useful and
necessary for them in their professions; 2° to help each other by the loan of equipment as
much as is in their power (. . .)” (Blanc, Histoire, 516).
166 pierre vernus

accessories for setting up looms according to the various articles; he will

also find the means to get work without going all over the place knocking
on the doors of the different establishments. The workers will be less afraid
of being laid off because the special information bureau being the centre
where all the requests for workers end up, they will know where to go and
will not vegetate, going from one workshop to another looking for an avail-
able loom. But it is in the bad times that the information bureau will be
most useful, because that is when there is less work and as a result, the
worker is more likely to change workshops.
The merchants will be able to get the number of workers they need for
an order more easily. What we suggest has been tested by some commercial
houses for whom we supplied the number of workers they needed in two
days.
So, in the interests of commerce and industry, a special information
bureau for the production of silk goods will be installed from today in the
offices of the ‘Echo de la Fabrique’. We will take care of I° the merchants’
requests for looms; 2°, the placing of workers in the different workshops
according to the articles; 3° requests for placing apprentices; 4° the sale of
looms, harnesses and accessories for all kinds of production, and finally all
requests concerning the ‘Fabrique’.60
By reducing the cost of access to information and by guaranteeing bet-
ter adjustment of the internal needs and resources in the ‘Fabrique’, the
information bureau was supposed to reduce waste of time, to ensure the
quality of production and in fine, to improve the global functioning of
the ‘Fabrique’ which would thus be in a position to face up to foreign
competition.
The improvement in the circulation of information should also reduce
the imbalance penalizing the workshop head in relation to the merchant
in the contractual relationship, and partly contribute to balancing the
positions of the two parties. This argument was clearly expressed in a peti-
tion launched in 1840 by the workshop foremen and the workers.
“The merchant-manufacturer is informed of everything that happens
among his colleagues; he knows the sales price of the fabrics, the time he
spends gathering information is well used and produces its share in
the profits”. On the contrary, for the worker, “every hour spent outside the
workshop is time lost. Badly informed and driven by need, he must accept
the conditions he is offered”.61

60 EF 18 (1832), 7–8.

61 Pétition.
 contractual relations, tariffs and customs 167

Fighting Abuses

Even if the workers saw tariffs as a way to compensate for their weak
position by framing their contractual relations with the merchant-manu-
facturers, they could not reduce the abuses the merchant-manufacturers,
or at least some of them, were accused of. In fact, fixing the price for
transforming silk, covered by a tariff or not, only covered one aspect of the
contractual relations linking the workshop head and the merchant. They
included other parameters (amount of waste, overheads for setting up the
looms, quality of the material, etc.) which affected the workshop head’s
pay and conditions. For Pierre Charnier,62 these abuses stemmed from
free competition and were the major cause of the Fabrique’s difficulties.63
Following Alain Cottereau, we should underline that although some
of the decisions taken by the ‘conseil des prud’hommes’ in its general
assemblies, or some of its judgments, set up a local quasi-legislation which
sometimes revived certain elements of statutes from the eighteenth cen-
tury, these practices were also continually submitted to criticism, more or
less intensely depending on the period.64 When these attacks came from
the merchant-manufacturers who could act collectively, they were consid-
ered by the weavers or their spokespersons to be abusive. At the end of
the 1820s this question was particularly acute when, during the difficulties
at the end of the 1820s, some merchants tried to free themselves from the
traditional practices, especially as they could appear to be unsuited to a
form of production which was constantly being renewed.
In a series of articles in 1832, Joachim Falconnet65 denounced the most
glaring of these abuses. The first one dealt with surcharges or modifica-
tions put on the workers’ account booklets by certain merchants or their
assistants, taking advantage of the fact that the rule stipulating that the
noting of all elements should be written by both parties on the accounts
booklet of the other, was not respected. In fact it was quite common for
the workers to leave their booklets with the merchant and pick them up

62 Workshop head born in Lyon in 1795, son of a grocer; founded the ‘Mutual society for
supervision and information’ in September 1827. Member of the ‘conseil des prud’hommes’
1832–1852.
63 Rude, “L’insurrection”; La Révolution de 1848 164(1938), 18–49, 165 (1938), 65–117, 166
(1938), 140–179.
64 Cottereau, Fate, 51–55.
65 Born in Lyon September 1798, son and grandson of workshop heads and himself a
workshop head, founder and manager of L’Echo de la Fabrique. He was a member of the
‘conseil des prud’hommes’ 1832–1858.
168 pierre vernus

only a week or two later, sometimes more. This period could be used to
modify the written conventions in the booklet. Added to this the work-
shop heads sometimes failed, deliberately or not, to state the balance of
accounts, which left him in a doubtful position towards the merchant.
Moreover, some merchants only paid their workers on a fixed day and not
on the day the roll of material was delivered. This practice favoured the
merchant’s bank balance but weighed on the weaver’s who often had to
down tools and go once more to the merchant’s office and leave his work
for that time. It was particularly unpopular because the merchants on their
side, if there was a delay in delivering the rolls of material, would charge
a penalty on the price of transforming. Another big cause for recrimina-
tion by the weaver was over the percentage of wastage accepted by the
merchant during the making of fabrics.
In fact some merchants, mainly for new articles, tried to question the
practices which had been in place even before the Revolution, which gave
a level of wastage of 3%, and recognised that the selvage depended on the
weaver.66 The setting up of the looms was also a recurring source of con-
flict. For fashioned goods the preparation work for weaving took a long
time, several days or even several weeks, and this imposed heavy costs
on the workshop head. On the other hand, the manufacturer generally
agreed to order a minimum number of rolls to be woven. But this practice
was also questioned. Some merchants who had cancelled part of their
orders, refused to compensate the worker for his overheads. Others, some-
times the same, attempted to make the weavers pay the cost of repairing
the lacing of the cards used with Jacquard mechanisms, or even that of
correcting mistakes in the piercing of holes, although the cards belonged
to the merchants.67
The workshop heads demanded the establishment of stable regulations
to counteract the abuses. There were two parts to this demand. First of all
they wanted to keep the old customs against the attempts by certain mer-
chants to question them, or more precisely against their practices which,
by creating precedents, threatened to nullify the practices in place. So
then it was a question of defining new practices for the new fabrics.
At the beginning of the 1830s the ‘conseil des prud’hommes’ found it-
self at the heart of the weavers’ efforts to fight against the abuses. The

66 Selvage, or edge, the weaver had to make to adapt the fabric to the roller of the
loom.
67 EF 15 (1832): 2, 16 (1832), 3, 17 (1832), 2–3, 18 (1832), 2–3, 19 (1832), 2–3, 22 (1832), 3–4,
23 (1832), 2–3.
 contractual relations, tariffs and customs 169

tribunal seemed, to the weavers, to be their main means of redress, espe-

cially since they had made progress on two points they had criticized. In
February 1831 a petition criticized the fact that the general bureau’s meet-
ings were held in camera. They called it ‘an occult bureau [. . .] favourable
to the merchant-manufacturers’.68 The following month the meetings
were publicized and, despite some attempts to question it, this lasted till
the end of the second Republic.
The weekly sessions of the general bureau attracted many people. The
weekly minutes were published in the ‘Echo de la Fabrique’ and in the
similar journals which followed it. Another criticism was of the numeri-
cal majority of the merchant-manufacturers on the board. Following the
revolt in 1831, the government’s only concession was on the organization
of the ‘conseil des prud’hommes’. A statute of 1832 enlarged the electoral
college of workshop heads and increased the number of ‘conseil’ members
from the silk working industry.69
So, after the rejection of the 1831 tariff, the workshop heads put their
hopes in the new reformed ‘conseil’. They were very active in the April
1832 elections, hoping that the institution would define regulations out-
lawing the worst of the abuses. But, although during the first few months
of the new ‘conseil’s activity the path of collaboration and conciliation of
the interests of the workshop heads and the merchants seemed possible
and allowed the adoption of decisions favorable to the weavers, tensions
appeared as early as the last few months of 1832. It was particularly the
case in the 1833 elections, where merchant-manufacturers were elected
who were not disposed to grant concessions and who sought to block the
working of the ‘conseil des prud’hommes’.70
The relations between the two major elements of the institution finally
calmed down and the ‘conseil des prud’hommes’ was able to continue its
work. Its activity reached its high point in the middle of the 1840s with
more than 6000 cases heard in privy session. This number dropped later
because of the fall in the number of urban workshops due to the develop-
ment of rural weaving. In the first half of the 1840s about 40% of cases were
between merchants and workshop heads. It is worth noting the stability

68 Rude, Insurrection, 231–241.

69 It still did not establish equality of members since the merchant-manufacturers went
up from 4 to 9 and the workshop heads from 3 to 8, which meant, for the council as a
whole, 14 merchant-manufacturers against 11 workshop heads instead of 9 and 6. More-
over, only the heads of workshops who owned more than 4 looms had voting rights, only
20–25% of the total.
70 Bezucha, “Aspects”.
170 pierre vernus

of a certain number of ‘conseil’ members, especially among workshop

heads where some members from the 1832–1834 period continued to sit
during the 1840s and even till 1852 for J. Falconnet and 1858 for Pierre
Charnier. The questioning of certain customs continued to bring about
moments of tension, like in 1843 when a group of merchants contested
once more the assigning of selvage, or in 1860 about ‘peignes’ (cards),71 the
percentage of waste and the setting up of looms.
As early as 1831 the weavers asked the ‘conseil des prud’homes’ to set up
a ‘fixed jurisprudence’, which, being written down, would stabilize their
acquired positions or confirm the older customs, and would play the role
of regulations for the ‘Fabrique’ It was from this standpoint that Marius
Chastaing72 inaugurated the publishing of the minutes of the meetings of
the bureau of the ‘conseil’ in the first edition of the ‘Echo de la Fabrique’ in
October 1831. One year later he presented them in the form of ‘notices of
jurisprudence’, in which precedents were presented before the final deci-
sion. Then, in the first four months of 1833, he managed to persuade the
‘conseil’ to have its decisions systematically noted as minutes by the clerk,
whereas this was not the case before (at least, it would seem, before Octo-
ber 1832) unless it was requested immediately. At the same time Chastaing
started to collect the ‘conseil’s decisions “so that the accused can use them
if necessary and mention them as precedents. That would partly make
up for the lack of a fixed jurisprudence, which we would have liked, and
which we continue to demand and which the ‘conseil des prud’hommes’
has a duty to set up”’.73 He took up this work again, with a short break in
July 1835, in ‘La Tribune Populaire’. His ‘notices of jurisprudence’ were col-
lected in 4 then 5 series74 and were aimed at serving as a basis for a ‘code
of the council’ or of ‘the silk workers’.

71 A certain length of chain which could not be woven at the end of each roll.
72 A journalist, Chastaing was the son of a bailiff who had himself followed law studies
(Popkin, Jeremy D. Press, revolution, and social identities in France, 1830–1835. University
Park: Pennsylvania state university press, 2002, 143).
73 EF 2 (1833), 12; 4 (1833), 1–2 and 9 (1833), 69–70.
74 When, from October 1832 Chastaing started to collect the council’s decisions and
to publish them in ‘L’Echo de la Fabrique’, he grouped them in 4 categories (which he
called series): first series; the workshop heads in their relation with the merchants; second
series; the workshop heads in their relation with the workers or journeymen; third series;
the workshop foremen in their relation with pupils and apprentices; fourth series; the
workshop heads with various other persons. Later Chastaing stopped working for ‘L’Echo
de la Fabrique’ in 1833. He took up publishing his ‘notes of jurisprudence’ of the Lyon
‘conseil des prud’hommes’ in a new paper, the Tribune prolétaire in July 1835. The fourth
series was then called: Workshop heads in their relations among themselves or with vari-
ous people and different questions concerning skills and others. He created a fifth series
 contractual relations, tariffs and customs 171

In 1843 in his ‘Information directory for the Lyon Silk Industry’ J. Falcon-
net carried on Chastaing’s work, re-publishing his ‘notices’ and adding his
personal notes. For him it was also the first draft of what should be a fixed
jurisprudence for the industry, like those which already existed for civil
and commercial affairs Despite moments of tension around the question
of wastage it seems that there was little change on it during the 1840s.In a
new edition of his work in 1849–1850, Falconnet, who, as a councilor in the
‘conseil des prud’hommes’ was well placed to judge, thought it unneces-
sary to reprint the ‘notices’ since the ‘jurisprudence of the council has not
changed much despite some changes’.75
However, the idea of an industrial code continued to progress. In 1860 a
petition submitted to Napoleon III with 561 signatories, of whom the first
were seven newly-elected worker counselors, took up the slogans both
of an ‘industrial code’ (for the silk industry), which would stabilize prac-
tices, and also of fixing the correspondence with the commercial code.
This petition was not heard, but from the end of the 1860s the merchant-
manufacturers, worried by the rise in collective action by the workshop
heads and keen to put a brake on what seemed to them to be ‘abuses
[which] infiltrate[d] the’ conseil’s decisions were attracted by this desire
for codification,76 and put it into practice by publishing ‘Practices of the
industrial tribunal of the town of Lyon for the silk industries’ which was
widely read by the merchant manufacturers. This collection of articles
dealt successively with the relations between the merchant-manufacturers
and the workshop heads, between the latter and their workers and appren-
tices, this group making up the largest part of the book, and then the
relations of the manufacturers and workshop heads with the other profes-
sions which took part in the transformation of silk and the production of
fabrics (reelers, warpers, throwers, dyers). So the structure of the work was
similar to that of Chastaing then Falconnet. Afterwards, the modifications
of the customs were negotiated between the merchant-manufacturers
employers’ federation and the workshop foremen, like in 1885. A new
modified version of these works was published in 1891.77

entitled: Merchants and their relations with each other or with people other than work-
shop heads.
75 Falconnet, Indicateur, 1843 and 1849.
76 Pariset, Histoire, 328; Enquête, 371.
77 Usages du conseil; Recueil des usages.
172 pierre vernus

Conclusion

Behind the apparent stability of the organizational framework of the Lyon

silk ‘Fabrique’, the relations between merchant-manufacturers and work-
shop heads as well as their relative positions within the framework of these
relations, varied greatly. Between 1744 and 1786, the master weavers had
noted the domination of the merchant-manufacturers over the commu-
nity imposed by the 1744 statute. The abandoned their claim for a return
to equality in the management of the community and reoriented their
demands, calling for the setting up of a compulsory general minimum
tariff designed to put a frame round the merchants’ economic power.78
The Revolution disturbed the projects for restructuring the ‘Fabrique’,
which, like the one presented by the intendant Terray after the 1786 upris-
ing, aimed to cancel the intermediary status of the master workers and
reduce their status to that of the journeymen. In fact, despite the disap-
pearance of the community of the Ancien Régime, on the legal plane the
special position of the workshop heads was confirmed by the law of 1806
which created the Lyon ‘conseil des prud’hommes’. Moreover this institu-
tion revived, although only partially, the demands for joint management
of the ‘Fabrique’ by the merchants and workshop heads. Added to this, the
political situation, the new principles for organizing society, in particular
equality and the prevalent lack of manpower during the first years of the
nineteenth century reinforced the position of the weavers towards the
merchants.
However, at first, contract law which framed the contractual relations
between merchants and workshop heads left open the debate about the
meaning of ‘interpersonal negotiation’. The interpretation given by the
weavers’ spokespeople, according to which contractual relations could
only be freely established if the positional imbalance between the two
parties was balanced by barriers limiting the power of the dominant party,
could still be heard at the local level up until the 1820s. It was even sup-
ported by the most influential, if not the majority, of the representatives
of the merchant-manufacturers, who saw it as a way of regulating com-
petition inside their group and of avoiding that behavior based on purely
individual interest should lead in the long term to the disintegration of
the ‘Fabrique’ itself. As well as that it was tolerated if not supported by
local and government authorities. But the progress of liberal economic

78 Garden, Lyon, 572–582.

 contractual relations, tariffs and customs 173

ideology led the weavers’ idea of mutual agreement to become less and
less acceptable both by the merchants and the authorities. The authori-
ties henceforth refused to approve any tariff or even nullified any such
agreements. This was the cause of the 1831 events. Having said that, the
confrontation between merchants and weavers was not only around tar-
iffs; it also concerned practices aimed at clearing up certain aspects of
contractual relations of work which caused uncertainty and sometimes
greatly influenced the wages earned.
Several parameters contributed to fashioning the strategies and limit-
ing the scope of means available to the different parties to push forward
their demands. First of all the institutional, legal and also political frame-
works defined the formal space and the types of action the workshop
heads could use. The variations in this framework opened or closed room
for mediation or negotiation, helped or hindered compromises and alli-
ances and influenced their form and content. They allowed or prohibited
certain strategies of the parties and conditioned their ability to attract
political support. Moreover, the state of the job market and the direction
of demand constituted parameters which weighed on the workers’ ability
to impose a certain amount of control over the rules governing their pay
as well as the intensity of their work, and to balance their position in the
process of collective or individual negotiation which led to a contractual
relationship. The use by the merchant-manufacturers of rural manpower,
then from the 1880s, the acceleration of the mechanization of weaving
encouraged by the direction of the markets, in the long term weakened
the urban weavers position by reducing their ability to create a balance of
power which could improve their position relative to the merchants.
 The Circulation of Commercial Manpower in an
Indian Worldwide Trading Network in the
Early Twentieth Century

Claude Markovits

The existence of a significant worldwide circulation of commercial man-

power from India has been little noticed in the existing literature. In India
itself, those employed in shops and other commercial establishments are
largely absent from standard accounts of labour history, although they
probably represented and still represent altogether one of the largest seg-
ments of the overall labour force. Labour historians have tended to focus
their attention almost exclusively on agricultural and industrial labour,
to the detriment of a category that suffers from an ambiguous socio-
economic position, straddling as it does the boundary between lower
middle-class and working class, and has attracted little attention from
colonial rulers, resulting in a paucity of source materials. As regards the
circulation of Indian labour outside India, the existence of a sizeable move-
ment of commercial manpower is also generally overlooked by historians
of the South Asian diaspora. Their focus has been mostly on the move-
ment of indentured and kangani labour (two forms of contract labour)
that flourished between 1834 and the late 1930s, resulting in the emigra-
tion of millions of Indian workers towards the sugar-producing countries
of the Caribbean (Guyana, Trinidad, Surinam, Martinique, Guadeloupe)
and of the Indian Ocean (Mauritius, Reunion, South Africa), as well as
the tea plantations of Ceylon and the rubber plantations of Malaya. These
migrations are particularly well documented in the colonial archives, and
have given rise to a vast secondary literature.1 However, in addition to this
migration of workers strongly regulated by the Indian colonial govern-
ment in coordination with the imperial government (inasmuch as 90% of
the indentured workers and 100% of the kangani migrants made their way
to British colonies, less than 10% going to French and Dutch territories),
there has existed another movement, much smaller in volume but never-
theless significant, of traders and commercial employees. That particular

1 On this migration, see, amongst a vast and growing literature, two classics: Tinker A
New System and Northrup, Indentured Labour.
176 claude markovits

stream has origins that go back to the medieval and early modern period,
but was considerably amplified in the nineteenth century when British
India was integrated into the world capitalist economy and new trading
circuits were created linking India with the Far East (China in particu-
lar), while already existing circuits connecting India with the Middle East
and East Africa were reorganized and significantly expanded. It remain-
ing largely unrecognized in the literature has to do with the fact that the
colonial government of India did not organize it directly or even monitor
it closely, with the result that there are few official sources available.
In a previously published article,2 I used some little-known published
material to evaluate the importance of that movement, that belongs to the
universe of circulation rather than of permanent migration, and I came up
with a global figure of some 1.5 million men (it was an almost exclusively
male stream) for the period 1830–1950. I found that it was characterized
by a great diversity in the geographical origins of its participants, most of
whom however hailed from coastal regions of the subcontinent, in par-
ticular Gujarat on the West Coast and Tamilnadu on the South-eastern
coast. Surprisingly, I came across evidence that some inland areas and
towns also sent commercial migrants abroad on a significant scale. This
was the case in particular of the town of Hyderabad in the province of
Sindh (not to be confused with its better-known homonym, Hyderabad-
Deccan), that belonged to British India from 1843 to 1947 and was then
made a part of Pakistan. Although situated inland on the lower valley of
the Indus at a distance of 150kms from the seaport of Karachi, Hyderabad
was the cradle of a particularly dynamic trading network that, between
1860 and 1914, extended its operations practically to the entire world.

The Rise of a Worldwide Trading Network

I have given in a book3 a detailed historical survey of the rise and growth
of that network, which I shall here briefly summarize. The origins of the
network go back to around 1860, when some Hindu merchants from that
middle-sized town with a population of 40 000 (a majority of them being
Hindus, while Sindh as a whole was a Muslim-majority province) started
going from Bombay to Egypt on annual voyages. In Egypt, they sold a cer-
tain kind of craft goods produced by the (Muslim) artisans of Hyderabad

2 Markovits, “Indian Merchant”.

3 Markovits, The Global World.
 the circulation of commercial manpower 177

and some neighbouring Sindh towns, which had become known in Bom-
bay in the 1850s as “Sindwork”. Hence the appellation of “Sindworkies”
under which these traders were generally known. They benefited from
the fact that their arrival coincided with the beginnings of modern tour-
ism in Egypt in the form of the first tours organized by Thomas Cook.
Their customers were thus mostly affluent European and North American
tourists. In the following five or six decades, that network transformed
itself into a worldwide phenomenon. From Egypt, the Hyderabad mer-
chants gradually extended their trading operations to the entire Mediter-
ranean basin, and, from Gibraltar, penetrated into the Atlantic, reaching
West Africa, the Canary Islands, Central and Southern America, and even
making forays into North America. Other merchants of the same town,
departing from Calcutta and Madras, reached Southeast Asia, and Singa-
pore emerged as an important hub for them. They soon reached China
and Japan, where, from the 1870s onwards, they started buying silk goods
and various artisanal products that gradually replaced the productions of
the small workshops of their native province. By 1907, it was estimated by
a British official4 that there were approximately 5 000 so-called Sindwor-
kies dispersed across the world in locations as widely scattered as Kobe
in Japan and Panama in Central America. They specialized in the sale of
silk goods and curios to a mostly European and North American clientele
of affluent travellers and tourists, and had establishments at the various
ports where the great transoceanic steamers alighted. These Sindworkies
however were far from being a homogeneous group of merchants, as the
standard literature on “trading diasporas”5 would have us believe. They
actually formed a highly diversified community in which capitalists, small
and big, who were proprietors of family firms of varied size, salaried man-
agers, and different kinds of employees were represented. The core of the
network consisted of a certain number of family firms based in Hydera-
bad-Sindh, a hundred in all, of which five or six were really big (the largest
being the firm of Pohoomull Bros that, circa 1911, had more than twenty
branches outside India), the rest medium or small-scale. There were also
small independent merchants, who peddled goods supplied by firms
for one season, or purchased in India directly from artisans. Firms how-
ever accounted for the bulk of the turnover of the network, and the vast

4 Aitken, Gazetteer.
5 The term “trading diaspora” was first used by the Africanist scholar Abner Cohen in
his seminal article “Cultural Strategies in the Organization of Trading Diasporas”, in Meil-
lassoux, The Development, 266–278, and further popularized in Curtin, Cross-Cultural.
178 claude markovits

majority of the circulating men were in the employ of a firm. The princi-
pals of the firms themselves generally resided in Hyderabad and only con-
ducted inspection tours of their branches abroad at various intervals. The
circulating Sindworkies belonged to three main categories: 1) the manag-
ers, some of whom were working partners of the principals, others sala-
ried employees (and the proportion of salaried managers increased as the
network expanded to new areas); 2) the shop assistants, by far the largest
category; 3) the so-called “servants”. There was actually a fourth category,
of pedlars, but its importance diminished overtime. For the purpose of
this paper, I shall ignore the managers, since they uncomfortably straddle
the boundary between capital and labour, as well as the pedlars, about
which too little is known, and concentrate on the latter two groups, who
were clearly workers, although of very different origins and status.
My information about the commercial workers comes mostly from doc-
uments of a judicial nature, files concerning disputes between Hyderabadi
proprietors of commercial firms and their employees that came before the
Cairo consular court and other British consular courts in Egypt during the
period 1900–1925, which I traced in the Foreign Office Records at the Pub-
lic Records Office (now the National Archives of the UK). The reason why
disputes between Indian employers and employees in Egypt were referred
to the consular courts is simply that Indians in Egypt, being British sub-
jects, were subject to the jurisdiction of the Consular Courts, in the matter
of disputes with other British or foreign subjects, in accordance with the
stipulations of the “Capitulations” under which Egypt, still nominally a
part of the Ottoman Empire (till 1914) fell. The files contain in particular a
certain number of contracts of employment signed in Hyderabad between
employers and employees, which figured as exhibits in the court cases.
These contracts define with a certain degree of precision the conditions
under which employees of commercial firms travelled between Hydera-
bad and Egypt, and their terms of employment in Egypt. My sample of
cases is small, but I have no reason to think that it is not representative.
Apart from the fact that Egypt was a most important destination for Sind-
workies prior to the First World War, being really one of the major hubs
of that worldwide network (the other ones being Gibraltar, Singapore and
Hong Kong-Canton),6 the specific location of the cases has probably no

6 In 1915–16, out of a total of 572 certificates of identity delivered in Hyderabad to Sind-
workies to allow them to travel abroad, 69 were given for Egypt, then the fourth desti-
nation for these men, after the Dutch East Indies (108), Gibraltar (104) and the Straits
Settlements (79). See Markovits, The Global World, Table 4.1, 127.
 the circulation of commercial manpower 179

direct bearing on the form or content of the contracts, which I assume to

have been relatively standardized.
The reason why employees of Sindwork merchants were provided, at
least after 1900 (no clear evidence has come to light regarding the earlier
period, c. 1860–1900, in the history of the network) with written employ-
ment contracts, a practice rarely found in other similar commercial net-
works, either those of other Indians merchants, or those of Chinese or
Lebanese traders, can only be guessed at. Growing tensions between
employers and employees in Egypt were noticed from 1899 onwards by
the British authorities in Egypt,7 and they led to a spate of court cases,
which may have prompted an institutionalization of previously more
informal arrangements. It must have become apparent in Egypt to the
employers that the “normal” mechanisms for solving disputes between
employers and employees in India through caste panchayats or other
“customary” institutions were not available, and that the existence of
written contracts, which were clearly skewed in favour of the employers,
would come in handy in case of disputes in the courts. Another possible
factor is that the big firms employed an increasing proportion of non-
family labour; lacking the means of pressure normally available to heads
of firms to squeeze family labour, which consisted mostly in applying a
moral and financial pressure, they tended to seek redress in the courts in
case of theft, absconding, etc. by their employees.
The fact that private contracts signed in British India were deemed to
have legal value in British-occupied Egypt (that was officially annexed
only in 1914) and could be used as corroborating evidence in court cases
tried before British consular courts reflects the existence of what I would
call an imperial legal space, in which norms were largely standardized.
The homogenization of legal norms across an imperial space does not
mean however a unified labour regime. While employment contracts
signed in Hyderabad conformed to a certain general pattern, which was
directly derived from aspects of British labour law, in particular the whole
body of legislation known as the Master-Servant Act,8 they actually reflect
the existence of different labour regimes, in their turn linked to different

7 See the correspondence between the British Consulate in Cairo and the Foreign
Department of the Government of India in May 1907, reproduced in Bombay General (Mis-
cellaneous) Proceedings Emigration, November 1907, Serial no A 112, India Office Records,
Asian Pacific and African Collections of the British Library, London.
8 On this question, see the following magisterial survey: Hay and Craven, (eds),
Masters.
180 claude markovits

occupational and social profiles. To illustrate the point, I shall present a

comparative analysis of two contracts.

Two Labour Contracts

Contract no 1, dated 1901 has been partly reproduced in my book under

the heading “employment contract of a shop assistant.”9 While nothing in
the contract itself actually identifies the employee specifically as a shop
assistant, other evidence at the trial makes it clear that it was the case.
Besides, the salary scale is clearly that of a shop assistant, as known from
other sources. There are altogether eleven clauses to the contract. Clause
I concerns duration of service (two and a half years), location (“any place
they may be willing to send me to”) and salary (45 Rs per month to be
increased to 50 after 15 months), which was high by the standards of the
time (the average wages of an agricultural worker in Sindh were 12 rupees
per month, a figure that was higher than the Indian average). Clause II
concerns travelling expenses between Cairo and Hyderabad, to be paid
for by the employer. Clause III defines exactly the period of payment of
the wages. In Clause IV, that is probably lifted word for word from the
Master and Servant Act, the employee pledges honesty and obedience to
the employer. The tasks however are not described in detail (“I shall do
all the works they impose on me”). Clause V concerns the place where
accounts will be settled, which can be either Hyderabad or Bombay, a
clear indication that accounting took place in the city of residence of the
principals of the firms and not in Cairo, where the employees worked.
Clause VI mentions English calendar months as the unit of accounting,
a trait that sets that kind of contract apart from more widespread Indian
contracts using Indian calendar time. Clause VII concerns sanctions in
case of loss of or damage to goods. Clause VIII concerns dismissal in case
of breach of contract by the employee (nothing is written about breach of
contract by the employer) in which one can find an echo of a 1859 Indian
Act. Clause IX prescribes the large fine of 200 rupees (a sum equivalent to
four or five months of wages) in case of the employee not joining the firm
at the appointed time. Clause X is repetitive of Clause IX, concerning any

9 “Pohoomull Brothers, Plaintiff and Fatehchand Kayomull, Defendant”, Foreign Office

Records, Embassy and Consular Archives: Egypt, Cairo Consular Court Records, Dossier no
25 of 1902, FO 841/72, Kew, The National Archives of the United Kingdom (TNA), partly
reproduced in Markovits, The Global World, 298.
 the circulation of commercial manpower 181

breach by the employee of the clause concerning the date of expiry of his
contract. Clause XI concerns an advance of 25 Rs to be given by the firm
to the family members of the employee for their “maintain”, which hints
at the fact that the salaries earned in Sindwork were a crucial resource for
many lower middle-class families in Hyderabad.
Contract no 2,10 dated 1905, makes clearer use of a lexicon of master–
servant relationship. The employee is thus explicitly called a servant and
the employer a master, which was not the case in contract no 1. Clause 1
defines the work of the employee as that of a “general servant” and men-
tions Cairo explicitly as his place of work while not excluding “any other
place . . . wherever the said master may be pleased to send him”. Clause
2 fixes the duration of the contract as three rather than two and a half
years. Clause 3 gives the amount of salary to be paid, which is much lower
than in contract no 1: 10 Rs to be increased to 12 and 15 after one and two
years. Of this 6 Rs are to be paid to the parents of the servant, and the rest
kept by the master and given only on completion of the period of service.
Clause 4 is about conveyance to Cairo at the expense of the employer.
Clause 5 gives a detailed account of the allowances to be received by the
servant: board, lodging, washing, shaving, to be provided at the employer’s
expense, and a yearly allowance of 18 Rs for the purchase of clothing.
Clause 6 is the usual one about obedience and honesty. Clause 7 goes
into a detailed evocation of possible forms of misconduct (“gambling,
hard drinking, debauchery” and “any such act not sanctioned by the soci-
ety”) which it then defines as reasons for immediate dismissal. Clause 8 is
concerned with the keeping of trade secrets and prohibits the employee
from disclosing any information to competitors. Clause 9 is about breach
of contract, and, although it does not mention a fine, it includes the threat
of non-repatriation to India. Clause 10 is about the advance given to the
family of the servant.
A rapid comparison of the two contracts brings out firstly the large
gap in salaries, with a ratio of 4,5 to one in favour of the shop assistant’s.
Another striking feature is how much more paternalistic contract no 2
is: it purports to be prescriptive at the level of conduct, but also provides
more in terms of living expenses. Thirdly, regarding the payment of wages,
it appears that, contrary to the shop assistant, the servant is in fact not

10 “Kesumal Sobhraj, plaintiff, vs Chellaram Vasanmal, defendant”, Foreign Office

Records, Embassy and Consular Archives: Egypt, Cairo Consular Court Records, Dossier
no 7 of 1907, FO 841/91, TNA, partly reproduced in Markovits, The Global World, 299–300.
182 claude markovits

going to receive any money: part of his salary will be paid to his parents,
and the rest kept by the employer. This is at least in part linked to the fact
that there was a significant difference in age between the two employees;
while the shop assistant was 33, and therefore a mature adult, the servant
was much younger, probably under 18, although already past adolescence,
judging from the fact that shaving expenses were one of his perquisites.
This comparative exercise thus brings into focus the existence of a two-
tier labour market in Hyderabad for Sindwork. Although no detailed job
description is offered in the contracts, other evidence from court cases
shows that servants generally did not serve clients in the shops, contrary
to shop assistants, but were entrusted with menial tasks, such as cleaning
the premises, shelling pearls, or attending personally to the managers. It
seems also that, after a certain point in time, they were expected to do
the cooking, a task which, in an earlier period, was accomplished by high-
caste cooks specifically recruited for that purpose.
As to what the principles of this segmentation of the mercantile labour
market were, we are somewhat left guessing, because of a lack of docu-
mentation. Apart from age, which may or may not have been an impor-
tant differentiating factor (we do not know for instance whether it was
common for youths who had started as servants to later graduate to
the position of shop assistants), two obvious criteria that come to mind
are skills and family origin. Regarding the former, shop assistants were
required to be literate and numerate, to have some basic knowledge of
goods and markets as well as some rudiments of accounting, so as to be
able to manage the relatively complex tasks involved in selling goods to
an international clientele, and were also expected to have a smattering
of English, and even of other foreign languages, so as to be able to talk
to customers who were not likely to know Indian languages. Such quali-
fications could be found only amongst youths having frequented school
for several years. The existence of adequate educational facilities up to
the secondary level in Hyderabad, whose Hindu population had one of the
highest male literacy rates in the whole of British India, explains that the
merchants were able to find locally a readymade supply of shop assistants
amongst the educated youth. The latter later acquired on the job the kind
of practical knowledge necessary to the performance of their tasks. On
the other hand, servants, given the kind of simple manual tasks they were
expected to perform, did not need any literate skills, and certainly had no
need of English. It is probable that they were illiterate like the vast major-
ity of Indian workers at the time. Access to education being dependent on
socio-economic position, one can infer that shop assistants came mostly
 the circulation of commercial manpower 183

from merchant families with some resources, probably in part families

already engaged in Sindwork, while servants belonged to a lower social
strata. The exact sociological profile of Hyderabad town is unfortunately
not known to us in detail, in the absence of reliable studies. The only
quantitative data available come from the decennial Censuses, but the
latter are of little use as far as the socio-professional aspect is concerned.
What can be said with a measure of confidence is that, in Hyderabad,
contrary to most regions in India, caste cannot be treated as a proxy for
class, since most of the Hindu population of Hyderabad belonged to one
so-called caste, the Lohana. The latter was divided between two segments,
an upper strata of clerks and professionals known as Amils, and a lower
strata known as the Bhaibands, which encompassed the entire spectrum
of those engaged in the trading professions, from humble shopkeepers
to affluent bankers. Although practically all the Sindworkies belonged to
the Bhaiband group, that in itself was no indication of the exact position
occupied by an individual on the socio-economic scale.
The most remarkable feature of this system of labour circulation, espe-
cially when compared to the largely contemporary system of indentured
labour, is that it was not directly supported or even underwritten by the
colonial state. It was purely a product of Indian private enterprise, and
actually the colonial state appears to have become conscious of its exis-
tence only in the late 1890s when, as already mentioned, an influx of Sind-
workies in Egypt resulted in a sudden spate of court cases which led the
British authorities in Egypt to take up the matter in correspondence with
the Government of India. The authorities in Calcutta seem to have given
some passing attention to the question, but not to the point of actively
interfering in the organization of the circulation of the employees. The
latter did not fall under the purview of the vast corpus of emigration leg-
islation, developed from the 1830s onwards to regulate the movement of
indentured labour from India to the colonies. It is only after the abolition
of indenture in 1917 and the promulgation of a new Indian Emigration Act
in 1922, that all movement of working personnel from India became an
object of state regulation. From that date onwards, contracts which had
been signed in Hyderabad before a notary public had to be registered in
Karachi (the capital of the province of Sindh, and an important port of
embarkation for Sindworkies, although many of them left India through
Bombay or Calcutta) with the Protector of Emigrants, a magistrate specifi-
cally in charge of overseeing migration and of checking abuses. The shift
from a purely private system to a state-regulated one did not however
entail significant modifications, as shown by a detailed comparison I have
184 claude markovits

done of the 1901 contract already mentioned with a contract signed in

1929 at Karachi.11
Regarding the question of legal constraints on labour mobility, a spe-
cific dimension that has to be taken into account is that of cross-border
travel, crucial in the case of an international network. Between 1870 and
1920, there occurred a gradual transition from a regime of relatively few
restrictions on travel for commercial people to one in which restrictions
were many. While, in the 1870s, Indian merchants and their employees
could move practically everywhere, without any documents proving their
identity, and simply trust in the protection afforded by their status as
British subjects (following Queen Victoria’s proclamation of 1858, all resi-
dents of British India were considered subjects of the British Crown), the
growth, from the 1880s onwards, of “anti-Asiatic legislation” in the United
States, and in the white settler colonies of the British Empire, (Natal, the
Australian colonies), while meant mostly to stem the migration of cheap
Chinese (and Indian) labour, inevitably affected commercial manpower.
Starting in the 1880s, travelling Sindworkies had to carry some legal proof
of identity, mostly in the form of certificates of identity, which district
magistrates in Sindh were empowered to issue under a regulation dated
1863. The First World War really sounded the death knell of the free travel
regime, and, after that date, Sindworkies had to have proper passports to
travel abroad, which did not however always allow them to enter their
country of destination, including British territories. In the interwar period,
growing restrictions on travel led to calls from the Sindworkies for inter-
vention by the British government to ease them, and therefore left traces
in official correspondence.12
The organization by the merchants of a small inland town in India of a
system of worldwide circulation of employees, without any intervention
or support from the colonial state prior to 1917, is undoubtedly a puzzling
fact, and raises different kinds of questions. The only comparable example
that comes to mind is that of the commercial network originating from
the town of Wenzhou, in the Chinese province of Zhejiang, specialized in
the sale of leather goods, which was more or less contemporary. A com-
parison between Hyderabad and Wenzhou could be highly instructive but
unfortunately it has not been possible to trace studies of the “Wenzhou”

11 Markovits, The Global World, 172–73.

12 For a few detailed case studies, concerning in particular the Philippines and Panama,
see Markovits, The Global World, 230–240.
 the circulation of commercial manpower 185

in Western languages. It is possible however to speculate that, in the case

of Wenzhou, family and clan must have shaped the network in more
direct ways than in Hyderabad. Regarding the Sindworkies, it is striking
to note how the sphere of labour relations acquired a degree of autonomy
vis-à-vis the sphere of family and caste relationships. Their case is clearly
different from that of another well-known network originating from India,
that of the Nattukottai Chettiar bankers, who dominated the field of rural
credit in Burma and Ceylon between 1870 and 1930, and also played an
important role in Malaya, Thailand and French Indochina. In his study of
this community, American anthropologist David Rudner emphasized the
central role played by caste in its success (Rudner 1994). In the case of the
Chettiars, it was their caste institutions, based in Saivaite temples, that
ensured the regulation of financial transactions within the community.
Nothing comparable can be found in the much more amorphous social
and religious universe of the Sindhi Hindus, characterized by a certain
degree of syncretism with Sikhism and Sufi islam. Such a “soft” structure
appears to have created a space in which it was possible for class antag-
onisms to emerge within an apparently homogeneous community. The
degree of polarization that existed should not be exaggerated: the shop
assistants, the group most capable of manifesting an antagonism vis-à-
vis the employers, not excluding forms of violent behaviour, were often
themselves potential small capitalists. It was common for Sindworkies,
after having served an employer for years, firstly as shop assistant, then as
manager, to create their own firms and to become in their turn employ-
ers of labour. In spite of the existence of forms of mobility internal to the
community, there remains the fact that a confrontational stance was fre-
quently present in the relationship between employers and employees.
Regarding the question of the degree of freedom these workers
enjoyed, a comparison with indentured labour is in order. At first glance,
shop assistants would appear to constitute a clear case of free labour, in
striking contrast to the indenture regime, while servants could be seen
as being closer to indentured labourers. Yet both groups of employees
had proper contracts, signed in front of witnesses, while it is known that
generally indentured labourers did not actually sign and take cognizance
of their contracts before their departure from India.13 While there is still
a debate about whether indentured labourers were voluntary or forced

13 Thus, in the 1840s, indentured labourers going to Mauritius would sign contracts only
after their arrival in the colony. Mentioned in Northrup, Indentured Labor, 63.
186 claude markovits

emigrants, there seems no doubt that Sindworkies, be they shop assistants

or servants, were not coerced into entering into agreements with employ-
ers. Lack of open coercion is not however sufficient to conclude to the
“free” character of a labour contract. Firstly it is possible that some of the
so-called servants were actually employed as chattel to repay the debts
incurred by their families to their employers. There is nothing explicit in
the sources about it, but the specific payment arrangements in contract
no 2 leave at least that possibility open, although it remains necessarily
speculative. Regarding the actual terms of employment, a fundamental
question is whether these workers were free to terminate their employ-
ment before the expiry of the period for which they had signed, without
risking penal sanctions. It is not clear from the contracts themselves, but
other evidence from court cases and other sources shows that it was not
uncommon for a shop assistant or a servant to change employer or to
start business on his own when in Egypt, where there seems to have been
a steady demand for labour in Sindhi shops as well as for pedlars. The
degree of freedom of the workers would actually have depended more
upon the state of the labour market than upon the legal nature of contrac-
tual arrangements. One has to take into account the fact that labour in
commercial establishments in a big city like Cairo could not be as tightly
controlled as labour on a sugar plantation in Mauritius or Guyana, where
mobility was actually very restricted for Indian indentured workers. There
remained nevertheless a significant difference between the cases of the
shop assistants and of the servants. The former actually pocketed their
wages regularly and, given that their ordinary expenses were much lower
than their wages, could accumulate some savings. Their effective saving
rate however depended on two variables: on the one hand the amount
they remitted to their families, and on the other hand their lifestyle; if
they drank, gambled and frequented prostitutes, three temptations which
loomed large in a city like Cairo, (well known for its large number of bars,
brothels and casinos frequented by a cosmopolitan clientele), they could
easily run into large debts. Provided however they were prudent enough,
they could easily survive a period of unemployment between two jobs. On
the other hand, the servants did not actually pocket their wages, which
were partly paid directly to their families, partly retained by their employ-
ers, and had no way of surviving dismissal in Cairo, except if another Sind-
work merchant took them immediately in his service.
The shop assistants can be seen as forming a sort of labour aristocracy,
their salaries being much higher not only than those of most commercial
employees in India (about which little reliable information is nevertheless
 the circulation of commercial manpower 187

available) but also than those of many such employees in Britain. On the
basis of the data provided in a book published in Manchester in 1910 by
British trade-unionists dealing with the situation of commercial employ-
ees in Britain14 it appears that they formed one of the most exploited
categories of British workers. Their wages were low, they worked very
long hours (between 60 and 85 per week, depending on branches and
categories of workers) and had precarious life conditions. Some lived on
the premises of the commercial establishments, a trait that is found also
in Cairo amongst employees of Sindworkies. The latter also worked long
hours (although no detailed information is provided in the judicial docu-
ments), but their salaries were relatively high, a consequence firstly of the
prosperity of the trade in semi-luxury articles in which the Sindhis were
specialized, and secondly of the fact that they had qualifications that were
more than average (in particular linguistic competence that was essential
for dealing with an international clientele). Such an “aristocracy” how-
ever cohabitated in the Sindhi shops of Cairo with a group that had more
resemblance to a kind of sub-proletariate, with employment conditions,
as known through the contracts, that were close to debt servitude, a sys-
tem that was and is still widespread in India. The cohabitation of these
two very different kinds of manpower did not fail to generate certain ten-
sions, about which we know however too little.

Conclusion

To conclude this presentation, I shall address the question of the relation-

ship of this system of labour circulation to colonialism. Was it a direct
product of the colonial situation, or the outcome of a mixture of “tra-
ditional” features and of innovations due to the colonial regime? The
legal framework as already mentioned, was decidedly British colonial:
the contracts, especially contract no 2, are direct derivations of contracts
under the Master and Servant Act. They were apparently written in Eng-
lish (although there may also have been a version in Sindhi), and signed
before a notary public with three witnesses. They followed a so-called
“English” calendar, different from the various Indian calendars, whether
Hindu or Muslim, that regulated most of commercial activity in India.
They included financial clauses which seem to belong more to a British

14 Hallsworth and Davies, The Working Life.

188 claude markovits

than to an Indian universe. Payment of wages on a monthly basis in par-

ticular, as far as can be ascertained, was not a very widespread practice
in the commercial world of India. On the other hand, the “moral” clauses
included only in contract no 2 could be reflecting either Victorian Puri-
tanism or traditional Indian merchant ethics, which were actually not
that different. The system of advances on wages does not seem either to
be specifically British or Indian, but could probably be found operating
in both settings. The legal framework under which the employees of the
Sindwork merchants circulated and laboured between India and Egypt
in the early twentieth century appears thus to be the product of a prag-
matic adjustment between British legal norms and Indian practices, a sort
of “hybrid”, which reflects the capacity of Indian merchants to adapt so
as to be able to function in an international setting. On the other hand,
the actual working of the system in Egypt was the product of the inter-
play between market forces, which seem to have somewhat favoured the
employees, at least the shop assistants, because of the existence of a short-
age of qualified manpower, and a legal system which was clearly skewed
in favour of the employers, producing a contentious work environment in
which disputes were rife, and violence not unknown. And of course it is
because of the disputes that there are archival traces of this story, which
would otherwise have remained hidden. Whether any kind of broader
conclusion as to the problem of labour mobility in the Eurasiatic space
can be derived from such a limited case study is a point I shall leave to
the appreciation of the readers.
 Part three

Dangerous Ties:
State, Landlords and Longue Durée Servitudes
 Constrained Labour in Early-Modern Rural East-central
and Eastern Europe: Regional Variation and its Causes

Markus Cerman

Introduction

Several generations of rural and agrarian historians broadly agree that the
establishment of a commercial demesne economy and a deterioration of
the legal and social situation of the rural population were central elements
in the characterization of early modern East-central and Eastern European
agrarian structure. This conception lies at the heart of the construct of
an ‘agrarian dualism’ between early modern Eastern and Western Europe
and formed an important pillar in assessing an alleged long-term East-
ern European economic backwardness in conventional but also in world-
system analyses. By contrast, there is no agreement about the exact timing
of the establishment of demesne lordship and a demesne economy nor
about the main causes for these developments. Likewise, approaches
to this alleged special agrarian system differ widely. While traditional
schools of thought place most weight on legal factors, such as the rights
of demesne lords and the weak position of villagers, other approaches,
including Marxist ones, primarily focus on economic aspects. Among
the latter, the attempts of historians of the former GDR, former Czecho-
slovakia and Hungary focussed on the prevalence and burden of labour
rents (forced labour services, corvée) (see below), which villagers had to
perform for demesne farms and the demesne economy.1 The actual bur-
den of labour rents has been used in research to differentiate between
‘mild’ forms of demesne lordship and ‘harsh’ ones which could include
the establishment of serfdom for instance in Schleswig-Holstein, Meck-
lenburg or Pomerania. The threshold has been set at two to three days of
labour rent per week for a full tenant farm.2

1 Harnisch, “Gutsherrschaft in Brandenburg”, 138, 145–46; Pach, “Bauernaufstand”; Válka,

“Druhé nevolnictví”. See also Bush, “Serfdom”, 211, 213; Knittler, “Between east and west”.
2 Harnisch, “Gutsherrschaft in Brandenburg”, 146–47; Harnisch, “Probleme”, 252, 272–73;
Henning, Dienste, 46–70, 116–19; Bush, “Serfdom”, 213.
192 markus cerman

It remains questionable whether regimes of serfdom are accurately

defined by this one variable alone, but this is not the concern of the pre­
sent paper. Instead, it mainly refers to the organization of work in the
demesne economy and whether or not forced labour was really the domi-
nant source of labour in early modern East-central and Eastern Europe. In
order to analyze this it is necessary to discuss, in a first step after an ini-
tial discussion of the terminology, the causes for the rise of forced labour
services after 1450/1500 and the ways, lords succeeded in raising them.
Es­pecially in the light of studies published in recent years it becomes clear
that landlords had to negotiate higher forced labour services and make
concessions to villagers. In some circumstances, higher labour services
may have been more favourable for the economy of tenant households
than an increase of rents in cash and kind. Secondly, the paper will high-
light the fact that it is sometimes very difficult to distinguish between
‘forced’ and ‘free’ labour in the context of the early modern demesne econ-
omy and will discuss areas where characteristics overlap. Hence, studies
which claim the prevalence of one or the other form need to be evalu-
ated carefully as to whether this problem in distinguishing these forms
is sufficiently considered. Thirdly, the main argument developed here is
that the demesne economy in general and demesne farming in particular
could not rely on forced labour services alone, but that a mix of various
forms of labour—with considerable swings in the importance of one or
the other form—prevailed. With the exception of Russia, however, there
seems to have been a tendency during the eighteenth century to increas-
ingly convert a part of the abundant labour rents into cash payments and
to use wage labour for demesne farming to a higher degree.
Labour rents or corvée represent (usually unpaid) compulsory work bur‑
dens of tenants which were mainly used for demesne enterprises. They
were part of the ordinary rent obligations of tenant farmers, smallholders
and cottagers in the areas surveyed here. It is significant that cash rents and
rents in kind were typically small and the major burden for tenant hold-
ings was represented by the obligation to render labour rents, ­especially
in regions in which they were actually demanded and were higher than
one or two days a week per holding. In the area surveyed here, labour
rents are referred to in the literature as Arbeitsdienste or Scharwerk, robota,
pańszczyzna or barshchina. These forced labour services or labour rents
have to be differentiated from two other forms of compulsory work used by
demesne lords, i.e. compulsory agricultural life-cycle service ­(Gesindezwang
in German) and compulsory wage labour (see below). In certain terri‑
tories, it was part of the rights of demesne lords to force children of tenants
 regional variation and its causes 193

into employment as life-cycle farmhands and maids (for which, however,

the individuals received wages) for a certain number of years. Households
of smallholders and cottagers could be asked to supply compulsory wage
labour, e.g. during agricultural peaks, as part of their tenancy regulations.
As areas of different languages are covered in the following, the paper will
use the approximate English translations introduced before.3

Why and How Did Demesne Lords Increase Labour Service Demands?

With the exception of Russia, the rural population of East-central and

Eastern Europe was confronted with seigniorial demands of rising forced
labour services (corvée) at some point between the second half of the fif-
teenth and the early seventeenth century. As with regard to actual levels
of forced labour services, there was also considerable regional variation
in the occurrence of these demands (in the timing or in the strength in
which these claims were presented) and in the reasons for this trend.
As ‘classic’ major cause, the literature identifies the growth of the com-
mercial demesne economy in East-central and Eastern Europe from the fif-
teenth century onwards. Confronted with the erosion of conventional rent
income during the crisis of the Later Middle Ages and the strong devalu­
ation of (fixed) customary rents during the sixteenth century, landlords
all over Europe were desperate to generate additional revenue. While the
flexibility of tenures and the development of property rights in the West
allowed for systematic rent increases (such as for short- or medium term
leases) or the introduction and increase of entry fines, which soon became
important sources of additional income also for lords in Central Europe,
various theories of demesne lordship suggest that due to the abundance
of land demesne lords in East-central and Eastern Europe systematically
turned to direct management and the expansion of commercial demesne
farming geared at international markets where grain demand grew from
the late fifteenth century onwards.
The fact that increases in forced labour services were reported system-
atically in the sixteenth century would suggest that this period did indeed
represent a watershed in this respect. Typically, the literature refers to
estate registers or to negotiations on territorial Diets. These documents
may reflect a shift in norms or in general views on the matter, but they

3 For an excellent discussion of the relevant German and Polish terminology see Hagen,
“Village life”, and Hagen, Ordinary Prussians.
194 markus cerman

cannot be regarded as sufficient evidence that higher services were ac‑

tually generally imposed. The impression remains that demesne lords may
have needed labour of their rural tenants desperately, due to the lack of
servants or wage labourers and high wage levels. Complaints on Diets and
princely regulation of the terms of agricultural life-cycle service in this
period reflect this concern.4
Contrary to often repeated images in the literature, the existence of
forced labour services was not unusual in late medieval East-central and
Eastern Europe, either. Yet, traditional accounts were right to suggest
that apart from special social groups directly attached to demesne farms
average service obligations usually were limited to a few days per year.
With the tendency to abandon demesne activities during the fourteenth
and early fifteenth century, they were further reduced or converted into
payments. When the tide turned, lords succeeded in raising labour rents
beyond the medieval levels by different steps whose timing and mechan­
isms could vary.
Recent studies investigating the rise of forced labour services after
1450/1500 do not ignore the use of force by lords in this process, but also
stress the element of negotiation with villagers throughout the early mod-
ern period. According to this argument, it seems clear that in an envi-
ronment of high levels of village desertions and a strong competition for
tenants—of which measures of the Estates in various territories, which
were aimed at forcing lords to return fugitives to their original lords give
evidence (and which, in a re-interpretation that took a number of steps,
would become the basis of legal mobility restrictions in the sixteenth cen-
tury and later)—the use of violence would get lords only so far, before,
e.g., tenants left elsewhere. The bargaining position of villagers had defi-
nitely increased (although legislative changes were again intended to
strengthen lords, see below). The gradual change therefore should be
interpreted as a complex process of negotiations in which lords had to
offer something in exchange for the services they wanted to receive. It has
also been suggested that the original demands of forced labour services
and simultaneous pressure for the increase of other rents occurred at a
time of growing income for tenant farmers during the sixteenth century.
They were thus able to absorb some of this pressure in a period of rising
living standards among farmers and smallholders without any significant
loss of income.5

4 See, e.g., for Prussia Aubin, Geschichte, 120–22, 124, 130–31.

5 Kula, Theory, 125.
 regional variation and its causes 195

Initially, tenant farmers and smallholders may have consented to render-

ing forced labour services occasionally, which were requested rather than
commanded or were introduced by means of formal contracts between
demesne lords and villages in return for the reduction of other payments
and dues or for additional land.6 During this period, when inflation had
eroded much of the original value of cash rents, the rural population may
have agreed to rendering corvée, especially when levels were still low (a
few days per year). Based on this, they could keep most of their produc-
tion for commercial sales, which was more favourable for their income in
an environment of rising agricultural prices. A recent study suggested a
different argument for late medieval Poland. According to this, villagers
themselves preferred the change towards forced labour services during
the fifteenth century, because they were unable to sell grain at sufficient
prices and hence raise the necessary cash to service their rents. By giving
up some of their land as well, they succeeded to reduce overall rent levels
without jeopardizing their subsistence livelihood. The earliest legislation
of maximum labour services on Diets within parts of the Polish Common-
wealth in 1520 is thus seen as a means to prevent that all villagers could
exchange their rents in cash and kind against labour rents rather than as
a push of lords to increase overall labour rent levels.7
The empirical literature reports a large number of cases which prove
that increases of forced labour services in the fifteenth and sixteenth cen-
tury were based on negotiations between lords and villagers. For instance,
in the Brandenburg estate of Stavenow farmers only agreed to provide
more labour services, because their rents in cash and kind were held con-
stant or reduced.8 For the needs of a new demesne farm in Hindenberg
in Mecklenburg in 1517, the Duke established a new general service of one
day in every third week in two villages. In return, villagers’ payments in
lieu of services were reduced by half and they no longer had to perform
transport services which had existed before.9 In a formal contract between
the lord and the villagers of Roxförde and Wannefeld in the Brandenburg
Altmark, both parties held the right to choose between labour services
and payments.10 As late as in 1770, a village in one of the Esterházy estates

6 Enders, Uckermark, 179, 193–200; Enders, Prignitz, 400nn; Enders, “Frondienste”; May-
baum, Entstehung, 130; North, “Entstehung”, 57.
7 Mikulski/Wroniszewski, “Das Vorwerk”, 122.
8 Hagen, “Peasant rents”; Hagen, “Village life,” 168; Hagen, Ordinary Prussians, 32nn,
65nn; Scott, Society, 190.
9 Maybaum, Entstehung, 134.
10 Enders, “Frondienste”, 90.
196 markus cerman

in Hungary chose to offer labour rents to avoid the introduction of a rent

in kind.11
In other cases, villagers may have accepted forced labour services when
lords offered certain ‘help’. A village of the Mecklenburg monastery of
Dargun agreed to being ‘protected’ by a neighbouring noble lord and vil-
lagers were ready to perform some work for him under the condition that
they be served meals on the days of work. When the village wanted to end
this arrangement some time later, the lord turned against them and he
insisted that these services had to be rendered on a regular basis. In 1573,
the conflict was resolved and the village consented to paying a consider-
able sum in exchange for the labour rents.12
Some of the examples show that the development was more often ac‑
companied by conflicts than not and demesne lords exerted various
degrees of pressure, including outright violence or imprisonment of those
who resisted their claims.13 It was the rising seigniorial bargaining power
based also on state concessions to the Estates during the sixteenth cen-
tury which slowly and persistently weakened the position of villagers.
Constant resistance against increasing demands by demesne lords, who
tried to undermine earlier agreements, became the rule.14 Often, it was
the rural population that insisted that a specified level of forced labour
services be written down once and for all. When previous agreements
stated that they were to help at harvest or at haymaking, it left them vul-
nerable, when demesne lords extended their demesnes or built new ones.
By setting an explicit number of days or clearly specified tasks, further
demands for increases could be prevented and it may have been because
of this that the well-known general regulations of a certain number of ser-
vice days per week first entered contemporary documents or legislation.15
Brandenburg tenant farmers and smallholders fought fierce legal battles
with their demesne lords over binding arrangements for forced labour
services from the 1540s on. In many cases, they turned to the territorial
prince that a solution be decreed or existing contracts be obeyed.16 Some-
times, outright injustice prevailed. Thus, despite the fact that the Meck-
lenburg Duke agreed to relieve his tenants on the island of Poel from their

11 Domanovszky, “Gutsherrschaft”, 459.

12 Mager, Geschichte, 100.
13 Enders, Uckermark, 179, 195, 198–99; Harnisch, “Gutsherrschaft in Brandenburg”,
135–39; Maybaum, Entstehung, 135–36, 149.
14 See e.g. Enders, Uckermark, 193–201; Enders, “Frondienst”, 145.
15 Enders, “Frondienst”, 92, 94.
16 Enders, “Frondienst”, 97–9.
 regional variation and its causes 197

labour obligations of four days a week in exchange for an annual payment

of a certain quantity of barley in 1639, his administrators continued to
demand services. When the villagers complained to the Duke, he insisted
to uphold the original burdens despite the agreement.17 Such moves often
were accompanied by the use of violence to intimidate villagers or crush
their resistance.18
Apart from the ruling of Diets and some princely concessions, tradi-
tional accounts often refer to the prior acquisition of jurisdictional lord-
ship as an important basis to claim and receive new or increased services.19
With these powers, lords could police arrangements and gained access to
forced labour services which were originally reserved for the prince (by
his power of jurisdictional lordship). With respect to the medieval foun-
dations of early modern developments, this seems an interesting point
for future analysis. A careful study for Mecklenburg, however, describes
this as a relatively complex process which also displayed important
discontinuities and was far from being a uniform pattern even within
this country.20
The evidence presented reveals that the increase of forced labour ser-
vices frequently came in return for freezing or reducing other rents. Vil­
lagers were often prepared to accept this, because they could minimize
cash and commodity rents in periods of low agricultural prices or reap
maximum commercial profits during agricultural booms, because their
market quota was not affected by seigniorial demands. The significant
general erosion of rent value in cash and kind in real terms between the
fifteenth and early seventeenth century was recently emphasized in sur-
veys on Brandenburg and Poland21 and is evident in numerous case stud-
ies for regions in East-central and Eastern Europe.22

Identifying Forms of Forced and Wage Labour

One of the difficulties of research regarding a clear picture of the organ­

ization of work within demesne enterprises is that it is often impossible
to differentiate between forced labour services, compulsory wage labour

1 7  Mager, Geschichte, 101– 102.

18 Enders, Uckermark, 195, 198–99; Göttsch, Leibeigene.
19 Blum, “Rise”, 822–26, 833–35; Schmidt, Leibeigenschaft, 127–30.
20 Maybaum, Entstehung, 42–3, 49–86, 87–95, 97–107, 126–30.
21 Guzowski, “Changing economy”; Hagen, “Village life”; Hagen, Ordinary Prussians.
22 Muszyńska, Gospodarstwo; Wajs, Powinnośći.
198 markus cerman

and wage labour.23 For instance, life-cycle servants on demesne farms

are often regarded as being in waged employment, which, in the strictest
sense, they were. Demesne lords, however, often had the right to recruit
servants from the children of their tenants’ households and force them
into paid service for a certain number of years.24
The border line is equally difficult to draw in the case of day labourers.
Demesne accounts often show expenses for the wages of day labourers.
Yet, in many cases, these may have been smallholders or cottagers, who
had to supply compulsory wage labour when they were summoned as
part of their feudal obligations. To complicate matters further, custom
often stipulated that villagers had to be compensated even for their ten-
urial forced labour service obligations. As a consequence, people often
received food or beer on the days they worked on demesne farms (and
these expenses would be registered as well).25 Compulsory wage labour
was also systematically used to supplement labour services, such as in
Lithuania when demesne farms were further extended during the eight-
eenth century.26
When studies emphasize a widespread use of wage labour in the
demesne economy for instance in certain Central European territories
prior to 1650, the existence of overlaps between forced and wage labour
does not always seem to be considered in a sufficient manner.27 It thus
seems problematic that the exact nature of ‘wage labour’ is not character-
ized in necessary detail in these examples. For instance, the employment
of servants on demesne farms would usually include villagers’ children
in forced life-cycle service as well as waged day labourers; hence, distin-
guishing between compulsory wage labour and wage labour in general
terms seems difficult.28
Sixteenth- and seventeenth-century laws enabled demesne lords to
force villagers’ children to enter agricultural life-cycle service on demesne
farms (Gesindezwang). The term, during which they were paid wages

23 Pach, Agrarentwicklung, 30; Pach, “Problematik”; Żytkowicz, “Trends”, 71.

24 Hagen, “Village life”, 168.
25 See e.g. Pach, “Problematik”; Soom, “Gutswirtschaft”, 46–7; Stutzer, “Ertrags- und
Lohnverhältnisse”, 20; Wächter, Domänenvorwerke, 63.
26 Kahk/Tarvel, Economic history, 71.
27 See Čechura, “Gutswirtschaft”; Čechura,“Dominium Smiřických”; Čechura, Grund-
herrn, 56–62, 114–17, 150–54. 166nn.
28 Válka, Hospodářská politika; Válka, “Druhé nevolnictví”.
 regional variation and its causes 199

as servants, was usually limited to three to five years.29 For instance in

Brandenburg, it was first established in 1518 and confirmed in various new
ordinances during the sixteenth and seventeenth centuries.30 In Western
Pomerania, a universal forced life-cycle service was usual from the begin-
ning of the sixteenth century. The East Prussian Land Ordinance of 1577
introduced forced life-cycle service for children of subject tenants which
were not needed on their families’ farms.31 In Upper Lusatia, forced ser­
vice with prescribed wages was first established in 1539 and regulations
were regularly repeated and extended in land ordinances until 1767.32
After 1550, regulations on Bohemian crown estates repeatedly confirmed
the possibility to hire subject children as servants. As examples and pro-
tests show, forced life-cycle service was usual in noble estates at this time,
too.33 Resolutions of the Silesian Diet governing mobility restrictions, ordi-
nances for servants, and compulsory life-cycle service were passed in 1502,
1512, 1528, and 1565. The provisions of these were standardized in the 1652
Ordinance for Servants and Subjects.34 Yet, compulsory life-cycle service
was far from universal in East-central and Eastern Europe. In Poland, Slo-
vakia or Hungary, for instance, comparable regulations did not exist.35 In
Mecklenburg, compulsory service was usual, but does not seem to have
been formally introduced by law.36
Due to the existence of compulsory life-cycle service on demesne farms
and of compulsory wage labour it is therefore difficult to conclude from
wage expenses recorded in seigniorial accounts that these were only
incurred for ‘free’ wage labour.37 A study on estate management in Upper

29 Boelcke, Bauer, 113–15; Henning, Herrschaft, 142–46, 157–58; Kaak, Gutsherrschaft, 20,
40, 103–106; 141, 156, 399; Maur, “Poddanská otázka”; Maur, “Das Gesinde”, 80–1; North,
Amtswirtschaften, 83; Wächter, Domänenvorwerke, 53–5.
30 Enders, “Entwicklungsetappen”, 124; Harnisch, Boitzenburg, 69; Kaak, Gutsherrschaft,
103–6; Kaak, “Erbzinsrecht”, 76.
31 Aubin, Geschichte, 134–35; Henning, Herrschaft, 142–46, 157–58; North, Amtswirt-
schaften, 83; Wächter, Domänenvorwerke, 53–5.
32 Boelcke, Bauer, 113–15.
33 Maur, “K demografickým aspektům”, esp. 31–2; Maur, “Gesinde”, 80–1; Urfus, “Právní
postavení”.
34 Feigl, “Entwicklung”; Weber, Polizei- und Landesordnungen, 150nn, 197nn.
35 Horváth, “Poddanská otázka”, 11–2; Lubinski, “Everyday work”, 148; Špiesz, “Vývoj”.
36 Kaak, Gutsherrschaft, 20, 40, 141, 156, 399; Mager, Geschichte, 87, 163–65; Rudert,
“Grenzüberschreitungen”, 361.
37 Míka, “Problém”, 233, 241–42; Válka, “Nevolnictví”, 320–21; Válka, Politika, 100. See
also Jirásek, “Venkov”, 98nn; Maur, “Gesinde”, 80–1; Válka, “Stavu”, esp. 200. For compulsory
wage labour on Rožmberk estates see Longfellow, “Bohemia”, 7. Also North, “Lohnarbeit”,
argues for East Prussia that many wage labourers from sub-peasant strata were forced into
employment and were paid in bread and beer rather than money.
200 markus cerman

Silesia at the end of the eighteenth century, for instance, shows that ten-
ant farmers were compensated in cash and kind for the statutory forced
labour services (corvée) they had to render for the demesne economy and
total wage costs including wage labour would represent up to 33 per cent
of seigniorial income. The money tenant farmers received approximately
made up for their maintenance costs of the team of draught animals
which they had to provide and for food for one person. This compensa-
tion was relatively lower than that offered to cottagers as wages as they
also received free housing and did not have to bring draught animals.38 As
Zs. P. Pach already concluded: “Paid work was (. . .) sometimes burdened
with the chains of extra-economic coercion, corvée was occasionally inter-
woven with threads of wage labour relations.”39

Forced Labour and the Demesne Economy

The previous section established that East-central and Eastern European

demesne lords could rely on various forms of labour to operate their com-
mercial demesne economy. In the light of this, it seems only straightfor-
ward to assume that forced labour services were almost never the only
source of labour for the demesne economy and very often not even the
dominant one. It is far more likely that it worked on the basis of a mix-
ture between (mostly unpaid) forced labour services and (compulsory)
wage labour of one or the other form,40 the balance of which was sub-
ject to enormous variation between regions and periods. The two poss­
ible extremes were characterized by GDR-historiography as Teilbetrieb
(operation with forced labour services) and Eigenbetrieb (operation with
servants and wage labour only).
Several structural reasons can account for the variation and degrees of
overlap between the individual modes of operation. To begin with, the
most valuable variant of labour rents is associated with tenant farmers
whose holdings were large enough to supply ploughteams or draught ani-
mals for transport services. Thus, during the initial phases of the exten-
sion of forced labour services up until the sixteenth century, certain social
groups, who were unable to provide these services may have been exempt.
Among them were cottagers and lodgers, unless they had been specifi-
cally settled as demesne workers during the later Middle Ages, which

38 Stutzer, “Ertrags- und Lohnverhältnisse”.

39 Pach, Agrarentwicklung, 30. See also Rusiński, “Remarks”.
40 Melton, “Gutsherrschaft”, 329; Pach, “Problematik”.
 regional variation and its causes 201

was not exceptional in some areas, such as in Prussia or in Silesia. In all

other regions, in which they had no prior connections to the demesnes,
they were only gradually integrated into the demesne economy, first as
wage labourers in voluntary but also compulsory wage labour and later
for manual services. In Estonia, as late as 1796 and 1798 laws were passed
ruling that also cottagers and crofters generally had to supply forced
labour services.41
The actual use of forced labour services depended on the extent of the
demesne economy and particularly on the proportion of demesne arable
land. The ratio between demesne and tenant land varied strongly between
territories but also between individual regions. As a consequence, the
exact ratios established on the basis of territorial land surveys or cadas-
tres can only serve as an approximation. But even these indicate that it
is very difficult to generalize. For instance, in the Brandenburg Kurmark,
the average proportion of demesne arable was found to be a third of the
total arable land in 1797, but it was as little as 16.1 per cent in the district of
Luckenwalde and as much as 43.8 per cent in the district of Uckermark.42
Proportions of more than 50 per cent were not rare on the level of indi-
vidual estates or even villages, but certainly the exception for entire ter-
ritories. Thus, the demesne share of about 60 per cent established for late
seventeenth-century Swedish Pomerania is an extreme example within
East-central and Eastern Europe.43 There, the expansion of demesne at
the expense of tenant land paired with the strong decline in tenant num-
bers as a consequence of the Thirty Years’ War meant that demesne farms
did not have at their disposal an adequate number of villagers to provide
forced labour services and thus had to turn to wage labour. An unusually
high proportion of 25 per cent of demesnes in Swedish Pomerania was
operated in Eigenbetrieb in the late seventeenth century.44 Similar devel-
opments were also observed in the Brandenburg Uckermark in the same
period. It seems as if there was an inverted-U-relationship between the
use of forced labour services and the proportion of demesne land.
Leaving aside Russia, Belorussia and the Ukraine, a comparative sur-
vey on the extent of forced labour services in East-central Europe in the
eighteenth century distinguished between three different zones. In
the first, tenants had to provide services on a demesne land area that

41 Kahk, Peasant, 32.

42 Harnisch, “Gutsherrschaft”, 230; Müller, Landwirtschaft, 44–5.
43 Schilling, Schwedisch-Pommern, 19–21, 23, 50–1.
44 Rudert, Gutsherrschaft und Agrarstruktur, 118; Schilling, Schwedisch-Pommern, 49–50,
54, 64, 108.
202 markus cerman

represented less than fifteen per cent of the land of their own holdings.
This would hold for Denmark, much of western and northern Schleswig-
Holstein and most areas of Hungary and Slovakia. The maximum burdens
occurred in those regions, in which forced labour services required ten-
ants to work on demesne land larger than 40 per cent of the size of their
own holdings: eastern Holstein, Mecklenburg, Pomerania, the Branden-
burg Uckermark, Greater and Little Poland and Mazovia. Finally, there
was a group, in which the work burden amounted to caring for demesne
land of between 15 to 40 per cent of their own land, such as in Branden-
burg, Upper and Lower Lusatia, Lower Silesia, Bohemia, Moravia, Prussia,
Courland and Livonia.45
The type of farming activity could influence the organization of work
within the demesne economy. For instance, the importance of viticulture
in parts of Hungary, Slovakia and South-eastern Austria limited the use
of forced labour services except for transport purposes. As low quality
work usually associated with labour services could be damaging, primar­
ily wage labour was used for demesne viticulture.46 Similarly, seignio-
rial profits from agricultural exports in the sixteenth century were not
based on cereals only, but in Denmark, Schleswig-Holstein, Hungary and
some areas of the Polish Commonwealth dairy and oxen exports were
far more important. Dairy and pastoral activities involved tasks for which
demesne farms usually employed servants in continuous employment. As
a last example, in certain areas of Bohemia and Moravia commercial pis-
ciculture was more important than arable farming especially in the early
phases of demesne expansion. The building and rearing of ponds was car-
ried out by well-paid specialists and there were only few tasks for which
forced labour services could be used in this context.47
Finally, local conditions could influence whether the inhabitants of
particular villages were called to forced labour services or not. Even in
areas with a strong reliance on corvée, individual villages could be entirely
exempt and paid cash rents in lieu of services, because they were not
close enough to demesne farms. For an estate in Mecklenburg in 1620,
the average distance villagers had to cover to reach the demesnes they
were assigned to was calculated at five kilometres.48

45 Henning, Dienste, 46–70, 116–19.

46 Horváth, “Der Charakter”, 92–3; Kiss, “Der Agrarcharakter”, 148; Pach, Agrarentwick-
lung, 57–9, 74–80, 98–9, 110, 130–31, 133–34, 143–44; Pach, “Problematik”; Zimányi, Economy,
81–2.
47 Míka, “Problém”, 241nn; Míka, Lid; Petráň, Výroba, 161; Vacek, “K dějinám”, 10.
48 Baumgarten/Bentzien, Hof, 221.
 regional variation and its causes 203

Wage labour of day labourers, farmhands and servants on demesne farms

as defined above was always used for the operation of demesne farms.
As indicated, the share it represented relative to the total demand for
labour varied strongly between regions and over time.49 There is a ten-
dency in most recent research to rehabilitate a rather widespread use of
wage labour within the early modern demesne economy in East-central
and Eastern Europe in order to overcome traditional generalizations of
an overall dominance of ‘serf’ forced labour economies. While several
detailed studies succeed in establishing this point for individual cases or in
a broader perspective, we need to be aware of the fact that the mere pres-
ence of full-time servants on demesne farms cannot always be equated to
the absence of forced labour service demands. In these cases, the question
needs to be answered, whether the available servants and farmhands were
actually employed for the works connected with the dominant demesne
activity of arable agriculture, for which forced labour services were usu-
ally called.
For example, in the Moravian estate of Telec in 1591, only 13 of 50 ser‑
vants on five demesne farms were employed for arable agriculture, because
the others were explicitly shepherds or dealt with the other livestock.50
According to the terrier (estate register) of the Bohemian estate of Rum-
burk 1587, three demesne farms with an arable area of 200–300 hectares
employed six maids for dairy farming and as cowherds (for a total of nearly
200 cattle and 1,200 sheep), nine further maids and only three male farm-
hands (apart from the three farm managers).51 In the property of Brložec
in the Bohemian estate of Toužim in 1634, there were 32 servants on five
demesne farms, eight of which were cowherds, six were maids (who very
likely also cared for livestock) and two male servants were responsible for
building maintenance. The farms kept 100 cows and 129 calves in total.
Nine further servants were the administrators of the farms and their wives.
In fact, there were only three farmhands for the work in arable agricul-
ture (for a cultivated arable area of more than 300 hectares). In Toužim
in 1638, there were eight shepherds, eight cowherds and 19 maid servants

49 Harnisch, Boitzenburg, 170; Harnisch, “Probleme”, 263; Harnisch/Heitz, “Einleitung”, 17;

Melton, “Gutsherrschaft”, 329; North, “Lohnarbeit”, 17; Peters, “Ostelbische Landarmut”,
280nn. For demesne operation with servants and day labourers in Poland and Hungary
see Geremek, “Problem”; Izydorczyk-Kamler, “Lohnarbeit”; Kamler, Folwark, 65nn; Melton,
“Gutsherrschaft”, 304–305, 314; Pach, “Labour control”; Wyczański, Studia, 104.
50 Tiray, “Jak se žilo”, hier 221–22.
51 Hausarchiv der Regierenden Fürsten von Liechtenstein im Schloß Vaduz, Urbar 74;
Richter, Geschichte, 33, 36.
204 markus cerman

out of 64 servants in total. Four servants were explicitly mentioned to be

responsible for the oxen. Apart from the administrators, there were only
seven male servants for a demesne area of 480 hectares. Livestock con-
sisted of 164 cows and 205 calves.52 In the estate of Stráž p. R. in Northern
Bohemia, 63 servants were employed on six demesne farms in 1630. Only
19, less than a third, were adult male servants, whose duties could have
included work in arable agriculture, while 20 female servants must have
at least partly or perhaps even mainly cared for the livestock but for the
busiest days of harvesting. The rest of the servants was either responsible
for the management of the farms or were explicitly listed as shepherds,
cowherds etc. A very similar structure prevailed in three demesne farms in
the neighbouring estate of Lemberk. There were only eight male and three
female servants out of a total of 23 persons who were not explicitly work-
ing with the livestock or carrying out other specific duties.53 Around 1600,
there were fifty servants employed on five demesne farms in the estate of
Frýdlant situated in the same region (Table 1). From twenty male servants,
three were responsible for the carriages and four were horse servants, who
would probably only partly engage in field labour.

Table 1 Structure of the Servants in Demesne Farms, Estate Frýdlant, 1600/1601

and 1628/1629
Group of employees 1600/1601 1628/1629
Servants N (per cent) N (per cent)
Manager 4 (8.0) 6 (16.7)
Full male servants 20 (40.0) 10 (27.8)
Male youths 1 (2.0) —
Full female servants 7 (14.0) 15 (41.7)
Shepherds/cowherds 9 (18.0) 1 (2.8)
Cooks 4 (8.0) 3 (8.3)
Others 5 (10.0) 1 (2.8)
Sum 50 36
Source: SOA Děčín, HS, Kart. č. 12a;54 Kart. č. 34.55

52 Ryba, “Hospodářství”, 218–24.

53 Cerman, “Venkovské společnosti”, 372–73.
54 Daß erste jahr. Einnamb (Ausgab) register der Herrschafften (. . .) vom Tage Michaeli
des 1600 jhars (. . .) 1601, pag. 19–23.
55 Register uber alle einnamb undt ausgaben des geldes vom July bis zum 30. Decem-
ber auff ein halb jahr von der Herschaft Friedlandt 1628; Register uber alle einnamb undt
ausgaben des geldes vom 6. Januari bis zum 30. Junij a[nn]o 1629. Auff ein halb Jahr von
der Herschaft Friedlandt 1629.
 regional variation and its causes 205

A comparison with 1628/29 reveals that the proportion of female servants

grew strongly, whereas that of males decreased by more than a quarter (a
decrease even when one considers three deputy managers newly hired).
In total, the number of servants must have remained unchanged (a lower
number of demesne farms was recorded in the 1628/29 survey). Until
the first half of 1641, the number of servants on the three demesne farms
declined to sixteen (a year before, there were still 35 servants); among
them only two male hands for the oxen and two maid servants.56
In 1651, eighteen servants were employed on two demesne farms of the
estate of Frýdlant. Thus, there was no apparent reduction in the number
of servants. Apart from four female servants aged 18 to 23, there were only
three male servants who most likely were responsible for the field works.
All the other predominantly female servants were either cooks or working
with the livestock (cows, calves and pigs). The structure of the servants in
fourteen demesne farms held in feof is summarized in Table 2. The table
reflects, that apart from a group of female servants only about a third of
the workforce was mainly engaged in field labour.57
In 1730, 151 servants were employed on 17 demesne farms in Frýdlant. 38
were maids mainly occupied with rearing the livestock, a further 39 were
maids in domestic service and 18 were cooks, all of whom would not have
been called for arable duties on a regular basis. In the neighbouring estate

Table 2 Structure of the Servants in Demesne Farms, Estate Frýdlant 1651

Servants Number (per cent)
Full male servants 36 (38.3)
Male youths 17 (18.1)
Full female servants 23 (24.5)
Female youths 4 (4.3)
Shepherds 5 (5.3)
Cooks 5 (5.3)
Nannies 4 (4.3)
Sum 94
Source: NA Praha, SM R 109/45, Bl. 10.

56 SOA Děčín, HS, Kart. č. 34, Rechnungk uber die einnahmb undt außgaab der gelder
in den rentten der Graff Gallaßischen Herrschafft Friedtlandt vom 29. Decembr: 1640 bieß
zum 29. Junij 1641, expenses no. 6; Kart. č. 485, Rechnung uber die einnahmb undt außgaa-
ben deß geldes (. . .) vom 17. Martÿ bieß zum 30. Junÿ Anno 1640, expenses no. 14.
57 Národní archiv (NA) Praha, SM R 109/45, Bol. 10.
206 markus cerman

of Liberec, 24 servants were employed on demesne farms in the summer

of 1704. Two were shepherds, one cared for the horses and three were
cooks. Of the remaining individuals, 15 were maids responsible for the 168
heads of cattle.58
An analysis of 100 select demesne farms in Bohemia in 1651 revealed
that there were 194 maid servants (among them also kitchen personnel
and 72 shepherds and cowherds) among 840 employees in total. Various
categories of male farmhands formed the relative majority (44.6 per cent),
but more than four fifths of the 185 auxiliary farmhands were younger
than 20 years, more than one fifth younger than 14. They could thus not
be employed for the more heavy tasks of arable farming.59 Clearly, a more
detailed analysis is necessary to fully account for the concrete organiza-
tion of work on the demesnes.
Despite these problems, the literature frequently uses the incidence of
a low number of servants on demesne farms as an indicator that forced
labour services prevailed in operating demesne farms. Demesne farms in
Mecklenburg-Schwerin princely estates on average kept four horses and
employed only 4.5 servants at the beginning of the eighteenth century.
There were about 1.7 adult farmhands on these farms, so probably one
ploughteam could be mustered. With on average nearly seven servants
(among them 2.5 grown up males) and nine horses, demesne farms on
noble estates were better equipped. But they could usually rely only on
a much smaller number of tenant farms rendering labour rents, because
the expropriation of tenant farms was stronger on noble estates (there
was less re-settlement of tenants after 1650). In four noble administrative
districts, there were between 4.5 and 7.5 draught animals on average per
demesne farms in 1703. All of the demesne farms also employed servants
and farmhands, but it is difficult to determine how many were available
for the tasks of arable agriculture. The majority were female—the propor-
tion of female servants was higher than on tenant farms—and they were
presumably primarily employed for the livestock.60 The 1703 data prove
that demesne organization was highly adaptive to the availability of ten-
ant labour. Thus, princely estates mustered between 17 and nearly 24 ten-
ants but only 4.0 to 5.2 servants per demesne farm. By contrast, in noble

58 Cerman, “Arbeitsrenten”.
59 Further 200 employees were managers and their wives. Maur, “Gesinde nach dem
Soupis poddaných”, 122.
60 Heitz, “Die sozialökonomische Struktur”, 26–30, 42–3, 45–50, 70–3; Rudert, Guts-
herrschaft und Agrarstruktur, 120.
 regional variation and its causes 207

estates, only between seven and 9.4 tenant farmers were available, but 6.6
to 7.3 servants per demesne farm. The number of landless wage labourers,
which were attached to demesne farms and whose labour could be used
was higher in princely estates than in noble ones (3.0 vs. 2.3).61
Of more than 300 demesne farms surveyed in Swedish Pomerania in
the late seventeenth century, the average number of servants of 4.6 was as
low as in contemporary Mecklenburg. The average number of 9.6 horses
and 6.9 oxen (registered for 429 and 237 farms respectively) shows at
least that demesnes could supply ploughteams by themselves, which
were then probably operated by tenant smallholders who had to provide
manual labour as part of their tenurial forced labour service obligations.62
Already in the sixteenth and early seventeenth century, draught power
and servants were kept in demesnes in order to avoid overburdening ten-
ant farmers, who were theoretically obliged to perform unlimited service
from 1616 onwards. Yet, differences were strong in this respect: In eight
princely estates, demesne farms did not have horses as draught animals
at all, whereas in six noble estates the average number was 7.7 horses per
farm. Again, this may have been for lack of alternatives as many of the
smaller (noble) estates probably did not have enough tenant holdings to
work their demesne farms with labour rents.63
Even within a relatively small sample of estates of a case study for
Upper Silesia, there were profound differences in the management, as
some demesne farms were operated entirely with their own draught ani-
mals and wage labour. Thus, in the estate of Łubowice, 31 smallholders
and cottagers were employed and the demesne farm held 30 draught ani-
mals for 189 hectares of arable. In other estates, the number of tenant
farmers was not large enough to perform all the necessary tasks of arable
farming on the basis of forced labour services. Forced labour was thus
complemented by keeping draught animals directly on demesnes and by
the wage labour of cottagers.64

61 Rudert, Gutsherrschaft und Agrarstruktur, 101–106, 185–86.

62 Schilling, Schwedisch-Pommern, 24–5, 52, 56, 58, 100–101; Schleinert, Gutswirtschaft,
79, 96–7, 283.
63 Mager, Geschichte, 88; Schleinert, Gutswirtschaft, 92–3, 95–6, 130–37, 283; for the
Duchy of Szczecin see Wachowiak, Gospodarcze, 298–311; Wachowiak, Gospodarka,
60–92.
64 Stutzer, “Ertrags- und Lohnverhältnisse”, for the employment of landless in noble
demesne farms in early eighteenth-century Mecklenburg see Heitz, “Die sozialökonomi­
sche Struktur”, 32–4, 52–4.
208 markus cerman

Concentrating on the ‘classical’ period of the rise of the demesnes in

Poland in the sixteenth and early seventeenth century, the organization
of work was analyzed also in Polish studies. From the evidence, demesnes
entirely based on labour services were unlikely for Poland, either.
A case study of several hundred royal, church and noble demesnes in
three governorships in Little Poland confirms that the majority employed
servants and seasonal wage labourers in the period 1525–1642. In some
parts of the kingdom, more than 50 per cent were operated in direct
management without corvée (Table 3). The proportion of royal demesnes
entirely without servants was zero in 1525–53, 13 per cent in 1560–90 and
seven per cent in 1611–42. The highest share occurred in the governorship
of Lublin with 20 per cent in 1560–90 and 17 in 1611–42 respectively. In
the same periods, the proportion of noble and church demesnes oper-
ated entirely without wage labour was 24 and eleven per cent.65 Also in
other regions, wage labour for demesne farms was important and there
was some equipment with draught animals.66 Many demesne lords only
owned very small estates and had very little farmland to be worked. Wide-
spread commutation of forced labour services was still possible in Little
Poland in this period. In 1564, 30 per cent of the villages of royal estates
in the Cracow region paid money in lieu of services and a further 40 per
cents had only rent duties in cash and kind.67
It is also necessary to consider changes over time in the balance of the
use of wage labour and forced labour services. For instance in East Prussian

Table 3 Share of Labour Services on Polish Demesnes, 1551–1580

(in per cent of demesne land operated)
Region Operated with labour Operated in direct
services management
Mazovia 79 21
Little Poland 60 40
Western Great Poland 63 37
Eastern Great Poland 48 52
Average 63 37
Source: Adamczyk, “Weltmarkt”, 81; Wyczański, Studia, 109.

65 Adamczyk, Polen, 191; Izydorczyk-Kamler, “Die Lohnarbeit”.

66 Kamler, Folwark, 37–40, 52–76; Wyczański, Studia, 104, 122–38, 163–66; Żytkowicz,
Studia, 229–41; Żytkowicz, “The peasant’s farm”, 140–41, 143.
67 Trzyna, Położenie, 222–36.
 regional variation and its causes 209

princely estates, the formal increase of forced labour services after 1550
was accompanied by a shift towards using the manual labour of small-
holders and cottagers at a larger scale.68 With population growth and
stagnant agricultural prices demesne expansion stopped around 1600.
As a consequence, services which were no longer needed were abolished
and converted into cash rents. Also because of slacking grain demand the
estate administration tried to increase income from cash rents and used
the proceeds to pay waged servants on the demesnes where employment
grew.69 A change of preferences between wage and forced labour occurred
fairly frequently as the case of the estate of Ostróda illustrates: corvée was
converted in 1615 and then reintroduced in 1622. These developments fol-
lowed economic considerations regarding grain price trends. The area
of demesne arable farms reached a peak in the 1570s, was considerably
reduced until the 1590s, rose again and then decreased dramatically after
1610. However, when population levels fell, labour rents of the remain-
ing tenants could increase. Ten princely demesnes could use the labour
rents of 1,355 tenant farmers in 1610. In 1683, only 530 farmers were still
available. Simultaneous with the extension of their land, these demesne
farms reduced the number of continuously employed servants over the
course of the seventeenth century from 409 to 285. Thus services must
have increased also for this reason.70
The estate of the University of Greifswald is another example to illus-
trate major discontinuities in the estate administration and shifts in the
views about efficient management. With a major expansion of demesne
farming after 1620 the management shifted from being based primarily on
forced labour services of tenant farmers to introducing cash rents in lieu
of labour and dividing demesne farmland among tenant farmers between
1670 and 1750. Then after 1750, tenant farms were expropriated, demesne
farms rebuilt and labour rents re-introduced. This form of demesne oper­
ation lasted until the mid-1770s.71
The phenomenon observed above for seventeenth-century Prussia that
desertion of tenant farms increased the average burdens of the remaining
tenants also occurred in seventeenth-century Poland. In a case study on
the district of Korczyn in Little Poland, the average number of service

68 North, Amtswirtschaften, 80–1; North, “Lohnarbeit”, 12–4, 16.

69 North, Amtswirtschaften, 77–8, 81, 100; North, “Lohnarbeit”, 13–4.
70 Kaak, Gutsherrschaft, 284–87; North, Amtswirtschaften, 77–8, 81, 100; Wächter, Ost-
preußische Domänenvorwerke, 12, 56–9, 61, 64, 71–4, 143, 147, 153–85.
71 Kaak, Gutsherrschaft, 167–70.
210 markus cerman

workdays per household increased from 1.87 per week in 1572 to 2.25 in
1616/1629 and 2.54 in 1660. It was due to a declining number of tenant
farms and not due to further demesne expansion or a shift towards a rela-
tive increase in the use of forced labour in operating demesne farms.72
In the Brandenburg estate of Boitzenburg the lack of tenants after 1650
did not allow for an intensification of the demesne economy. Instead
of leaving the deserted farmland unused, however, it was consolidated
into nine new demesne farms until 1700, which were leased out to single
tenants, who ran them as Eigenbetrieb, i.e. with their own inventory and
with wage labour. When the demesne lord returned to demesne farming
in the eighteenth century after the population and agrarian prices had
consolidated, some of the farms continued to be managed independently,
while others started using forced labour services.73 It was on princely
estates of the region, where a rise of labour services was attempted in this
period.74
In many regions of eighteenth-century East-central and Eastern Europe
a shift towards reducing forced labour services for operating the demesne
economy can be observed. While this was often due to an abandonment
of demesne farms and their conversion into tenant farmland or due to a
temporary reduction caused by agricultural business cycles, in many other
cases demesne farms used servants and wage labour to a growing degree.
This was caused by population growth, the increasing resistance of vil-
lagers to accept new forced labour services resulting from the intensifica-
tion of demesne farming (partly caused by agricultural innovations) or by
landlord awareness that further burdens would threaten the livelihood
of tenant farms and thus undermine the whole system. In the Branden-
burg estate of Badingen, for instance, eight of eleven villages had to render
labour services in 1728. Until the end of the eighteenth century, services
per holdings decreased considerably in four villages (by between 15 and
80 per cent from an initially very high level of 298 days per year). Overall,
total available draught and manual labour service days fell from 24,615
per year in 1728 to 19,302 in 1763 and 11,457 in 1803 (nearly 55 per cent).75
When demesne farming expanded due to increasing grain demand, wage

72 Muszyńska, Gospodarstwo, 45, 69–72; Wyczański, Studia nad gospodarka, 106, 120–22;
see Inglot, Historia, 264, 363.
73 Harnisch, Boitzenburg, 137, 148–49, 158–60, 169; Kaak, Gutsherrschaft, 271–73.
74 Enders, Uckermark, 343–44, 373, 466–67, 469, 518–25.
75 Vogler, “Entwicklung”, 156–62.
 regional variation and its causes 211

labour was increasingly used in Brandenburg. In the estate of Stavenow

after 1763 this was promoted by the fact that the local tenant farmers sim-
ply refused to obey the verdicts of the royal chamber court and did not
perform the forced labour services they were ordered to. After two dec-
ades of costly legal battles over the extent of the services, the court had
ruled in favour of the lord—but in vain.76 Similar developments and more
conflicts occurred elsewhere, when rising grain prices caused a further
intensification of demesne activities.77
Also in eighteenth-century Upper Lusatia, demesne lords increasingly
used the wage labour of the rising number of cottagers to work their
demesnes78 along with the existing forced labour services. While forced
labour services were already abandoned in favour of cash rents in west-
ern Holstein in the fifteenth century, in the eastern parts, where demesne
farming dominated, reforms started early in the eighteenth century.
Under the influence of low grain prices demesne farming contracted and
103 demesne farms and estates had abolished forced labour services for
good by the 1720s. Serf resistance occurred where pressures grew again
when agrarian prices picked up after 1720/1730.79 Leaseholders of Danish
demesne farms did not invest in their own stock of draught animals, but
hired tenant farmers and smallholders to work for wages in addition to
what could be provided by forced labour services.80 In other areas, the
growing number of demesne leaseholders were simply denied access to
villagers’ forced labour services. A mixed regime is also obvious in East
Prussia, where in the eighteenth century only a part of the demesne farms
were operated entirely with forced labour services.81
Demesne lords often had no use for forced labour services in excess of
their demands, which automatically occurred as a consequence of popu-
lation growth, and thus an ever growing share of tenant households paid
money instead, which was then used to employ wage labour. In the North-
ern Bohemian estate of Frýdlant, around thirty per cent of the households
paid cash in lieu of services in the period 1720–1740, while in the neigh-
bouring estate of Liberec only 30 per cent of households were called to

76 Hagen, “Village life”, 181; Hagen, Ordinary Prussians.

77 Harnisch, Boitzenburg, 195; Harnisch, “Probleme”, 263–65; Kaak, Gutsherrschaft,
274–75; Kaak, Bauern, 167–200, 291–304, 338–40; Peters, Lebenswelten.
78 Boelcke, Bauer; Kaak, Gutsherrschaft, 264.
79 Kaak, Gutsherrschaft, 279; North, “Entstehung”, 234–40; Peters, Lebenswelten, 679–80,
715–17.
80 Christiansen, World, 270.
81 Aubin, Geschichte, 157, 166.
212 markus cerman

their full service duties in 1703 and a further 44 per cent presumably to
some of their duties. While these are significant proportions, the pay-
ments only represented between three and less than six per cent of total
seigniorial income in Frýdlant until 1765, casting doubt on the general
applicability of the theory that such payments attracted demesne lords
to abandon the commercial demesne economy in ‘proto-industrial’ areas
of demesne lordship.82 By contrast, in areas, where the demesne econ-
omy had not fully developed in the first place, resulting payments in lieu
of labour services were considerable, sometimes representing a quarter
of total landlord income, as seventeenth- and eighteenth-century Lower
Austrian examples show. While the value of forced labour services was
estimated at 10 per cent for estates of the Lower Austrian princely cham-
ber in 1570, they represented 14 to 21 per cent of total seigniorial income
of noble estates in 1750/51, half of which came from cash payments in lieu
of services not needed.83
The increasing use of wage labour also in regions where forced labour
services were high such as in Poland, Mecklenburg or in the Brandenburg
Uckermark finally resulted in labour rents being abandoned or becoming
a marginal phenomenon even before their legal abolition.84 Sometimes,
there was strong competition for wage labour, which was also needed in
tenant households.85 Like many other features, this was neither a uniform
nor a linear trend, as Zealand villagers experienced, who faced increases
in labour services as late as in the 1790s, partly because demesne farms
had been continuously extended over the previous two decades.86 In the
estate of Boitzenburg in the Brandenburg Uckermark, two thirds of the
demesne farms and three quarters of the demesne land were operated
in Eigenbetrieb with wage labour just before the agrarian reforms in 1800,
while the others were run on a mixed basis with slightly more than 50
per cent of the total work necessary being supplied by forced labour ser­
vices.87 Thus, from 1750 at least there was a widespread tendency among
demesne lords and leaseholders of demesne farms alike to turn to using
more wage labour.88

82 Cerman, “Arbeitsrenten”, 278–82.

83 Knittler, Nutzen, 42–7, 57, 217–18, 222–33.
84 Hagen, “Village life”, 178; Peters, Lebenswelten, 641–42.
85 Lubinski, “Work”.
86 Christiansen, World, 269.
87 Kriedte, Peasants, 112–13; Schmidt, Leibeigenschaft, 104.
88 Bentzien, “Fortschritte”, 151.
 regional variation and its causes 213

A survey of 657 demesne farms in Swedish Pomerania c. 1700 showed

that 421 (61.1 per cent) were operated with forced labour services and 165
(25.1 per cent) were run as Eigenbetrieb (for 71 further farms there was
no information). In particular small demesne farms were more likely to
be worked with their own means. (Eighty per cent of the demesnes were
smaller than 100 hectares.)89 In the Pomeranian case, studies have been
cautious to interpret this relatively high proportion in direct management
as a result of economically innovative behaviour on behalf of demesne
lords. It can rather be seen as a reaction to the lack of tenant farmers
due to expropriations and desertions, which forced lords into this mode.
On average, each demesne farm could only draw on 5.5 tenant house-
holds to perform forced labour services. Rather than agrarian progress and
awareness of the weaker productivity of forced labour services, it was the
extreme nature of demesne lordship that caused this result.90 Given that
the majority of demesne farms operated in this way were leased out, the
situation may also have been caused because leaseholders were denied
access to forced labour services.
The described tendencies in the operation of demesnes therefore oc‑
curred for a mix of reason and a belief in a model of linear progress fitting
all regions of demesne lordship would be wrong. Despite the trend, even
in the period of peasant emancipation demesne lords were often not con-
vinced of the superior productivity and organization of wage labour when
they compared the different forms of work in running their estates. In
the 1850s, Estonian demesne lords gave the lack of qualification, unstable
employment situation and high fluctuation of workers as reasons that their
expectations in the employment of wage labour were disappointed. Apart
from paying wages, the necessary investments in draught animals often
pushed demesne lords into delaying the introduction of wage labour.91

Conclusion

The development of the rural economy of early-modern East-central and

Eastern Europe was traditionally associated with the omnipresence of
labour constraints in ‘serf’ societies. Indeed, where labour rents or forced

89 Schilling, Schwedisch-Pommern, 49–50; Rudert, Gutsherrschaft und Agrarstruktur,

118.
90 Schilling, Schwedisch-Pommern, 54, 64, 108.
91 Dorošenko, “Herrenhof”, 199; Kaak, Peasant, 72–7, 89–91.
214 markus cerman

labour services (corvée) existed, they belonged to the strongest burdens

tenant households faced vis-à-vis their landlords. Arguments, conflicts and
resistance against forced labour services, their increase or the way they
were commanded occurred almost on a daily basis. In addition, also the
obligation to send teenagers into life-cycle service on demesne farms
(rather than having them available at home) was frequently experienced
as a serious burden. The economic consequences of such services in terms
of loss of workforce or draught power available to the tenant agricultural
economy has often been analyzed and especially regular burdens beyond
two or three days a week meant that households had to maintain addi-
tional ploughteams and farmhands.
By contrast, this paper focussed on the problems of such generaliz­
ations that became apparent in most recent research. It is now clear from
the evidence that the establishment of forced labour services was far from
being a one-sided process driven by seigniorial interests alone. Secondly,
it is not always possible to distinguish clearly between individual forms
of labour constraints and ‘free’ wage labour in practice. Thirdly and most
importantly, the early modern commercial demesne economy could not
and did not rely on forced labour services or other forms of labour con-
straints alone. Rather, the organization of work was based on a complex
mixture of forced labour services and wage labour, whose respective
importance could shift over time and between places. Also with regard
to the organization of work in the demesne sector, therefore, East-Elbian
rural societies of demesne lordship seem far from being a uniform area of
a (‘second’) ‘serfdom’.
 Rights and Bondage in Russian Serfdom

Alessandro Stanziani

Introduction

Already in 1921, Marc Bloch warned against the use of the word “serf” and
the expression “serf of the glebe”. From this point of view, the “serf of the
glebe” was used to identify a largely stylised feudal system and to oppose
it to an equally stylised liberal economic system.1
Ever since, medieval studies have widely adopted this conclusion for
France and Britain.2 More recently, a similar reassessment has been
made for the German second serfdom (notably in criticizing Brenner’s
argument).3
Quite surprisingly, discussions on Russian serfdom have focused on its
origin (the State4 and/or the landowners5) and profitability6 not so much
on the interplay between legal rules and economic activity. The most
remarkable contributions along this path are those of Hoch, Wirtschafter,
Moon and Melton who have strongly revisited the simplistic definition
and functioning of Russian “serfdom”. Its dynamics and rules have been
questioned by maintaining that serfdom was never officially introduced7
and that it was much more a set of practises than a system.8 We are will-
ing to push this same way and analyzing rules defining the legal status
of actors and their implementation. We shall show that in Russia, in the
decades preceding the official abolition of serfdom, a quarter of the whole
peasantry, that is, a half of private peasants changed their legal status and

1 Bloch,  “Serf de la glèbe.”

2 Bois, La crise; Duby, Les trois ordres; Bonnassie, From slavery; Scott, The peasantries.
3 Brenner, “Agrarian class structure”; Aston, Philpin, The Brenner debate; North, From
the North Sea; Melton, “Population structure”; Raeff, The Well-ordered; Shmidt, “über die
bezeichnung”; Hagen, Ordinary Prussians.
4 Blum, Lords; Gershenkron, Economic backwardness; Crisp, Studies; Bartlett, “Serfdom”;
Hellie, Enserfment; Kolchin, Unfree labour; Bush, Serfdom; Field, The End.
5 Hoch, Serfdom.
6 Koval’chenko, Russkoe krepostnoe; Blanchard, Russia’s age; Moon, The abolition; Den-
nison, “Did serfdom matter ”; Melton, “Proto-industrialization”; Melton, “Enlightened sei-
gnorialism”; Domar and Machina, “On the profitability.”
7 Barlett, “Serfdom”.
8 Examples of this in Hoch, Serfdom.
216 alessandro stanziani

became state peasant, free artisan, free worker, etc. As David Moon has
already shown, this means that the official abolition of serfdom has to be
put into a longer process, starting at the end of the eighteenth century
and that it will be not completed until WWI, when peasants got access
to private property while on the labor market criminal punishments were
repealed. However, we will show that shifts in the legal status of the peas-
antry before 1861 were not a purely administrative act (according to a top
down reform), but they accomplished through judicial and/or admin-
istrative procedures. Peasants and artisans were able to mobilize rules
originally conceived for nobles in order to prove their claims on estates.
Juridical conflicts within the elites and the reformist attitude of some tsar-
ist leaders made this issue possible.
Our work is based upon archival and published sources. We made use
of Russian archives, mostly archives of ancient Russia (RGADA) and local
archives in Moscow (TsGIAM)(with local law courts decisions). To this
one has to add the impressive collection of the Russian laws, decrees
and jurisprudence9 which gives the opportunity to capture the incredible
number of rules adopted on our topic, their connection with rules con-
cerning nobility and family, but also the main jurisprudential decisions
adopted by high courts and Senate in Russia.

Property Rules and the Legal Status of Russian Peasantry

The ways economic and social actor appropriate legal rules make the object
of huge works in sociology, economics, history and anthropology, and, of
course, in legal studies. In social and economic history, this approach has
been developed for different countries and different periods. Historians
of Russia have equally developed an ever increasing bibliography on this
topic.10 As regards “serfdom”, it is widely accepted that peasants had no
but a few legal rights until the reforms of 1861–1864. We argue that, if it
is true that peasants had not the right to attack masters in justice, their

9 Polnoe sobranie zakonov Rossiskoi Imperii (hereafter PSZ) (Whole collection of the
laws of the Russian Empire), three series: I: 1649–1825, 46 vv. Saint-Petersburg, 1830. II:
1825–1881, 55 vv. Saint-Petersburg, 1830–1884; III: 1881–1913, Spb, 33 vv 1885–1916.
10 Wirtschafter, Structures; Wirtschafter, Social Identity; Martin, Law and custom. Most
of the available literature on legal action in Imperia Russia focus on post-1864 period: 
Wortman, The development; Wagner, Marriage; Burbank, Russian peasants; Frank, Crime;
Solomon, Reforming; Frierson, “Rural Justice”; Frierson, “I must”; Czap, “Peasant-class
courts; Kritika, special issues, 6,1 (2005) and 7,1 (2006); Pravilova, Zakonnost”.
 rights and bondage in russian serfdom 217

legal rights increased not only in 1864, but largely before, when volost’
(district) courts where institutionalized in the 1775. These courts were cer-
tainly limited in their scope and rights, as landlords controlled them. Most
important, peasants had not the right to attack masters for physical pun-
ishment. At the same time, peasants could contest the rights of the estate
owner not only in requesting their labor service or money fees, but, even
more radically, her/his title of ownership. How was this possible?
In order to answer this question, we need to carefully examine the legal
entitlements of both landlords and peasants. Despite the lack of formal
rules, serfdom actually existed and this reflected the power of nobility
over state rules.11 Rules adopted in Russia since the middle of the six-
teenth century up through the famous Ulozhenie (Statute) of 1649 linked
the binding of the peasantry to the establishment of a cadastre in order to
improve fiscal income and military conscription, and to settle estates for
which different categories of nobles, the crown, the Church and monas-
teries advanced claims.12 Limitations to peasants’ mobility were not a goal
in itself but they entered these relationships between different groups of
landowners and the State.13 Nor peasants’ indebtedness14 nor State expan-
sion15 was the main cause of such a complex process.16 Indeed, most of the
nobility was hostile to territorial expansion, which was viewed a source of
instability and was accused to reduced available manpower.17
Thus, measures for delimiting property were not only dictated by tax and
military requirements of the State,18 but overlapped with a ­significant redef‑
inition of the relationships between social groups and the State, namely,
the role of land property as a social and political marker.19 The Russian
elites welcomed the request of the provincial nobility to legitimate their
properties in order to secure an important ally in the fight against the

11 Wirtschafter, Structures.
12 Eaton, “Cadasters”; Voznesenskii, Materialy.
13 Cadastres documents (Pistsovye knigi) are largely available in RGADA in a large
number of fonds, among which f. 1239, opis’ 3, chast 17, 69–72, 74, 76, 86–87 (cadastres
of Moscow district, 1674–1681); fond 396, opis’ 2, ch. 5 (1616–1732), Smolensk, and several
other districts; fonds 1209 (prilozhenie arkhiv premikikh votchnikh del, 1565–1692), opis’ 1,
ch. 1–3, opis’ 2, ch. 1–2, opis’ 16–72.
Cadasters documents published in Pistsovye knigi Moskovskogo gosudarstva (ed. N.V.
Kalachov), St. Petersburg, 1872 and 1877.
14 Blum, “Prices”; Man’kov, Tseny. Smith, Peasant farming.
15 Field, The End; Perdue, “Military.”
16 Hellie, Enserfment.
17 Jones, “The nobility.”
18 Fuller, Strategy; Leonard, Reform.
19 Moon, The Abolition. Zaionchkovskii, Otmena.
218 alessandro stanziani

boyari (high aristocracy). Until the first half of the seventeenth century,
it was assumed that the binding was temporary; by the 1630s landlords
could eventually allow their peasants the right to move, as expressed
in document they signed (the otpusknaia gramota).20 This did not stop
peasants from moving without authorization. Against this movement,
heartland landowners and provincial cavalrymen were those who most
complained of the peasants’ flying. Petitions multiplied between the 1620s
and the 1640s and the central state responded by lengthening the time to
recover fugitives from five, to nine to fifteen years (1637, 1641, 1645, 1648).
However, this process was by no means simple, which is evidenced
in numerous legal disputes and petitions21 drawn up by noble families
against other claimants on their properties (other nobles, merchants,
boyari, etc.).22 This game was complicated by the fact that landowners
were ready to keep runaways peasants and that the state was unable to
enforce the related rules.

This is where the famous Ulozhenie intervention of 1649 entered in. This
document actually sought to find a solution to the same problems already
regulated in previous rules, that is, the legal identification of a landowner
and fair cooperation within this group.23 In particular, many articles of
the section 11 were devoted to the documentary evidence the nobles had
to provide in order to recover “runaways”. Landlords who accepted to hide
fugitive peasants were sanctioned even more than peasants themselves.24
There was no time limitation to recover fugitive peasants.
However, as the litigations among landowners and between them and
urban merchants clearly show,25 the 1649 did not solve this problem. In
the following years and decades, the great landlords became notorious
for luring peasants away from smaller estates.26 This game became even
more complicated when urban elites (ford fiscal needs) and peripheral
authorities (interested in increasing the local population) pushed to keep
the “runaways” in place. Negotiations on this occurred at legal, adminis-
trative and political level.27 During all the eighteenth century, several new

20 Hellie, Enserfment: 108.

21 RGADA, fond 615 (krepostnye knigi mestnyjh uchrezhdenii XVI–XVIII v), opis’ 1, sev-
eral dela.
22 Crummey, “Sources”.
23 Man’kov, Razvitie.
24 Ulozhenie, chapter 11, n. 10.
25 RGADA, fond 615, opis’ 1; Fond 294, opis’ 1–3.
26 Melton, “Russian peasantries”: 239.
27 Kahan, The plow: 76–77.
 rights and bondage in russian serfdom 219

rules were adopted to assess these claims. As in the previous centuries,

all these rules took a clear orientation, that is, limiting the definition of
“hereditary nobles” (in opposition to service nobles on the one hand, to
non-noble landowners on the other hand) and providing the ground for
a “fair competition” among entitled landowners. The definitions were
made gradually: in 1727, servants and peasants were excluded from civil
service; a decree in 1730 prohibited servants, servants in monasteries, slugi
and peasants from acquiring land, whether through inheritance or not.28
Next, in 1739, soldiers and lower echelons of the administration who did
not own estates lost the right to acquire vol’nye liudy (house servant) and
in 1746, the clergy, merchants, urban guilds, Cossacks and raznochintsy
(people of various ranks) were prohibited from acquiring and transferring
inhabited estates and establishing krepost’ (deed) relationships.29
Several rules adopted between 1754 and 1758 prohibited non-nobles from
owning inhabited estates. Finally, in 1762, the factory owners were prohibited
from buying or transferring estates with urban or rural workers.30 This did
not prevent several merchants and manufacturers from acquiring populated
estates and firms,31 so that in 1812, the Senate was forced to confirm that the
decree of 1758.32 Least but not the last, the evolution in the legal status of the
peasantry was strongly affected by the secularization of ecclesiastic lands in
1763, which placed 20–25% of the peasants under the State authority.
The restrictions imposed on non-nobles regarding access to inhabited
estates were followed by numerous petitions drawn up by non-noble offi-
cers, manufacturers and merchants demanding that a distinction be made
between owning workers (urban or rural) and entire villages, which alone
should be prohibited to non-nobles.33 This explains why numerous court
proceedings also opposed landowners to the administration or landown-
ers to each other concerning the validity of the noble title used by one or
the other.34 Disputes occurred as well when estates were transferred or
dowries were constituted, in cases of inheritance35 and finally, in view of

28 PSZ (I), vol. 8, n. 5633. RGADA, fond 342, opis’ 1 delo 37, chast’ 2 and 3.
29 PSZ (I), vol. 12, n. 9332, 9367.
30 PSZ (I), vol. 15, n. 11490, vol. 16, n. 11638.
31 RGADA, fond 294, opis’ 1, dela 385, 386; opis’ 2, dela 135, 145, 168.
32 PSZ, II, vol. 3, n. 2378; RGIA, f. 1149, opis’ 2, delo 44.
33 Coquin, La grande commission: 110 and 161–163.
34 TsGIAM, fond 54, opis’ 1; RGADA, fond 615, opis’ 1, several cases; RGADA, fond 342,
opis’ 1, dela 64, 119, 120.
35 Several cases in TsGIAM, fond 54. and in RGADA, fond 1209, in particular opis 84.
On this Jones, The emancipation, Wirtschafter, Structures: 71–74; Marrese, “The enigma”;
Wagner, Marriage.
220 alessandro stanziani

admitting the landowners of the newly annexed regions such as Poland to

the rank of noble of the Russian empire.
These already difficult relations were dramatically sharpened by the
1762 emancipation of the nobility from compulsory service; this made the
traditional justification for serfdom disappear; nobles had no more to serve
the tsar and they were therefore supposed to manage their estates. In this
context, “entitled” nobles seemed to be the more and more interested in
selling properties to non-nobles elites, to present a loan as a selling or
viceversa, mainly for fiscal reasons, but also in response to financial diffi-
culties, conflicts within the family, etc.36 Thus, in 1799, a merchant bought
an estate inhabited by peasants from a noble landowner; the transaction
remained valid as long as the merchant did not try and transform his pos-
session into full ownership and therefore pass it on to his children. For
in such cases, other interested parties—nobles as well as the peasants
living on these estates—could sue. To return to the example of the mer-
chant: he gave part of his estate and serfs as a dowry for his daughter
(1810) and bequeathed the rest to his son (in 1817). The peasants protested
and brought a lawsuit against him, which they won, and the decision was
confirmed by the Senate.37 The reason for the decision was that the owner
did not have the right to transfer the estate acquired in 1799.
Non-nobles were not the only ones seeking to get around these provi-
sions, however; noble landowners were also interested in doing so for tax
reasons or simply to gain access to resources. The solution then consisted
in presenting the sale of noble estates (or a part of them) to non-nobles as
loan contracts; which was not necessarily false, insofar as the “merchants”
often financed indebted nobles and the latter were unable to repay their
debts. It was a form of mortgage. What would be normal in mortgage mat-
ters in Europe was now prohibited in Russia, for mortgages opened the
door to the acquisition of the estate by the creditor; in Europe, this acqui-
sition resulted in full ownership, i.e. transferable, which was impossible
in Russia. One solution consisted in transferring the indebtedness of the
noble and the seizure of the estate (non-transferable) to the heirs of both
the formal owner (noble) and of his creditor (non-noble). To prevent this
from happening, a law of 1816 (supported by both part of the nobility and

36 Senat still records a great number of these transactions in 1816. Arkhiv gosudarstven-
nogo soveta: 4, vyp. 1, ch. 2, 253–258.
37 RGIA, fond 1149, opis’ 2 delo 20.
 rights and bondage in russian serfdom 221

Russian officialdom) prohibited the recording of letters of credit in the

name of persons who were not entitled to own and transfer peasants.38
Despite these provisions, transactions between nobles and non-nobles
continued. To escape the law of 1816, non-nobles used verbal agreements
in order to secure peasants’ work or to grant loans to nobles, etc.39 These
offences multiplied in the ensuing years; numerous decrees tried to limit
the practice of verbal agreements between nobles and non-nobles on the
subject of peasants.40
Still in 1836, the “personal” nobles (non-hereditary) were confirmed in
their right to own inhabited estates but not to transfer them.41 These pro-
visions were followed by an increasing number of disputes among nobles,
service nobles and hereditary nobles, or between nobles and the adminis-
tration. According to the statistics of the Ministry of Justice, in 1845 alone,
6,400 requests for confirmation of the title of “noble” were submitted, only
half of which were validated.42
To a given extent, these nineteenth century rules defining who were
legally entitled in possessing and transmitting populated estates were
perfectly coherent with sixteenth through eighteenth century rules. They
confirmed the distinction between “entitled” owners and the others. At
the same time, they sought to answer to a different dynamic, that is, the
interest of some groups of “entitled” landowners in establishing transac-
tions with other social groups.
The sale of serfs has therefore to be explained in this context made
of official rules designing those who were entitled to own and transmit
inhabited estates. Many literary sources and a good number of other
political or administrative sources mention the sale of serfs. Classical
references in historiography such Blum, Madariaga, Hellie, just to men-
tion a few of them, refer of the market for serfs. Demographers currently
speak of the prices for serfs,43 while economic historians have also tried
to evaluate the trend in serfs’ prices.44 Still, most of these works identify
the prices of serfs as the result of statistical regressions and/or proxies of
other variables (land prices, fee to pay for mobility, dowries etc.). This

38 PSZ (I), vol. 33, n. 26469.

39 RGIA, Fond 1149, opis 2, several dela.
40 Decrees of 1825, 1828, 1840, 1847. PSZ (I) vol. 40, n. 30407, psz II, vol. 3, n. 1696, vol. 15,
n. 13051.
41 PSZ (II), vol. 11, n. 9203.
42 Ministerstvo iustitsii, Otchet: xix.
43 Bushnell, “Did serf owners”.
44 Mironov, The social history.
222 alessandro stanziani

necessity is partially due to the fact that sources used in this literature are
mostly official sources (ministerial reports and data, reviziia, etc.).45 This
actually is an intriguing puzzle: why did official sources did not speak of
the price of serfs?
Estate sources provide a first answer: registers, accounting books, etc.
often indicate the price of serfs; at the same time, when addressed to offi-
cial institutions, these same estate documents often spoke much less of
price of serfs or of sale of serfs than of fees, dowries, transfers, gifts, etc.46
Notarial records offer a quite similar attitude.47 The immediate comment
these findings lead to suggest is that for fiscal and/or legal reasons the sell-
ing of serfs was made under the cover of other formal practises; however,
such a legal fiction would not deny the fact that a real market for serfs
existed in tsarist Russia. As Isabel de Madariaga put it, it was a de facto if
not de jure slave trade.48
However fiscal reasons alone cannot explain the lack of serf prices and
a market for serfs in official sources. Another reason for that, was that
Petr I already disliked the sale of individuals, in particular the break up of
families. It imposed a tax of 10% of the price in roubles on sales without
land (decrees of 1701, 1705, 1720). This was confirmed in 1773.49
Last but not least, the lack of official legal transactions concerning serfs
was that all the sets of rules we have mentioned strongly limited those
who could own and transmit inhabited estates. They forbade the sale of
peasants without lands, of peasants of mortgaged estates, of those of land-
owners with no heirs, etc.
At the same time, the fact that so many transactions of serfs were made
despite official rules, was not without consequences about the effective-
ness of these practises. Members of noble families, state administrations,
but also merchants and ultimately peasants contested titles of ownership.

Changing Legal Status through Judicial Proceedings

In 1836, while confirming that personal, non-hereditary nobles were not

entitled to transfer their peasants, a new law encouraged them to redeem

45 For example, Blum, Lords, Domar and Machina, “Profitability”, and Mironov, The
Social.
46 RGADA, fond 1287 Sheremetevy, opis’ 3 chast 1, delo 582; fonds 615, opis’ 1. Fonds
615, opis’ 1.
47 RGADA, fond 615, opis’ 1; fond 1273, opis’ 1.
48 Madariaga, “Catherine II”, in particular, 38.
49 PSZ, 19, 13950, 21 febr. 1773.
 rights and bondage in russian serfdom 223

those peasants in exchange for a compensation paid in part by the State

and in part by the peasants themselves.50 The adoption of this law was
followed by accelerated voluntary “emancipations”, which also occurred
during the years prior to the official abolition of serfdom.51

However, changing legal status also resulted from the increasing number
of legal proceedings brought by the peasants themselves against estate
owners. In general, illegal enserfment was an offence under the criminal
code; while before 1845, the punishment was limited to a fine after that
date the usual penalty was imprisonment, in addition to emancipation
without compensation.52 Even more than the statutory penalties, legal
proceedings and arguments may legitimately be seen as conditioning
peasants’ chances of winning their cases. Thus, until the 1770s, there were
no special courts in Russia for peasants, who were forced to go through
nobles and their courts to request emancipation, which considerably
reduced any attempt to initiate proceedings. The situation changed in
1775 with the introduction of peasant courts, accompanied by a clear-
cut legal separation between ownership of things and rights over human
beings. These courts and their scope of jurisdiction were to be expanded
by reforms of 1837–1841, when peasants were given increased rights and
the possibility of denouncing invalid ownership of estate and, thus, land-
owners’ entitlements.53
Throughout the first half of the nineteenth century, conflicts between
peasants, peasant-workers and estate owners on both peasants’ obliga-
tions and owners’ titles of ownership were on the agenda of the courts.
Usually, peasants who won were reclassified either as state peasants or as
urban residents.54 The arguments evoked by the plaintiffs testify first of
all to the peasants’ knowledge of the legal rules governing the possession
of estates and that we mentioned earlier. For example, in many cases,
plaintiffs emphasised that the transfer of the estate took place through
the intermediary of a non-noble (which was prohibited, as we have seen),
or by an emancipated peasant; in other cases, plaintiffs demonstrated
that the landowner was not a noble but rather a merchant and as such,

50 PSZ (II), vol. 11, n. 9203.

51 Khovanskii, “Pomeshchiki”, 60, 75, 118.
52 PSZ II, vol. 20, n. 19283.
53 Efremova, Sudoustroistvo.
54 Decisions of Senate in that sense in PSZ (II), vol. 3, n. 2378, vol. 8, n. 6129, vol. 9, 6941,
vol. 10, n. 8539, vol. 11, n. 9203, vol. 15, n. 13051, vol. 17, n. 15693, vol. 20, n. 19283, vol. 22,
n. 20825.
224 alessandro stanziani

prohibited from possessing serfs. In still other cases, the landowner was
a non-hereditary noble who transferred his estate and its population to
his children, etc.55 In short, in their legal disputes, the peasants took
advantage of every case involving the definition of the beneficiaries of
rights to own inhabited estates (false nobles, widows, merchants, indebted
nobles, etc.).56
Together with the owner’s title of property, the peasant’s obligations
could make the object of litigation. For example, in 1815, workers in a fac-
tory relied on these provisions to win their case, demonstrating that the
widow of their former boss was not registered in the local guild and hence
his apprenticeship contract was null and void and along with it, his rights
over the workers.57 Coercion in manufactures, shops and cottage industry
were quite often sanctioned.58
Other cases of sanctioning the title of the owner and the obligation
of the peasant-worker were linked to the civil status of the concerned
people. A married peasant, who had fled with his parents before the age
of fourteen and therefore could not be considered a peasant of a private
estate; she was thus redeemed of any obligation to the estate owner.
Similar issues included widows or wives of private peasants who married
someone belonging to another legal category; a given peasant who was
taken while still a minor in the provinces of Voronezh, Kharkov, Stav-
ropol, Don Cossack or the Caucasus; a non-Christian peasant owned by a
non-Christian master who converted to the Orthodox religion; a peasant
who was made a prisoner of war; a peasant that had denounced his mas-
ter for treason. The use of classifications was essential in these proceed-
ings. According to the law, servants and peasants of a bankrupt landowner
who owned only domestics and peasants without land were also trans-
ferred to the category of state peasant. The peasants of a landowner who
mortgaged his estate separately from his peasants, to whom he left less
than 4.5 desiatine (1 desiatina= 1.10 ha) per family, became state ­peasants.

55 TsGIAM, fond 54, 1783–1917, opis’ 1, for example: dela 56, 284, 966, 1509.
56 On transfer to married women and involved questions (legal statute): TsGIAM, fond
54, opis 1 et 2; for ex. Opis 1, delo 284; RGIA, fond 1149, opis 2 delo 20 and delo 44. The
issue was not always favourable to the peasants and workers. Example in GARF, fond 109
opis’ 3, delo 1885.
57 Quoted in Wirtschafter, Structures of society, 81. Peasant petitions in Materialy dlia
istorii.
58 TsGIAM, fond 32 (Moskovskoi gorodskoi Magistrat), op. 1 (1842–1866), dela 5, 13, 46,
118, 153, 284; opis’ 2 (1823–1844) dela 31, 40, 82, 124, 146.
 rights and bondage in russian serfdom 225

Finally, the peasants of a landowner who had illegally gone abroad were
also redeemed.59
The proceedings brought by peasants became so numerous that
between 1837 and 1840 the Senate even decided to put a stop to cases
involving serfs still living with their masters.60 Overall, between 1833 and
1858, the Senate recorded 15,153 cases of illegal ownership and “illegal
bondage” (as listed and qualified above), whereas the provincial courts
dealt with 22,000 cases of this type.61
Notaries’ archives and estate archives provide supplementary informa-
tion. Looking at some of these archives, the impression is that changes
in the legal status of private peasants made at the initiative of the estate
owner accelerated during the nineteenth century in response to both
fiscal support and legal simplification as provided by the tsarist rules.62
According to estimates at the time, the ninth census in 1851 in the twelve
provinces (guberniia) resulted in transferring eleven thousand peasants
into the category of meshchane.63 The validity of these figures is hard to
determine, for, in any case, they do not take into account all the emanci-
pated people that did not register on the urban lists of meshchane. Though
not cumulative, relatively detailed data on these individual acts was col-
lected by the Tsarist authorities between 1860 and 1868 in 25 volumes on
the state of the population in the Russian provinces and the Empire.64

Naturally, these positive outcomes should not obscure the difficulties

confronting peasants who tried to bring proceedings against a noble or
someone claiming to be noble. The rulings of local courts often differed
considerably and several judges considered peasant petitions inadmissible
and even refused to grant them the right to appeal.65 A number of cases
of corruption of judges by noble landowners were also recorded. Finally,

59 Svod zakonov, vol. 9, 957 (1842), arts. 544, 1040, 1042, 1078, 1107, 1185–97, 1199, 1201, 1205 
(1857); PSZ, I, vol. 30, n. 23806 (1809); PSZ, II, vol. 2, n. 906 and 1144 (1827), vol. 3, n. 1696,
(1828), n. 2052 (1828), vol. 7, n. 5425 (1832), vol. 8, n. 6163 (1833), vol. 10, n. 7982 (1835), vol. 11,
n. 9203 (1836), vol. 16, n. 14152 (1841), vol. 23, n. 21929 (1848), vol. 24, n. 23508 (1849), vol. 29,
n. 28846 (1854).
RGIA, fond 1149, opis’ 2; TSGIAM fonds 16, 32, 54, 342.
60 RGIA, fonds 1149, opis’ 2, delo 90.
61 PSZ (II), vol. 20, n. 19283, vol. 22, n. 20825. RGIA, fond 1149, opis’ 3, delo 125.
62 Notarial archives in RGADA, fonds 615, opis’ 1; fonds 1253 and 1274.
63 Keppen Deviataia reviziia: 6, 7, 21, 88, 95–100, 127, 142–144, 152, 159.
64 Glavnyi General’nyi Shtab, Materialy.
65 RGIA, fond 1149, opis’ 2 delo 20.
226 alessandro stanziani

the investigations of proceedings went slowly and often took ten years to
reach a conclusion. Measures aimed at changing this state of affairs were
not adopted until the end of the 1840s, when a new law facilitated the
legal proceedings of all those who considered their obligations towards
estate owners as illegal.66
However limited they were, judicial procedures ending with a change
in the legal status of concerned peasants entered a broader process in
which administrative law (or changes from above) strongly supported
these shifts.

New General Rules for Local Emancipation

The general issue of procedures concerning the legal status of peasants
consisted in transferring private estates’ peasants into the category of
state peasants, and, sometimes to other categories (mostly raznochintsy,
people of various ranks). Since the seventeenth century, and more clearly
in the 1720s, peasants of nobles and private estates had been distinguished
from state peasants. The latter, which most historians, especially in the
West, immediately included in the serf category, only had certain obliga-
tions to the State, e.g. to pay an annuity, usually in money, sometimes in
product. They also had an obligation to perform work of common interest
(which is usually mentioned to justify the existence of forced labour and
serfdom in this category). At the same time, state peasants could engage
in activities in the cities in trade and in industry, provided they had the
necessary documents.67 Catherine II even included representatives of the
state peasants in her legislative commission assigned to define the rights
and obligations of the main social groups.68 In 1842 state peasants were
freed from any obligation (except tax) towards the State. With this dis-
tinction in mind, we may now examine the ways a private peasant can be
“switched” to another category.
Voluntary redemption of peasants’ obligations and thus the change of
her/his legal status by the estate owner was already provided for in the
seventeenth and eighteenth centuries; a noble landowner could free a
peasant from his obligations with or without land. In the latter case—the
most widespread—the act could be freely carried out; the civil authori-
ties gave their consent and the will (in the case of emancipation after

66 Law of 1847, in PSZ II, vol. 22, n. 20825.

67 Druzhinin Gosudarstvennye.
68 SIRIO, X, 285–8.
 rights and bondage in russian serfdom 227

the landowner’s death) was recognised as valid.69 The noble landowner

was also responsible for paying the capitation or head tax on his peasants
until the next revision, unless a different agreement in the emancipation
contract stipulated that the involved peasant would pay his own tax. The
latter possibility had been adopted at the time of Catherine, precisely to
facilitate changes in the legal status of private peasants70 and it was sub-
sequently encouraged by the laws of 1844 which exonerated freed servants
from paying any tax until the next census (a time period reduced to six
years for those who registered on urban tax lists as meshchane).71 Indeed,
many laws were passed during the first half of the nineteenth century
to facilitate administrative procedures for changes of the legal status of
private peasants. While responding to contingent situations we are going
to detail hereafter, these rules as a whole expressed the particular stakes
and worries of a part of the Russian elites. Political stability, economic
efficiency, paternalistic criticism to serfdom, concrete economic interests
of the State and a part of the nobility, all these variables took part, to a
different degree, to these reforms.72 A first reform was adopted by Alex-
ander I who ascended the throne in 1801. This same year he allowed mer-
chants, townspeople and state peasant to own unpopulated estates. This
was intended as a partial breaking of the nobles’ monopoly on landowner-
ship. In 1803, a decree created the svobodnye khlebopashtsy (free farmers),
i.e. those who, by will of a landowner, were exempt from any obligation
to him, aside from those resulting from the attribution to peasants of plots
of land belonging to the owner. The agreements were legally binding for
both peasants and landowners.73 The influence of German and Russian
academicians, as well as the pressure of some of the noble elites encour-
aged the tsar’s reforming attitudes. In particular, Alexander was pushed
to adopt this provision by count Rumiantsev who sought permission to
free some of his peasants of any obligation and sell them land. The posi-
tive answer to this request became the ground for a broader reform. This
provision was supposed to lead to an improvement of agriculture while
being advantageous to estate owners. According to the estimates of Hoch
and Augustine, between 1833 and 1858, 58,225 private peasants changed

69 Svod zakonov, vol. 9, art. 674–680, 1833, art. 1148–1184, 1857.

70 PSZ, I, vol. 20, n. 14294 (de 1775).
71 PSZ, II, vol. 19, n. 17977 (12 juin 1844) and n. 17985, 12 juin 1844.
72 Becker, Nobility; Dukes, Catherine; Jones, The emancipation; Lieven, The Aristocracy;
Lincoln, The great reforms; Pintner, Russian.
73 PSZ, I, 27, n. 20620 (20 février 1803).
228 alessandro stanziani

their legal status on the basis of this law,74 a figure that rose to 114,000
male peasants between 1803 and 1855.
However, despite several attempts to go further in this direction, Alex-
ander did not adopt other reforms. The Napoleonic wars and the opposi-
tion of some groups of nobles and high officials seem to have played a role
in his attitude. The greatest worry was not so much the change of the legal
status of peasants than their possibility to sell the acquired lands to other
“non-noble” owners.75
The next tsar, Nicholas I, also spent some time in projecting reforms
before adopting concrete rules. These occurred in the broader realm of
Kiselev’s reforms that between 1838 and 1842 radically modified the condi-
tions of state peasants. Peasants’ general and professional education was
taken into account, while the fiscal burden and the obligations toward
the State were no longer assessed per head, but in conformance with the
value of peasant’s land or their incomes from non-agricultural activities.76
Kiselev ordered a general enquiry on agriculture and the peasantry; the
final document reached the conclusion that the labor of peasants with
no statutory obligations toward the landowners was more productive
than existing forms of labor. On the ground of this report, Kiselev and
Nicholas I maintained that the present difficult situation of agriculture
and the existence of coerced labor was the result of the fact that some
landowners had abused their power and extended their seigniorial power
over the peasantry. Limitation of the abuses thus became a precondi-
tion for deeper reforms. At the same time, peasants had to be given the
opportunity to own their own land. In fact, as the experienced reforms
in the Western borderland had proved, the end of statutory obligations
without land would lead to a worsening of the peasants’ condition and to
social and political instability. According to this, in 1842 a new decree was
adopted; it was conceived as direct filiations of the 1803 decree, but with
the major difference that the contractual agreement between the estate
owner and peasants would not be left to the free will of the former but
would be regulated by the law. Until when the involved peasants had not
redeemed the price of land, estate owners would have retained the right
of full patrimonial ownership of the land and the peasant would receive
plots of land for use. A particular category of peasants was created; they

74 Hoch, Augustine, “The tax censuses”, 208–217.

75 Moon, The abolition, 42; Mironenko, Samoderzhavie.
76 Druzhinin, Gosudarstvennye.
 rights and bondage in russian serfdom 229

were the peasants with a temporary and specified obligation (obiazan-

nye krest’yanie) towards noble landowners, who fit neither the category
of private estate peasants in the strict sense nor that of servants (dvornye
liudi). As previous rules, this decree too accurately spoke of peasants, pri-
vate peasants and rural inhabitants, not of “serfs”. These peasants were
subject to other legal rules; they had no more general obligation towards
landowners, except those concerning the plot received at the signing
of the contract with the estate owner.77 From the adoption of this law
until 1858, 27,173 male peasants (and thus their families) were affected by
the decree.78
In 1841, peasants that private nobles had freely transferred to monas-
teries or charity institutions were also re-classified due to the prohibition
against ownership and transfer of peasants by these institutions.79 About
8,900 people were thus transferred to the category of state peasants and
the institutions to which they belonged were compensated by the State.80
In 1844, two new laws facilitated the re-classification of dvorovnye liudi
(servants); the first one freed the master from any responsibility regard-
ing the payment of the taxes of these people, whereas the freed servants
were not subject to military service or taxation until the next revision. The
second law facilitated the change in legal statute in cases in which the
landowner mortgaged his estate to certain credit institutions. According
to a study at the time, in 1851 alone, 11,000 meshchane (merchants) from
eleven provinces are said to have benefited from these rules (they were
previous peasants of private owners).81
A great many worker-peasants attached to private factories were
exempted from any obligation toward the factories owners between 1840
and 1851; about 19,000 people were concerned. During the same period,
many mines had recourse to “temporary workers” (nepremennye ou
urochnye rabotniki) registered as private peasants at the time of the eight
revision. The facts were denounced years later and a court decision was
made only for the ninth revision in 1851. At that point, 53,900 men work-
ing in the mines were transferred to state estates as peasants.82 For both
factories and mines the argument behind these provisions was that the

77 PSZ, II, vol. 17, n. 15462.

78 Hoch, Augustine, “The tax”, 410.
79 PSZ,II, vol. 16, n. 14669, 19 June 1841.
80 Mel’nitskii, Sbornik, vol. 2, vyp. 3, 187, 217; vyp. 4, 52, 119.
81 Keppen, Deviataia reviziia, 6–7, 21, 62, 88, 95, 100, 127, 142–144, 152–159.
82 Ryndzinskii, “Vymiralo”.
230 alessandro stanziani

owners were not entitled to get statutory (that is non contractual and
timeless) obligations and services from the workers.
Administrative re-classifications sometimes occurred for other reasons.
State and noble elites agreed on this, indebted landowners had to be
helped and avoiding their lands would fall into the hands of “speculators”
and “bourgeois”. To this aim, in 1841, a new law prohibited the succession
of landless nobles or landless owners of peasants. Consequently, between
1841 and 1858, the number of landless nobles dropped from 17,700 to 3,633
and their peasants from 62,000 to 12,045 (even taking into account the
possibility of some of the serfs having been sold to other nobles).83
To this same aim, mortgaged estates put up for auction were especially
targeted. Local peasants were granted the priority (with the State itself) to
acquire lands. Otherwise, the “public treasury” (udel) and, then, the Min-
istry of State dominions, intervened and bought estates. In all, between
1833 and 1858, the legal status of 34,3575 male peasants changed because
of these operations.
Ordinary and administrative law regarding emancipation had a cumu-
lative effect not only on mortgaged estates, but also on the terms of suc-
cession. Two new laws adopted in 1836 and 1839 stipulated that when a
landowner did not leave any heirs (or heirs that could legitimately have
serfs), the peasants would become state peasants whereas the servants
could register in cities in the category of meshchane.84
Other provisions admitted changes in the legal statute of peasants as
a consequence of political crimes made by landowners. Noble estates of
Western Borderlands, Poland, West bank of Ukraina who took part to
uprising were confiscated and their peasants (estimated 264,000) entered
the category of state peasants.85
Ultimately, beyond peasants voluntary redeemed by their masters and
private peasants turned into state peasants following a judicial litigation
or an administrative act, massive changes in the legal status of peasants
was linked to conscription and military service.

83 Hoch, Augustine, “The tax”, 420.

84 Svod zakonov, vol. 9, art. 462–3 (1857); PSZ, II, vol. 11, n. 9203 (1836), vol. 14, n. 13012
(1839).
85 PSZ, II; vol. 11, n. 9053; See also Ministerstvo gosudarstvennykh imushchestv,
Istoricheskoe, part. 2, 80.
 rights and bondage in russian serfdom 231

The Army As a Factor of Emancipation?

Conscription was the most important source of administrative change

of the legal status. Once a conscript had completed his military service,
he was classified as raznochintsy and could therefore move about freely,
settle in the city, etc. Hoch and Augustine estimate that about 433,750
peasants changed their legal status (from private to state peasants or even
urban residents) in this way between 1833 and 1858 alone.86 This outcome
was hardly considered favourable by the peasants, however. Indeed, the
length of conscription (20–25 years according to the period) implied first
and foremost a decrease in workforce for the family, which also had to
continue paying a tax for the conscript until the following revision (which
might take ten to twenty-five years). Moreover, upon their return, con-
scripts were relatively old and hence less productive and advantageous
for both their families and landowners. Re-classified ex-conscripts (from
peasants to raznochintsy or others) were not entitled to land, although
most soldiers requested it instead of freedom of movement, which they
did not know how to manage very well. Numerous requests and petitions
were then sent to village elders, noble lords and local authorities to be fully
reintegrated into the rural community, which were nevertheless refused.87
Of course, the perspective of changing the legal statute at the end of the
service did not imply that soldiers were allowed to escape before the term.
In this case, two crimes were simultaneously acted: desertion and run-
away (as a peasant belonging to a given rural community and a given
landlord).88
The new social definitions and redefinitions of the legal status of con-
scripts also affected their families. For example, children born after the
soldier entered service were placed under the guardianship of the Ministry
of War and destined for military service. This provision and its application
gave rise to a number of disputes, petitions, etc. brought by both peas-
ants and nobles who saw themselves deprived of considerable manpower
for the benefit of the army.89 Most of these disputes were based on the
definition of “entry into service” and “birth”; military leaves were indeed

86 Hoch, Augustine, “The tax”.

87 Some examples in TsGIAM, fond 16, opis’ 1, dela 1293, 1328, 1352, 1801.
88 TsGIAM, fond 16, opis 1, 2, 225, 227 (several files in each). On the importance of deser-
tion at the origin of “runaways”, Kahan, The plow, 76; Stevens, Soldiers.
89 TsGIAM, fond 16, opis’ 2; delo 192; fond 54, opis’ 1, delo 1618; fond 16, opis’ 2, delo 192;
fond 54, opis’ 1, delo 1618. Also: TsGIAM, fond 16, opis’ 227, dela 227, 938, 948, 976 980, 986,
1573, 1571, 1588, 1589.
232 alessandro stanziani

evoked, as well as the interpretation of the laws, some referring to birth,

others to conception. False birth dates, encouraged by both peasants and
noble landowners, were commonplace.90
The other subject of dispute concerned the peasants’ illegitimate chil-
dren; their families and the nobles considered that these children did not
fall into the category of those destined for the army and therefore recom-
mended a restrictive interpretation of the rules in this area; the Ministry
of War, on the contrary, considered that these children belonged to it.
Thus, in 1808, the Senate made an example of Princess Golitsyna due to
her rank, finding her guilty of illegally enserfing two illegitimate children
of the wife of a soldier from her estate.91 These conflicts intensified with
the Napoleonic wars, so that a 1816 decree declared that all children of
soldiers (whether legitimate or illegitimate) who had been mistakenly
enrolled in the estates of the State or of a noble in the previous revisions
would retain their status.92
On the whole, the number of children who changed their legal status
by this method has been estimated at about 12,000 between 1833 and 1858.
Added to the re-classification of soldiers, the number of peasants who
changed their legal status in connection with the status of soldier or child
of soldier rose to about 450,000.93
The army thus acted as a powerful source of changing the legal statute
of peasantry, mainly private owners’ peasants, who became either state
peasants or enter the urban population. As such, this provision joined sev-
eral other rules the tsarist regime adopted in order to proceed in admin-
istrative changes in the legal statute of the peasantry.

In all, between 1833 and 1858, i.e. between the eight and the tenth revi-
sion, about a million men were released from the status of peasants of
private owners due to administrative acts (the army plus the other cases
mentioned), out of a total of about ten million serfs.94 More than a quarter
of all male peasants was switched to other categories and became either
state peasants or urban dwellers.

90 RGADA, fond 294, opis’ 1, delo 567.

91 Wirtschafter, Social, 578.
92 PSZ, I, vol. 33, n. 26376, PSZ, II, vol. 12, n. 10727. Also TsGIAM, fond 16.
93 Hoch, Augustine, “The tax census”.
94 Hoch, Augustine, “The tax”.
 rights and bondage in russian serfdom 233

Conclusion: Labour and Economic Dynamism in Imperial Russia

Two main elements are usually called upon to explain “serfdom”: the State
and noble landowners. Historians talk about the State in two different
ways, which are totally incompatible. For some, serfdom is held to be
due to the strength of the autocracy, whereas others emphasise, on the
contrary, the weakness of the Tsarist State, which is said to have left the
nobles free to enserf the peasants.
These two interpretations are partial, if not erroneous, in that they do
not take into account the close interaction between the State (or rather,
certain administrations and certain leaders) and Russian society. The
State was more than an administration and above all, the latter was not
limited solely to taxation and the police; the State was also the law, and
from that standpoint, the dominant interpretations miss the essential
point when they assert that the autocracy imposed a law that was not
really law or, on the contrary, that it was incapable of controlling rela-
tionships in the countryside. In reality, Tsarist law recognised and relied
on other legal sources, including seigniorial law and its courts, and these
rules mutually legitimated each other. The disputes between nobles and
merchants or the considerable emancipations of peasants through legal
and/or administrative channels testify to the importance of this element.
However, the law was hierarchical; the various social groups had varying
access to the courts and to jurisdictions. In other words, while it is not
true that Russian society had no law, it would be a mistake to equate
it with the rule of law as defined, let us say, in France or in the United
States at the end of the twentieth century. The way law was conceived
and implemented shaped social and economic hierarchies. These last
cannot be seen as a simple coerced exploitation of “serfs” made by noble
landowners.

The peasants on private estates were subject to considerable restrictions

on their mobility. However, these restrictions were never definitive and
they changed according to the estate and the economic situation. As a
whole, these provisions corresponded to a reorganisation of the adminis-
tration and the introduction of a land registry in order to settle conflicts
between nobles and merchants, or even within noble families, over the
transfer of estates and the attribution of “noble” status.
Our approach leads us to raise doubts on the two main interpretations the
historiography provided of 1861 reforms, one stressing their revolutionary
234 alessandro stanziani

role and the other undermining their partiality and ultimate failure.95
The first interpretation underestimates previous reforms, exaggerates the
importance of serfdom and the development of Russian rural economy
after 1861. The second interpretation shares the same perception of a
tough and stable system called “serfdom” and puts the accent on peasants’
indebtedness and miserable standards of life after 1861. The first interpre-
tation has difficulties to explain the huge mass of “emancipation” acts we
have dealt for, the economic growth before 1861 as well as peasants’ hos-
tile reactions to the emancipation act. The second approach seems unable
to take into account the labor institutions after 1861, the economic growth
of pre-revolutionary Russia and nobles attitudes towards reforms as well.
Indeed, the reforms and judicial acts we have detailed testify to the fact
that the reforms of 1861 were part of a long reform process dating back at
least to the last quarter of the eighteenth century, which gained strength
during the first half of the following century and finally accelerated signifi-
cantly starting in the 1830s. This dynamic also expresses the complexity of
a process that it would be an exaggeration to describe as purely admin-
istrative, i.e. as a “reform from the top down”. First of all, because, as we
have seen, behind the many laws and decrees modifying the conditions of
emancipation stand social groups, both noble and bourgeois.
The second reason that leads us to question the argument of “top-down”
reforms is that the aforementioned laws and decrees were situated at the
crossroads of administrative law and civil law procedure. The administra-
tive authorities adopted measures based on common law rules pertaining
to the transfer and mortgaging of estates, and above all, defining those
entitled to full ownership of estates and their inhabitants.
The reforms of 1861 thus do not mark a break insofar “serfdom” was
much more flexible than usually asserted and legal constraints on peas-
ant mobility and peasant labour did not disappear after 1861. In fact, the
reform program as it was laid down in 1858 was largely inspired, as regards
peasants’ obligations and legal status by the decrees of 1803, 1842 and 1847.
As we have seen, these decrees fixed the rules for peasants’ legal eman-
cipation and land acquisitions; they had supported the emancipation of
around 240,000 peasant families. The “emancipation act” of 1861 repro-
duced the main features of these previous rules and only added a broader
financial state support to both peasants and landlords in order to facilitate
peasants’ access to land.

95 Gershenkron, Economic backwardness; Blum, Lords and peasants; Field, The end.
 rights and bondage in russian serfdom 235

All the same, between 1842 and 1856, the ministry of state domains
assembled impressive quantities of data about life in state peasant vil-
lages and the distribution of land between peasants and landlords and
among peasants. These surveys covered almost 11 millions peasants and
more than 65 millions acres of land in 25 Great-Russia provinces. Least
but not the last, the Russian government imposed inventories defining
schedules of peasants’ duties to landlords and landlords’ obligations to
their bondsmen.96
Our approach equally provides a solid justification to recent estima-
tions of economic growth and demographic trends between 1750 and 1861,
as well as from 1861 up through 1914, both having revised upwards.97 These
dynamics are much less explained by the possibility of further squeezing
peasants under serfdom, than by the coexistence of coercion and flex-
ibility. This was all the more true that before 1861 a large amount of the
peasantry had already been transferred into the category of state peasants
or urban inhabitants. This widely explains the good performance of Rus-
sian agriculture now widely accepted in historiography.
Our perspective provides a ground for justifying peasants and nobles
reactions to the “emancipation act”. Most of the peasants negatively
reacted to the 1861 act in which they saw nothing but a smooth shift from
previous conditions. We must to be clear: this does not mean, in our view,
that the reforms of 1861 were a “failure” and that they gave landowners a
new way for exploiting peasants. Quite the contrary, the reforms went on
along the previous path and aimed to strengthen the mixture of coercion
and flexibility—mostly to the social appropriation of legal rules—already
experienced since the early nineteenth century. Peasants’ had much less
duties, a better legal protection and rights and a scheduled access to full
property. Still, this was often perceived by the peasantry not as a “revo-
lutionary break” but rather as a further step along the previous reform
lines.

What about Landlords?

Several scholars have been surprised by the fact that, beyond clearly
documented opposition, several nobles largely expressed their assent to
reforms.98 This was so not only because of their political weakness and

96 Lincoln, The great reforms, 66–69; Field, The end, 77–83.

97 Gregory, Before command; Gatrell, The Russian economy; Stanziani, L’économie.
98 Field, The end; Moon, The end.
236 alessandro stanziani

the social differences among them, but also because nobles perceived that
the ongoing reform was in a strong continuity with the previous path.
Landlords kept part of peasants’ duties while receiving considerable finan-
cial resources from the State. Peasants’ obligations were mostly kept in
life while the state financed the transfer of less valuable lands to the peas-
antry at an extremely interesting price for noble landowners.
This for nobles’ support or at least for their lack of reaction to the
reforms. At the same time, the 1861 act introduced a new dimension
(beyond the financial support the state gave to both peasants and the
state) into the reformatory process. This new dimension was at the origin
of nobles’ hostility towards the reformatory process. In fact, labor rela-
tionships in agriculture were transferred from landlords and estate courts
to district courts. This was seen by several nobles as an intrusion of the
state into peasants-landowners relationships and marked a considerable
change as regards all the previous three and a half centuries. The end of
seigniorial justice was not only in Russia, but everywhere in eighteenth
and nineteenth century Europe one of the major breaks with the old
regime.
A second major break and a reason for nobles’ dissatisfaction was
that the reform erased any legal relationship between land ownership
and social (and political) status of “noble”. By now, no-one was entitled
to transmit inhabited estates and, conversely, merchants, kuptsy and
urban elites had access to land. Nobles clearly understood this shift and
appealed to counter-reforms. Tsarist elites attempted to take into account
this last request by strongly limiting the market for land (peasants could
not sell their lands to non-peasants or to peasants from other villages),
while maintaining an important status and political difference between
nobles and merchants (and, of course, peasants and workers). However,
these last defences to nobles privileges were progressively abandoned, the
latter with the institution of Duma and the former with the introduction
of a free market of land at the eve of WWI.
 Acting as Master and Bondservant:
Considerations on Status, Identities and the Nature of
“Bond-servitude” in Late Ming China

Claude Chevaleyre

Introduction

In early 1910, following the recommendations of the Commission on Consti-

tutional Government, an edict abolishing slavery and trafficking in human
beings in the Chinese empire was officially promulgated in the name of
the four-year-old Xuantong emperor (r. 1908–1912).1 A few months later,
the publication of the new Penal Law of the Great Qing Currently in Use,
compiled by the Commission for the Revision of the Code, confirmed the
emancipation of all slaves within the empire. A new sub-statute appended
to the law entitled “Households should be established according to the
registers” (renhu yi ji weiding 人戶以籍為定), stipulated that “house-
hold slaves” ( jianu 家奴) were granted permission to redeem themselves,
whatever their origins and the number of generations their families had
been in bondage. In order to enforce this decision, local authorities were
ordered to register redeemed slaves as commoners and to let their sons
and grandsons take the official examinations. For those who could not
redeem themselves, the law stated that they should be considered as hired
laborers. This raised their legal status from that of unfree laborer to that
of free laborer, without entirely erasing the personal tie that linked them
to their former masters.2

1 For more details on the process that led to the proclamation of the ten regulations for
the prohibition of the purchase and sale of human beings, see Edward T. Williams, “The
Abolition of Slavery in the Chinese Empire;” see also his translation of the documents
related to this emancipation, “Report to the Throne of the Imperial Chinese Commission
on Constitutional Government Recommending the Abolition of Slavery.” A former Prot-
estant missionary and official translator in China, Edward Thomas Williams (1854–1944)
was at that time Assistant Chief of Division of Far Eastern Affairs in the State Department.
A short mention in the Political Records of the Xuantong Reign, dated January 1910, con-
firms that decisions were effectively taken in order to enforce “the prohibition on the old
custom of trade in human beings” (  jinge renkou maimai jiuxi 禁革人口買賣舊習); see Da
Qing Xuantong zhengji shilu, 28:14a–b.
2 Shen Jiaben, Da Qing xianxing xinglü, 5:4a.
238 claude chevaleyre

Among the variety of forms of personal subjugation (including traffick-

ing in children for immoral purposes) that the collapsing Chinese empire
attempted to eradicate, the most widespread was slavery. Yet, this form
of slavery was not the same as that abolished earlier by European coun-
tries. It was not a form of transnational trade in human beings based on
race. It was, rather, the remnants of a practice of domestic bondage that
was rooted in the ancient past and that had remained basically the same.
Since the Ming dynasty (1368–1644).
From its very beginnings, modern Chinese historiography, strongly
influenced by Marxist-Leninist—and then Maoist—theories, has shown a
great deal of interest in forms of personal bondage, and, more specifically
in the late Ming, in slaves, henceforth named “bondservants” in English-
language sinology. However, Ming bond-servitude has hardly been stud-
ied for itself, despite lengthy and impassioned debates led by Chinese and
Japanese scholars who rank among the most prominent in the field of
Chinese studies.3 And, since Marxist-influenced views strongly dominated
the debate, bond-servitude has remained subordinate to objectives that
transcended it. The subject has remained relatively unexplored, either
because it was deemed marginal, or because more thorough analyses were
likely to introduce undesirable elements into historiographic models.4
Bond-servitude was in fact less regarded as a topic worthy of study per
se, than as an instrumental—though critical and well-documented—
element in supporting ideologically and politically-oriented historical
schemes more concerned with the nature of late-imperial society or the
periodization of Chinese history.5
Recent trends in cross-cultural and interdisciplinary studies have now
created the conditions for a new dialogue among historians and sinologists

3 Such debates attracted the attention of historians mainly from the 1930s until the
1980s. Pioneering studies on late Ming bondservants and bondservant rebellions are: Xie
Guozhen, “Mingji nubian kao” (1925); Chen Shoushi, “Ming Qing zhiji shiliao (1)” (1927);
Chen Shoushi, “Ming Qing zhiji shiliao (2)” (1929); and Fu Yiling, “Mingji nubian shiliao
shebu,” (1949).
4 For example, despite the abundance of materials, narratives of “loyal servants” barely
drew the attention of historians. And as far as I know, while most studies focus on the
exploitative aspects of Chinese bond-servitude, no one has studied its place in corpora-
tions where technical knowledge was transmitted from patron to apprentice, the latter
sometimes being considered and treated as a bondservant. Finally, I would strongly argue
in favor of a deep re-examination of the so-called “bondservant rebellions,” especially of
their historical significance and even of their “bondservant” nature.
5 For an overview of twentieth century scholarship, see Joseph P. McDermott, “Bond-
servants in the T’ai-hu Basin,” 675–677.
 acting as master and bondservant 239

on universal problems such as personal bondage. However, drawing

superficial parallels between past and present or between China and
other areas can only be of limited interest.6 If there is any purpose or util-
ity in doing history—apart from pure erudition—it is not simply to pro-
vide lessons for the future. It is also to establish connections between past
and present, to identify dynamics and explain historical evolution. In the
case of China’s history, as in other cases, an essential purpose of histori-
cal research is to identify continuities and developments in order to ana-
lyze and understand long-term changes.7 In the same manner, comparing
different areas is not only about establishing broad differences or simi-
larities. Comparison is useless when it is only intended to demonstrate
the otherness of one of the entities compared. Historical comparison can
nonetheless prove fruitful when it aims at dissecting the complexity of the
various mechanisms at work in universally shared human experiences.
In the field of the history of bondage, China’s experience can thus be
compared with the experiences of other areas in two ways: It can either
be studied with the sole purpose to make it fit with Western categories
of “serfdom” or “slavery” considered as universal referents; or it can be
regarded as another, singular configuration, worthy of attention and likely
to enlighten our knowledge of the mechanisms of personal subjugation
and bondage. In this regard, the present book should be counted as a
forceful incentive to rethink our perception of these topics in the Chinese
context.
But before we can actually succeed in (re)integrating China’s overlooked
experience in a global history of bondage, there is an urgent need for
intensified interdisciplinary dialogue and exchange. If collecting sources
remains the priority of sinologists, there is an even more urgent need to
make Chinese materials accessible to a larger audience of scholars. Such
a goal can only be achieved through the translation and publication of
additional primary sources. As a preliminary step in this direction, I chose
to go back to basics and propose a rough introduction to the status and
identity of bondservants in late Ming China.8 I will begin with general

6 This is the case, even though it may be tempting to draw parallels between our global-
ized world and late Ming China, be it in terms of acceleration in commercial exchanges,
new questioning of social norms, increased competition over resources, financial crisis, etc.
7 For a more persuasive argument, see Pierre-Etienne Will, “L’histoire n’a pas de fin,”
7–8.
8 The useful but vague word “bondservant” is now conventionally used to translate
the Chinese terms nubi 奴婢, nupu 奴僕 and the like. However, as a result of the need
to think with Western categories, some scholars continue to call these persons “slaves” or
240 claude chevaleyre

considerations regarding their legal status, with late Ming social hierarchy
and the nature of social degradation. I will then turn to more specific con-
siderations regarding the core nature of Chinese bond-servitude through
an examination of the “bond” between master and bondservant as pre-
sented in late Ming judicial cases.
Bondservants were not the only category of unfree persons in China.
But they were the most numerous and, to a certain extent, everything
that was related to unfree labor in the late Ming was also, more or less,
directly or by analogy, related to bondservant status. As for the focus on
late Ming China, this period remains the best starting point for study-
ing bondservants. The seventeenth century was objectively the culmina-
tion of a protracted and multifaceted crisis. The most dramatic moment
occurred on April 23, 1644, when the rebel leader Li Zicheng (1605–1645)
took Beijing and established the short-lived Shun interregnum, which
was soon overthrown by the Manchu-Qing who were to rule China until
1911. During the decades before and after this moment, the provinces
were confronted with chronic violence and were caught in the midst of
political repression, widespread peasant revolts, urban movements, rent
resistance, craftsmen uprisings and, long after the conquest, bloody bat-
tles between Ming loyalists and troops of the Manchu banners.9 At the
same time, bondservants all over China also became a source of social
unrest.10 They took up arms (sometimes provided by the local gentry in
ill-considered attempts to organize local resistance) and rose up against
their masters to try and retrieve their bondage contracts, killing those who
resisted them and burning their houses.11 In the blink of an eye, this social

“serfs.” For a case study of bondservants in a late Ming locality, see Harriet T. Zurndorfer,
“Servage domestique, conflits sociaux et la transition Ming-Qing dans la préfecture de Hui-
Chou, 1644–1646.” In Alessandro Stanziani (ed), Le travail contraint en Asie et en Europe,
Paris, FMSH, 2010, pp. 241–264. For other general articles on bondservants and bondser-
vant rebellions in late Ming China see: McDermott, “Bondservants in the T’ai-hu Basin;”
Mi Chu Wiens, “Masters and Bondservants;” Masatoshi Tanaka, “Popular Uprisings;” and
William T. Rowe, Crimson Rain, 120–128, passim.
9 On the Manchu conquest, loyalism and peasant rebellions see Frederic Wakeman,
Jr., The Great Enterprise; Lynn A. Struve, Voices from the Ming-Qing Cataclysm; Jonathan D.
Spence and John E. Wills, From Ming to Ch’ing; and James B. Parsons, The Peasant Rebel-
lions of the Late Ming Dynasty.
10 Bondservant insurrections broke out in almost every eastern and southern province
between the 1630s and the 1650s, but the great majority occurred in the prefectures of the
lower Yangzi delta between 1644 and 1646.
11 The “search for contracts” (suoqi 索契), the destruction of masters’ houses and the
murder of masters’ families are recurring topoi in bondservant-rebellion narratives. Yet, if
every act of violence committed by bondservants is usually and indiscriminately regarded
 acting as master and bondservant 241

group usually not deemed worth a note in conventional sources jumped

from the shadows into the spotlight and became the subject of numerous
records and lengthy narratives, which now constitute a crucial corpus of
sources on bond-servitude.

Multiple Identities, Social Hierarchy and the Law

Twentieth-century Chinese, Japanese and Western scholarship has amply

demonstrated how complex and multifaceted the social identities and
conditions of Ming bondservants could be. From the most powerful
estate managers, who terrorized the country, to the sexually abused maids
described by the prefectural judge Zhang Mingbi (1584–1652);12 from
those who submitted their masters to their own will13 to the thoroughly
exploited ones depicted by Zhang Lüxiang (1611–1674);14 from the richest

by a majority of Marxist-oriented scholars as the expression of a new “class conscious-

ness” and as part of an intensifying “class struggle,” there is still much to do to distinguish
precisely those who fought for “emancipation” from more opportunist bandits. Of course,
narratives were written by educated elites rarely inclined to commiserate with bondser-
vants and depict them in a favorable light. But such sources bring to light ambiguities and
should warn us against misinterpretations. The uprising that shook Liyang County in 1645,
for example, is generally classified as a bondservant rebellion. The leaders, the Pan broth-
ers, were indeed bondservants (pu 僕, or cangtou 蒼頭) in the Peng family. Yet, when
reading the long week-by-week narrative depicting them as blind rebels plundering and
killing rich and poor indistinctly (even allying with the Manchus in a totally opportunist
move), one can hardly regard them as idealists fighting for progress and emancipation; see
Ming chushi yiming daoren, Laijiang jishi benmo.
12 In a narrative of Jintan County bondservant rebellions, Zhang Mingbi has the follow-
ing horrific description: “I heard about fierce and cruel masters in Jiangnan, who severely
and blindly manage [their bondservants] in accordance with the bondage agreements.
Many bondservants have their stomachs empty and nothing to eat, they have no clothes
to protect their ankles and their knees, and the skin on their buttocks has disappeared.
Before their daughter is married, [the master] takes her virginity. Before their wives have
borne children, he ruins their freshness. And when the [master’s] wife is jealous, she even
mutilates their vaginas, shaves them, or sews their skin; their cries of pain thus can be
heard far away.” See Zhang Mingbi, “Xiaobi ban ji.” Another classic narrative of masters’
cruelty can be found in Jin Zhen’s (1622–1685) 1658 memorial on Guangshan county bond-
servant rebellions; see Jin Zhen, “Tiaochen Guangshan panpu xiangyi,” 19:16a–18b.
13 Like Chen Peng and Guo Kun, the skillful bondservants of Minister Zhang Fuzhi
(1547–1629). According to the Brief Account of the Restoration Society, Zhang totally relied
on Chen for writing his official documents and on Guo for the management of his prop-
erty. They had the reputation of controlling every matter in Zhang’s household and having
so much power over him that they did not hesitate to reprimand his elder brother Zhang
Yizhi; see Lu Shiyi, Fushe jilüe, juan 13, 516.
14 In a poignant text, Zhang Lüxiang heaps opprobrium on masters who “don’t treat
bondservants as humans,” who let them starve and suffer from cold without an ounce of
242 claude chevaleyre

to the poorest; from the most educated to the totally illiterate; bondser-
vants assumed nearly every possible position in Ming society. They were
maidservants, peasants, shopkeepers, gatekeepers, bookkeepers, appren-
tices, nurses, estate managers, militiamen, yamen servants, tax collectors,
wine shop owners and, in the rarest cases, even minor officials.15 Some
lived with their masters and some had their own houses. They could be
born bondservants, sold by their parents or kidnapped as children, while
others voluntarily offered or sold themselves to a master. Bondservant
status could be life-long and hereditary, or only pro tempore. One entered
in servitude because of debt or poverty,16 to escape taxation, to avoid
punishment, to take advantage of a new masters’ influence, in search for
protection, or just because one was not rich enough to pay for a bride17
or a proper interment for one’s parents. And, depending on age, gender,
function and even local custom, bondservants were called by a wide vari-
ety of names, among which nubi and nupu were both commonly used as
well as the legal terms. As a result, and because modern scholars mostly
choose to regard concrete social conditions as a key factor in determin-
ing status, this very disparity has been the cause of radically divergent,
not to say opposing, interpretations regarding bondservant status. While
some have focused on the ignominious treatment suffered by a majority
of exploited bondservants, others have emphasized the fate of the most
powerful, influential and uncontrolled ones. Similarly, when attempting
to evaluate to what degree they were integrated or excluded from society,
some have come to the conclusion that bondservants—who were also
named jiaren 家人 (lit., “family members”)—were deeply integrated into

compassion, who abuse their wives and daughters, who appropriate their goods and who
even kill and burn them; see Zhang Lüxiang, Yangyuan xiansheng quanji, juan 19, 575.
15 The Essentials on the Institutions of the Ming mentions 16 officials who were actually
of bondservant (cangtou 蒼頭) origin in the 1450s; see Long Wenbin, Ming huiyao, juan
52, 969.
16 Poverty was a frequently invoked reason in bondage contracts. When in 1609 Hong
Sanyuan sold himself, his wife and their son to a master, the contract on which he affixed
his hand prints clearly stated that he did so because he had not enough to feed his fam-
ily. In return, Hong received 15 ounces of silver; see Zhang Chuanxi, Zhongguo lidai qiyue
huibian kaoshi, 929.
17 Self-sale was often motivated by the lack of resources to pay for a bride. Such con-
tracts were therefore presented as “self-sale and marriage contracts” (maishen hunshu
賣身婚書). Zhang Shenbao’s bondage contract (dated 1574) stated that he voluntarily
wanted to become the perpetual bondservant of the Wangs in order to marry Ai Gui, a
maidservant of theirs. See the contract held at the Dingling Museum and reprinted in Han
Dacheng, Mingdai shehui jingji chutan, 83.
 acting as master and bondservant 243

a family continuum, while others have strongly asserted that they were
socially excluded and marginalized.
Every pattern thus proposed is undeniably supported by concrete
examples and evidence. Thus, rather than focusing on an impossibly broad
range of observable configurations, it seems more fruitful to reconsider
the question from a new angle and ask what bondservants had in com-
mon, what were the characteristics they all shared. The first characteris-
tic shared by all bondservants was legal status. Bondservant was clearly
an acknowledged and legally enforced status. However, when attempt-
ing to trace the contours of this legal category, one is soon confronted
with the fact that neither the Great Ming Code, nor any other legal or
normative source ever provided positive or descriptive definition of any
social groups.

Bondservants, Hierarchy and the Great Ming Code

Promulgated in 1397 by the Ming founder Zhu Yuanzhang, the Great Ming
Code was mainly a body of penal legislation.18 It operated like a complex
legal chart or schedule: It characterized crimes and offences, prescribed
the adequate sentences and specified sentence modulations according to
the relative status of the offender and the victim. Status was therefore of
pivotal importance in the internal rationale of a code of laws that was
chiefly intended as a tool to preserve social order and ensure the durabil-
ity of the dynasty—rather than to ensure individual rights.19 But identities
were neither fixed nor absolute since they were deeply rooted in con-
ceptions of relative personal hierarchical relationships defined by Confu-
cian orthodoxy.20 One’s hierarchical position was thus eminently variable

18 It also included statutes that could be regarded as being of administrative or civil
nature. As for debates regarding the nature of late imperial Chinese law, see in particular
Jérôme Bourgon, “Uncivil Dialogue.”
19 For a good translation and introduction to Ming law, see Jiang Yonglin, The Great
Ming Code. See also Edward L. Farmer, Zhu Yuanzhang and Early Ming Legislation, 107; and
Farmer, “The Status of the Person in Early Ming Law,” 75–77.
20 Hierarchy derived from the principle of zhengming 正名, or “rectifying names,”
according to which denominations have to correspond to realities, the form has to match
with the essence. If names were “rectified,” people would accept their lot, and social order
would be secured. See Confucius’ Analects XIII:3 and the famous echoing sentence in XII:11:
“A ruler ought to behave like a ruler, a minister like a minister, a father like a father, and
a son like a son.” It was also reflected in the notion of mingfen 名分, distinguishing (lit.
“cutting up” or “dividing”) names or statuses, which was essential in the shaping of social
relationships and especially in the engineering of master-bondservant relationships.
244 claude chevaleyre

and depended on various factors such as personal relationship, kinship

ties, age, rank of birth, gender, or membership in a special category (like
scholar-officials or bondservants). The relative value of hierarchical posi-
tions was for example embodied in the system of the “five degrees of
mourning” (wufu 五服)21 which was intended to determine the close-
ness of kinship ties: the tighter the kinship ties, the longer the duration
of the mourning, and the greater the gravity of the crime or offence was
modulated. In this system, and depending on circumstances, a man was
a father, a son, a brother, a husband, a master, etc. Cursing one’s father
or master, for example, was considered a more serious offence than curs-
ing a stranger; while cursing one’s own son or bondservant had no legal
consequence.22
From our modern point of view, Ming law was obviously discrimina-
tory, in particular against women, juniors and inferiors. But the Code was
not the source of discrimination. It only reflected and acknowledged dis-
criminations that were constitutive of the social order it was designed to
preserve. Reading the Code only informs us on the existence and opera-
tion of this social order, not on its core nature. There is no need to scan
the whole content of the Code to come to the conclusion that the wider
the range of potential hierarchical positions one could assume, the higher
one’s social status. In presence of his parents, even the highest official
was a junior and inferior, but his position made him superior to ordinary
subjects. Depending on the period of his life and on the person he faced,
he could experience a wide range of hierarchical positions. If women were
inferior in most configurations, even the humblest ordinary woman could
also be a mother or the mistress of a bondservant. Her hierarchical posi-
tion was far less variable than that of a high official. But her inferiority
as a woman was, in theory, less absolute than that of bondservants, who
were inferior even to ordinary women. The problem, in our perspective,
comes from the fact the Great Ming Code was intended for an audience of
officials and ordinary people familiar with the concepts, values and cat-
egories it referred to. In the same manner as it was not deemed neces-
sary to specify who was a woman and why a woman was inferior, there
was no need to explain which persons were bondservants and why they
were inferior. As a consequence, one would search in vain for precise

21 See Norman A. Kutcher, Mourning in Late Imperial China; and Ann B. Waltner, Getting
an Heir, 53–54.
22 Gao Ju, Da Minglü jijie fuli, 21:3b (law no. 350, following the numbering introduced
by Jiang Yonglin in his translation of the Great Ming Code, op. cit.).
 acting as master and bondservant 245

definitions of inferior categories or the nature of their inferiority in the

pages of the Great Ming Code.
When dealing with social stratification in late imperial China, one
usually refers to the conventional assertion according to which Chinese
society was organized along “four categories” (simin 四民): that is, the
literati, peasants, craftsmen and merchants (shi, nong, gong, shang 士農
工商).23 Slight variants can sometimes be found, such as the assertion
in the Ming huiyao presenting the Ming realm as a three-level society:
“One [level] is called ‘the people’ (min), another one ‘the soldiers’ ( jun 軍),
and another one ‘the craftsmen’ ( jiang 匠).”24 However, it is usually
acknowledged that the major line of social discrimination was the line
separating scholar-officials (shi)—those who at least had qualified for
(or bought) the lowest grade in the Chinese examination system—from
the rest of the population—the ordinary people (min 民, pingmin 平民,
shumin 庶民 etc.).25 But if the four categories of people represented the
great majority of the population and were considered to be “good people”
(liangmin 良民, that is to say ordinary or common subjects), then what
about the relatively small category of “mean,” or “debased people” ( jianmin
賤民), to which the Code often refers in opposition to the good people,
and among which bondservants were clearly the most numerous?
The category of mean people was composed of a variety of social groups
who lived in very different conditions, whose fate could be either painful
or pleasant and who had little in common except the fact that they all
belonged to the same legal group. Yet, again, neither the Code nor any
other normative source explains who was included in this category and
what the source of their inferiority was.26 If we are to believe the statutes
in the Great Ming code, they were essentially prostitutes and professional
entertainers (changyou 倡優), and bondservants (nubi 奴婢).27 But it is
acknowledged that this category also included more marginal groups,

23 This reference to “four categories” dates back to antiquity; Gu Yanwu’s (1613–1682)

Record of Daily Knowledge traced its origins back to the Guanzi (seventh century B.C.E.).
See Gu Yanwu, Rizhilu jishi, juan 13, 439.
24 Long Wenbin, Ming huiyao, juan 50, 936.
25 To give just an example, in his modeling of social hierarchy in imperial China, Rom-
eyn Taylor does not mention inferior categories; see Taylor, “Chinese Hierarchy in Com-
parative Perspective.”
26 It shall also be added that some doubts remain about the internal hierarchy among
all these groups.
27 The Imperially Endorsed Collected Statutes of the Great Qing clearly stated it: “The
four [categories of] people are ‘good;’ bondservants, entertainers and yamen underlings
are ‘degraded.’ ” See Qinding da Qing huidian, 17:1b.
246 claude chevaleyre

such as the “musician households” (yuehu 樂戶), “fishermen of the nine

surnames” (jiuxing yuhu 九姓漁戶), a variety of yamen underlings (zaoli
皂隸), beggar households (gaihu 丐戶), etc.,28 to which statutes on mean
people were applied. But clearly, from a legal point of view, bondservants
and hired laborers were the most representative groups.29 There might be
a good reason for such a situation: Contrary to the other groups, bondser-
vants and, to a lesser degree, hired laborers were engaged in a personal
relationship with their masters or patrons. Without a master (either a
person specifically named “master” [zhu 主] or “family head” [jiazhang
家長]) and a master’s family to host them, bondservants could not have
existed. And, as mentioned above, the Code was designed to preserve a
social order in which familial hierarchy was essential. However, if the
Code applied harsh discrimination to mean groups, it never explained on
what this segregation rested. But the segregation was, in my opinion, far
more fundamental and, in theory, more impermeable than the one dis-
tinguishing scholar-officials from the common people, and the principal
expression of such discrimination was the exclusion of mean groups from
the examination system.

Exclusion from the Examination System and Personal Degradation

In his study of the fundamental characteristics of slavery, the French

scholar Alain Testart has argued that in every complex society, a slave
was always excluded from at least one of the (social) dimensions deemed

28 For more details regarding marginal categories, see Anders Hansson, Chinese Out-
casts. Hansson recalls that “there is no extant law or regulation from the Ming or early
Qing dynasty stating that the musicians’ households and the other groups were legally
mean.” Yet, the one thing they had in common is “the manner in which their legal eman-
cipation was initiated” in the Yongzheng (1723–1735) reign; Ibid., 2.
29 Another category frequently appears in the Code alongside bondservants: hired labor-
ers (gugongren 僱工人). We should also consider law no. 394 entitled “Slaves or hired
laborers committing fornication with household head’s wives,” or the numerous statutes
associating these two categories. Yet, it is still very difficult to ascertain that hired laborers
were also of mean status. They actually suffered a legal discrimination, but there was no
provision for crimes committed by hired laborers against commoners outside the family
that employed them. Moreover, when they committed a crime against the family head,
the sentence was in most cases lighter than the sentence prescribed for bondservants who
committed the same crime. And, finally, a new sub-statute was promulgated in 1588 that
specified that laborers hired on a contractual and short-term basis would be sentenced like
ordinary persons. See Wu Yanhong, “Dui Mingdai beiyou falü diwei de kaocha,” 382–390.
 acting as master and bondservant 247

essential by the society in which he lived.30 In China, one of the essential

prerogatives of which bondservants were deprived was that of attending
school and participating in the examination system. In so doing, they were
denied a fundamental common (though very theoretical) prerogative
shared by all ordinary (male) subjects: the potential access to academic
ranks, bureaucratic responsibilities, social prestige and political power. Of
course, lining up a series of legal impediments does not suffice to reveal
the nature of a social status, but tracing the origins of such incapacities
can help understand them.
In the Ming, bondservants were not explicitly banned from educa-
tion, examination or an official career. The Code simply made no such
provisions. However, a number of sources clearly suggest that the mean
categories (including bondservants) were effectively banned from the
examination system.31 A 1384 regulation promulgated by the Ming founder
stipulated: “As for the provincial examination, educational officials (xue-
guan 學官) and dismissed officials, prostitutes, entertainers and their rel-
atives, yamen runners and their companions, and those who mourn their
father or mother, are in no case allowed to take the examinations.”32 In
1449, the Zhengtong emperor (r. 1435–1449) stipulated that only outstand-
ing students could attend district and prefectural schools, and that it was
thenceforth “prohibited to improperly accept prostitutes and entertainers,
yamen runners and persons of degraded status.”33
Be that as it may, the laments of early and mid-seventeenth century lite-
rati over the schools’ decay also suggest that bondservants were included
in these dispositions. Wu Yuancui, for example, noticed that:
Sons of prostitutes, entertainers and yamen runners are not allowed to
enter schools. Yet, recently, the institutions have been undermined. There
are [too] many people in schools, among whom are numerous sons of
slaves. They are even more numerous in the Wu (Jiangsu) prefectures of
Su[zhou], Song[jiang] and Chang[zhou], and in the Zhe[ jiang] prefectures
of Hang[zhou], Hu[zhou] and Jia[xing]. In the worst cases they even attain

30 Alain Testart, L’esclave, la dette et le pouvoir, 24. This assertion does not imply that I
consider Chinese bondservants as slaves in the strictest meaning. But this approach seems rel-
evant when extended to the search for common characteristics shared by inferiors in general.
31 Considerations on the relation between exclusion from the examination system and
personal degradation owe a lot to the pioneering research of Kishimoto Mio. See in par-
ticular Kishimoto, “Mindai no shakai shūdan to ‘sen’ no kannen;” and idem, “Maojuan
maokao susong yu Qingdai difang shehui.”
32 Da Ming huidian, 77:2b; this regulation is repeated in a slightly different version in
Ming Taizu shilu, juan 160, 2468.
33 Ming Yingzong shilu, juan 183, 3560.
248 claude chevaleyre

top places at the examinations, enter the Hanlin academy and hunt for
the highest positions. Isn’t that an offence against the Court? Aren’t they a
shame for the scholars of our time?34
Wu makes an informative parallel between entertainers and yamen run-
ners on the one hand and “slaves” (nuli 奴隸, a common word for bond-
servants) on the other hand. None of those categories had their place in
schools and they were a shame for contemporary literati, even though
bondservants were not mentioned in early Ming legislation. Guan Zhi-
dao (1536–1608) shared the same views when he wrote that bondservants
were identical to entertainers and yamen runners, and that they could not
therefore “mingle with the class of officials”:
If we scrutinize laws and regulations, even though there are statutes on [the
way] bondservants have to pay homage to the family head35 and on the
prohibition against offending the family head, yet, only Dukes, Marquises
and officials of the third rank and above are allowed to keep bondservants.
[People] sentenced to servitude and whose property has been confiscated
are only offered as slaves to families of outstanding officials, but never to
[regular] officials. [Ordinary] officials and the common people name them
“adoptive sons,” not “bondservants,” because only meritorious and noble
[men] can govern ordinary people; ordinary people do not govern their own
kind. Those named “adoptive sons” are only one degree inferior to natu-
ral sons; they have never been prohibited to study and become officials.
Those who are named “bondservants” are identical to prostitutes, entertain-
ers and yamen runners; they cannot be allowed to mingle with the class of
officials.36
Then, if bondservants (and the mean categories in general) were mani-
festly excluded from the examination system, it may be interesting to ask
why and to inquire about the nature of this exclusion. Ming sources are
relatively silent, but the evidence leads us to the conclusion that: (1) the
mean status was deeply rooted in conceptions of personal degradation;37
(2) that this degradation resulted from a sort of stain or pollution; and
(3) that such pollution was either provoked by polluting activities (like
that attached to “coroners” [wuzuo 仵作], who were in direct contact with
dead bodies), or derived from criminal activities and physical injuries.

34 Wu Yuancui, Linju manlu, 2:5a–b.

35 This may refer to a 1393 edict on “Rituals of salutation among ordinary people” that
included regulations regarding the manner bondservants had to pay homage to ordinary
people; see Mingshi, juan 56, 1428.
36 Guan Zhidao, Congxian weisu yi, 2:61a.
37 Jian 賤 originally meant “depreciated” in the sense of “low-priced.”
 acting as master and bondservant 249

In 1522, the newly appointed Shanxi vice Educational Censor, Lu Shen

(1477–1544), expelled from school the son of the favorite court musician
(yuegong 樂工) of the Prince of Jin. He argued: “It is preferable to have one
less student at school, rather than having [a single] one who pollutes the
school.”38 Mean categories (here, in Shanxi, including the households of
musicians) were regarded as “polluted” (wu 污) and likely to corrupt pure
students. And when such impurity was not due to polluting activities, it
was intimately related to the degradation generated by criminal activity,
legal punishment and physical injury, as suggested in a 1454 proclamation:
“prostitutes, entertainers, yamen runners and persons who have suffered
punishments, who mourn [a parent], and who have committed faults or
crimes, are not allowed to take the examinations.”39 Alongside entertain-
ers and runners, criminals and those who had suffered legal punishment
were similarly excluded from the examination system.40 Physical injuries
and condemnation resulting from a crime were thus of the same nature
as the pollution suffered by mean people.
A very interesting source for the study of mean people’s exclusion from
the examination system is the 1814 edition of the Qing Imperially Endorsed
Complete Guide for Education Commissioners.41 A moral guide, practical
manual and compendium of legal jurisprudence on the functioning of
education and schools, this text had no equivalent in size and precision in
the Ming.42 The importance given to the compilation of such documents
covering every aspect of public education can be regarded as a demon-
stration of the Qing rulers’ strong will to control and reorder (I would
say re-interpret) Chinese society, notably by rectifying and clarifying the
contours of social status. This can be illustrated by the promulgation of
the so-called “emancipation edicts” of the Yongzheng era,43 or the sharp

38 He Liangjun, Siyou zhai congshuo, juan 17, 148.

39 Ming Yingzong shilu, juan 237, 5173.
40 The exclusion of those who had to mourn a parent was different by nature. In order
to fulfill the requirements of mourning, officials and students had to retire either from
office or from school during a fixed period.
41 For an overview of statuses and examinations as presented in Xuezheng quanshu, see
Kishimoto, “Maojuan maokao susong,” 148, passim.
42 Local guides for Educational Commissioners existed in the Ming, such as those col-
lected and reprinted in Yang Yifan and Liu Ducai, Zhongguo gudai difang falü wenxian.
However, these are short practical manuals that can compare with the Qing compendium
neither in size nor by nature.
43 The problems raised by the Yongzheng “emancipation edicts” are thorny and end-
less. These edicts, promulgated between 1723 and 1730, were essentially meant to eman-
cipate regional marginal groups such as Shanxi musician households, the “fallen people”
of Zhejiang, hereditary servants of Anhui, beggars of Suzhou prefecture, groups of boat
250 claude chevaleyre

inflation of sub-statutes on bondservant status in the Great Qing Code.

The Qing dynasty’s systematic attempts to deal with status strongly con-
trast with the relative laissez-faire practices that characterized the more
porous and mobile late Ming society.44 Still, I assume that the Complete
Guide for Education Commissioners of the Qing promoted principles and
ideas that were relevant—though less clearly formulated—in the Ming.
Chapter 43 is entitled “distinguishing statuses” (qubie liupin 區別
流品). It mainly consists of a series of judicial cases devoted to deciding
whether or not a candidate or his ancestors were of mean origins and, as
a consequence, if the candidate could be authorized to take the examina-
tion. Thus, some Qing documents explain more clearly than any Ming
source why the sons and grandsons of emancipated bondservants could
not attend school and take the examinations: Would they have been able
to do so, they could have become officials and posthumous honors could
have been bestowed to their parents or grand-parents (that is to say, to
bondservants).45 One pitiful case attracted my attention, as it implicitly
confirms that mean people were considered polluted in the same manner
as criminals and the physically injured.
Sometime after 1800, the Shandong provincial Governor was sent the
case of a young boy named Zhang Shanchang for deliberation. In 1797,
Zhang had been sexually abused (jijian 雞姦)46 by a man named Zhao
Jie. Zhao was sentenced to death by strangulation and executed after
the Autumn Assizes of 1800. But the case of Zhang himself still had to be
deliberated because he was about to take the examination:
[. . .] For students, taking the examinations is the beginning of official career.
This is the reason why the established regulation emphasizes before any-

people, etc. In my opinion, these edicts were not a first step toward a general emancipa-
tion of mean categories, but rather an attempt to clarify social organization by suppressing
abusive mean statuses. For an in-depth exploration of the process and the motivations of
the Yongzheng emperor, see Hansson, Chinese Outcasts, 163–170; see also Ho Ping-ti, The
Ladder of Success, 18–19, who regards it as a complete manumission whose effect was to
“open the main avenue of upward social mobility to them.”
44 Porosity and the blurring of social boundaries are hallmarks of late Ming society.
See, among others, Wiens, “Masters and Bondservants,” 57; and Ho, The Ladder of Success,
53–125. This feeling of social mobility and a general impression that status was more a
question of practice vividly appear in seventeenth century judicial casebooks. My own
reading of judicial cases led me to the conclusion that disputes and conflicts over social
status were often adjudicated on a case-by-case basis and that concrete configurations
prevailed over appellations.
45 See examples in Xuezheng quanshu, 43:4a–5b, 43:20a–22b, 43:24a–b, 43:25a–26b.
46 See Matthew H. Sommer, “The Penetrated Male,” for a study of homosexuality in
Qing China.
 acting as master and bondservant 251

thing the purity of [the student’s] person and family. If [the candidate] has
been [physically] wounded during a [legal] punishment or if he has com-
mitted a crime or an offence, he may not be permitted to take the examina-
tion. Zhang Shanchang, the son of Zhang Mu, a Shandong commoner from
Dezhou, has been physically polluted. [Such a demeaning] is in no way dif-
ferent from a wound caused by a punishment. Thus, it would be improper
to permit him to take the examinations.47
The case of Zhang Shanchang can certainly not be compared with that of
bondservants or other mean people. He did not belong to a mean category
but was the son of a Shandong commoner (min 民). Yet the argument
of the Governor establishes a striking parallel between criminals and
degraded individuals. It clearly states that persons who had committed a
crime or an offence were “polluted” (beiwu 被污), in the same manner as
those who had suffered the wounds of physical punishments or injuries.
As a consequence, all were denied the right to take the examinations, just
like mean categories, who were also denied that right for not being “pure”
(qingbai 清白, the prerequisite for becoming official). It is tempting (in
an assumed syllogistic approach) to regard every group of mean people
as suffering a sort of personal degradation comparable in nature to that
suffered by criminals and physically or sexually injured persons. Indeed,
the fate of many groups of mean people was supposed to be the result
of an original condemnation. In Zhejiang province, “fishermen from the
nine surnames” were supposed to be offspring of Zhu Yuanzhang’s enemy
Chen Youliang (1320–1363),48 and musician households in Shanxi (like the
favorite of the Prince of Jin mentioned above) were also deemed to be the
offspring of Jianwen loyalists.49
There is one last argument likely to support the idea that bondservants’
personal degradation was of a similar nature and intimately linked with
the debasement of criminals. When considering early Ming legislation, it
appears that bondservants must have been relatively rare and mainly for-
mer criminals. Neither coercion nor marriage could have produced new
bondservants. The Great Ming Code clearly protected commoners from
being constrained to servitude50 and also prohibited unions between

47 Xuezheng quanshu, 43:23b.

48 Hansson, Chinese Outcasts, 133.
49 The Jianwen emperor (r. 1398–1402) was the grandson and designated heir of the
Ming founder. He was deposed in 1402 by his uncle who reigned as emperor Yongle until
1424 (see Hansson, Chinese Outcasts, 60–61).
50 Laws no. 85 and no. 108; see Gao Ju, Da Minglü jijie fuli, 4:11b–12a, 6:6b.
252 claude chevaleyre

commoners and mean people.51 Even voluntary subordination must have

remained marginal. Only a minority of officials and dignitaries were per-
mitted to possess bondservants,52 and such possession by ordinary people
was formally forbidden:
Families of commoners who shelter and foster bondservants [will be sen-
tenced] to one hundred strokes, and [the bondservants] will be released and
recover [the status of ] good [people].53
As the Code also stipulated that relatives of certain categories of criminal
were to be sentenced to enslavement to serve officials (though this was
not one of the regular “Five punishments” [wuxing 五刑]),54 it is possible
to say that the intention of Ming legislators was to limit the possession
of bondservants to officials and dignitaries, and to restrict the origin of
bondservants to legally sentenced criminal relatives.
This is also confirmed by early Ming decrees promulgated to free peo-
ple who had been sold (or who had sold themselves) because of poverty
or as a consequence of wartime disorder:
During the fifth year [of Hongwu, i.e., 1372], it was decreed that ‘[now] that
all under heaven has been mostly pacified, rites and customs cannot be
[left] un-rectified. All those who have suffered from the disorders and have
been enslaved to other people will recover [the status of] commoners.55
Similar decisions that clearly enforced such principles can be found in
an “Imperial edict on rectifying rites and customs,” also dated 1372,56 or
in The Veritable Records of Ming Taizu.57 In extreme situations, local

51 Law no. 121; see Gao Ju, Da Minglü jijie fuli, 6:33b–34a. A woman’s status was deter-
mined by the status of her husband. This is the reason why the law especially focuses on
women of commoner origin that married or had intercourse with bondservants.
52 A regulation dated 1391 stipulates that Dukes and Marquises could not foster more
than 20 bondservants; officials of the first rank, no more than 12; officials of the second
rank, no more than 10; and officials of the third rank, no more than eight. It also states
that captured children and families of criminals were mostly offered as bondservants to
families of meritorious men. See Long Wenbin, Ming huiyao, juan 52, 969–970.
53 Law no. 84 on the nomination of an illegal heir; see Gao Ju, Da Ming lü jijie fuli, 4:9a-
b. In his commentaries, Gao Ju adds that “only the families of meritorious men could raise
bondservants;” ibid., 4:11a. However, these regulations were not precise enough to prevent
equivocal interpretations; for more details on that problem, see Wu Yanhong, “Dui Ming-
dai beiyou falü diwei,” 390, and Wu Zhenhan, “Mingdai de zhupu guanxi,” 147.
54 An example can be found in Fu Fengxiang, Huang Ming zhaoling, 2:7b. It stipulates
that those who employed castrated children were to be castrated themselves and given to
officials as bondservants.
55 Long Wenbin, Ming huiyao, 950.
56 Fu Fengxiang, Huang Ming zhaoling, 2:4b–5a.
57 Ming Taizu shilu, juan 73, 1352.
 acting as master and bondservant 253

administrations could even be ordered to redeem children who had been

sold.58 However, as is well known, the intentions of the Ming founder
soon became obsolete and were unscrupulously violated or by-passed.59
But what remained, despite the loosening of legal or social practice, was
a commonly assumed appreciation that mean and bondservant status
rested on the idea of personal degradation.
The elements described above are not sufficient to explain the differ-
ence between bondservants and other groups of mean people: personal
degradation, impurity, marginality, exclusion from the examination sys-
tem and a series of legal incapacities were features that all mean groups
(at different degrees) had in common. Yet, what distinguished bondser-
vants from other mean categories was the unique relationship they shared
with their masters. In exploring the core nature of their status, we first
ought to explore this particular link, rather than focus on the confusing
variety of conditions that can be observed when reading the sources.

A Concrete Relationship Based on Reciprocity

A bondservant, whatever his origin, the terms of his contract, the duration
of his servitude, his occupation, or the harshness of his condition, was
considered a family member subjected, like any other, to the authority of
the family head. As such, he or she was considered an inferior or junior
(beiyou 卑幼) who was to obey the orders of the family head and other
senior or superior family members (zunzhang 尊長). His or her relation-
ship with the master mimicked kinship ties, and when offences or crimes
were committed, the statutes in the Code usually cited him or her among
the other junior members of the family. In so doing, the law submitted
him or her to the highest possible familial authority and to the control
of the person one had to respect the most—one’s father. But, of course,
bondservants were not their master’s children, they could not have a
share in the property, and they would never themselves become senior
or superior members of their master’s family. Contrary to the master’s

58 See the decree on Henan situation, in Ming Taizu shilu, juan 177, 2688; or the decree
on Xuzhou prefecture (dated 1413) in Ming Taizong shilu, juan 140, 1687–1688.
59 To circumvent the law, commoners sometimes named their bondservants “adoptive
sons” and established adoption or marriage contracts rather than bondage contracts (see
the quotation from Guan Zhidao, Congxian weisu yi, cited above). These practices were
regularly denounced for diverse reasons by late Ming thinkers such as Hai Rui, Shi Tianqi
and Gu Yanwu. For more details, see Waltner, Getting an Heir, 86–87.
254 claude chevaleyre

son or grandson who, one day, would become master himself, a bond-
servant would always remain a bondservant. Yet, although the personal
relationship between master and bondservant was obviously dissymmet-
ric, and though it only mimicked kinship ties for moral and legal pur-
poses, it was nevertheless not simple and absolute subjugation. As will
be demonstrated in this second part, this relationship was deeply rooted
in Confucian-inspired principles of solidarity and reciprocity, which were
not always paternalistic and vacuous justifications, and were sometimes
taken very seriously by Chinese magistrates.

“Righteousness” and “Benevolence”:

The Basis of Master-Bondservant Relationship

In Ming China, the tie binding master and bondservant was supposed to
be based—at least this was the theory—on mutual aid and reciprocity.
In an empire where Confucian values were proclaimed as official ortho-
doxy and played the role of a universal matrix in the shaping of all human
relationships (both in family and in society), the pursuit of self-interest
was regarded with disdain.60 In this context, the subjugation of others for
the sole sake of personal profit could not be considered as morally valid.
In other words, exploitation was not openly assumed and was disguised
behind a more or less sincerely benevolent (or paternalistic) rhetoric. In
a passage of the Jianwen lu, Zhang Lüxiang (1611–1674) inveighed against
masters who abused their simple-minded bondservants at the cost of per-
sonal solidarity: “They are naïve (shiduan 識短),” he said, “and this often
leads masters to disdain solidarity (qu yi 去義) for the [sole] sake of per-
sonal interest (jiu li 就利).”61
In structuring the master-bondservant relationship, the principle of
yi 義 (often translated as “righteousness”) was undoubtedly essential.
Often opposed to self-interest (li 利) and as the corollary of the virtue
of “humanity” (or benevolence, ren 仁), yi was a primordial value in the
Confucian social order. It mostly referred to appropriateness in human
relationships, but also to loyalty and solidarity.62 For example, collective

60 Analects IV:16: “The mind of the superior man is conversant with righteousness; the
mind of the mean man is conversant with gain.” Translated in Legge, Confucian Analects,
170.
61 Zhang Lüxiang, Yangyuan xiansheng quanji, juan 31, 894.
62 These are broad simplifications derived from the scholarly and philosophical debates.
I only intend here to present how such principles were practiced, simplified and applied
in daily life, and how they shaped master-bondservant relationships. For more substantial
 acting as master and bondservant 255

estates kept by important clans to support clan schools, clan ceremonies

or for charitable purposes were named yichan 義產, a term convention-
ally translated as “charitable estates,” though they were dedicated to soli-
darity rather than to simple charity. Ming loyalists who refused to submit
to Manchu rule were in the same manner named yijun 義軍, literally “yi
armies,” that is to say troops of loyal subjects whose solidarity towards the
Ming court never failed. And bondservants who remained loyal to their
masters in extreme circumstances—sometimes to the point of self-sacri-
fice for the sake of one’s master—were named yipu 義僕 or yinu 義奴.63
Persons who had the sense of yi thus acted appropriately; their conduct
was in conformity with the role society had assigned to them; they dem-
onstrated a strong attachment to hierarchy and social order, while those
who did not were condemned for being “rebellious” (pan 叛) and a source
of social disorders.64
Indeed, yi was not only an abstract notion. In practice, it was the cement
of contractual or non-blood familial relationships. When an outsider
entered a new family as a wife, as an adoptive son or as a bondservant,
he or she engaged with the new family head (and his blood-relatives) in
a relationship based on reciprocity and on the sense of appropriateness
implied by the bond of yi.65 He or she thus accepted a shift of author-
ity and undertook to play a new role, and to act as a wife, as the son
of new parents, or as a bondservant.66 Such a shift of identity was not
only achieved through the performance of ritual ceremonies, but also
through the establishment of a contract.67 A contract was the symbol of the

references on the link between “humanity” and “righteousness,” see Mengzi 6:A:11 or 3:B:9;
translated in Legge, The Works of Mencius, 290, 155–160; and see the discussion between
Yang Shi (1053–1135) and Cheng Yi on Zhang Zai’s (1020–1078) perception of the relation-
ship between ren and yi, in Yang Shi, Er Cheng cuiyan, 33.
63 We can cite, for example, the courageous Wang Chang, who braved all dangers to
get food for his master’s family during the early 1640’s famine; see Zeng Yuwang, Yiyou biji,
7–8; or the case of Xu Da, the bondservant of Ma Shiqi, who, out of gratitude and loyalty,
sacrificed himself on the grave of his deceased master after the fall of Beijing in 1644; see
Wuxi Jinkui xianzhi, 25:12.
64 Pan 叛 is often used in the rhetoric devoted to condemn late Ming bondservants. It
is frequently employed in official sources such as local gazetteers.
65 For example, an adopted son was named yinan 義男, or a “son [bound by] yi;” regard-
ing wives, one of the reasons legally accepted to pronounce a divorce was the “rupture of
the yi [bond]” ( yijue 義絕); see law no. 123 in Gao Ju, Da Minglü jijie fuli, 6:37b.
66 This did not imply a total rupture with one’s blood-relatives. See Waltner, Getting
an Heir, 53–71, passim.
67 I use the word “contract” here despite the fact that there was no contract law in
Ming China. Contracts were thus more comparable to private agreements that could be
presented at court, but were not strictly legally binding.
256 claude chevaleyre

acceptance of a new identity, rather than a legally binding list of duties

(even though material conditions were often specified in such docu-
ments). When becoming a bondservant, one accepted not only to assume
the duties and respect the conditions listed in the contract to which one
had appended one’s hand-prints, but also to act as bondservant, that is
to say, to assume the role of a bondservant in such a way as to make the
name “bondservant” fit with the expected, corresponding reality.
The requirement to make names and titles correspond to realities was
indeed central to the Confucian conception of social order. The word that
most fits our conception of status in traditional Chinese was the word
mingfen 名分 (lit. “name-distinction”), itself a by-product of the idea of
“rectification of names” (zhengming 正名), central in Confucian ortho-
doxy.68 Regarding bondservants, this idea was expressed in the phrase
“name distinction between master and bondservant” (zhupu mingfen
主僕名分), which was used to recall that their actions had to correspond
with what was expected from someone named bondservant.69 Name dis-
tinction and righteousness (or solidarity) were thus complementary and
inseparable notions. Such a connection found its ultimate expression in
the legal statute entitled “violating names (or roles) and offending against
righteousness” (ganming fanyi 干名犯義).70 This statute was not exclu-
sively concerned with bondservants but with role transgression in gen-
eral. It notably specified the adequate sentence for whoever accused his
father or his master, a most serious offence against family hierarchy and
solidarity.

68 If the idea of “rectifying names” appears in the Analects, it is more developed and
more crucial in the Xunzi (book 22 is specifically entitled Zhengming). Differentiation plays
an essential role in Xunzi’s vision of social order and harmony. Distinguishing catego-
ries and names or appellations is crucial in that it permits the expression of the complex
realities of the world and the maintenance of social order. In book 9, Xunzi extended
the meaning of the famous sentence “the ruler ought to behave like a ruler, the minister
like a minister, etc.” and added “the peasant ought to act like a peasant, a gentleman like
a gentleman,” etc. See the translation in Homer H. Dubs., The Works of Hsüntze, 136. He
then demonstrated that the rationale underlying the correspondence between appella-
tions and realities was essential and could be extended to all human relationships. For a
deep investigation on “names rectification,” see Michael Lackner, “La portée des événe-
ments,” 79–80.
69 The sixth section of Xiao Yong’s Agreement of the Chishan Association, for example, is
dedicated to the requirement of “Rectifying Status” (zhengfen 正分); see Xiao Yong, Chis-
han huiyue, 7. Guan Zhidao’s Proposals for Safeguarding Customs also includes a text that
suggests to “correct master-bondservant [relationship] according to the laws of the Spring
and Autumns period” ( yi chunqiu zhi fa zheng zhupu 以春秋之法正主僕); see Guan
Zhidao, Congxian weisu yi, 2:60a–63a.
70 Law no. 360; see Gao Ju, Da Minglü jijie fuli, 22:30a–31b.
 acting as master and bondservant 257

“Name distinction between master and bondservant” implied that bond-

servants had to act as bondservants. They were expected to work hard, to be
obedient in all circumstances and to show respect at all times. Such expecta-
tions were frequently repeated in family or clan rules and in contracts.71 Yet,
as the master-bondservant tie rested on the principle of reciprocity, masters
were also expected to act as masters. First, it has to be recalled that the
Great Ming Code offered protection to bondservants against undue physical
aggression and murder. Even though crimes or offenses perpetrated against
bondservants were less severely punished (as a demonstration of social or
familial hierarchy), the principle of protecting bondservants remained in
effect, despite a flagrant dissymmetry unfavorable to bondservants.
Then, masters were supposed to care for their bondservants. They had
to “foster them with benevolence” (enyang 恩養), to protect them from
cold and hunger, to provide them with food and clothes and even with a
wife or husband when coming of age. In the section entitled “Commiser-
ate with Inferiors” (xuxia 恤下) of the rules of the Chishan Association,
Xiao Yong (  jinshi 1583) stated:
Make sure to provide them with clothes and food, and to marry them in due
time. The Law stipulates that they have to be sentenced on the same basis
as sons and grandsons, out of respect for the importance of benevolence and
righteousness. Considering that they regard their masters as their own father
and mother, how could masters not regard bondservants as their own sons?72
Or, as mentioned in a typical text of “family instructions” (jiaxun 家訓):
“[you] must provide them with beverages and food in a timely manner,
make sure that they are not starving or suffering from cold, and balance
their time of labor and leisure.”73 In the Ming, fostering bondservants
was even regarded as a natural act of benevolence. According to Wang
Mengji’s Family Admonitions:
Families with ample resources must keep bondservants. They rely on our
provisions and we rely on their labor; when relying on each other we form a
household. [Yet], as they can work, can they not work for anyone else? Thus,
it would be wrong to pretend that they cannot make a living without us.

71 See, for example, the general clauses in a late Ming “man-sale” model agreement
reprinted in Zhang Chuanxi, Zhongguo lidai qiyue huibian kaoshi, 1011; or the “Itemized
rules for governing one’s family” (zhijia tiaoyue) in Zhuang Yuanchen, Zhuang Zhongfu
zazhu (rare manuscript held at the National Library of Beijing).
72 Xiao Yong, Chishan huiyue, 14. Treating bondservants with commiseration was the
main topic of section 21 and also of section 7, entitled “Spreading Humanity” (guangren
廣仁) in the principles promoted by the members of the Chishan Society.
73 Pang Shangpeng, Pang shi jiaxun, 9.
258 claude chevaleyre

There are statutes in the Code that sentence [criminals] to servitude with
officials. But who allowed scholars and commoners to keep bondservants?
Therefore, male bondservants are called “adoptive sons,” female ones “adop-
tive daughters-in-law,” and young girls “adoptive daughters.” They are all
named after the same way as our own daughters-in-law, sons and daughters.
Though there is a distinction between high and low, it is not of the same
kind as [the distinction between] human and animals.74
For late Ming scholars, wealthy families had thus “to foster” (xu 蓄) bond-
servants not out of pure charity, but for the mutual benefit of those who
had wealth and those who had their labor. In this typical, ideal and pater-
nalistic vision of social harmony, which probably did not match with the
great majority of their contemporaries’ practices, masters and bondser-
vants were regarded as complementary and inseparable; they “relied on
each other” (xiangyi 相依) to form a household. The moral demand for
reciprocity could hardly be expressed in a clearer way. When stating that
“it would be wrong to pretend that they wouldn’t survive without us,” the
author of this document even recognizes that masters benefited more of
the situation than bondservants. His text also establishes an interesting
distinction between criminals sentenced to servitude and those (prob-
ably the great majority at the end of the Ming) who were owned by ordi-
nary people or in the private service of scholars.75 He thus confirms the
fact that early Ming legislation regarding the possession of bondservants
was obsolete. Though of inferior status, private bondservants (who were
sometimes named adoptive sons or daughters in order to by-pass the law)
were still of the “same kind” (tonglei 同類) as their masters, according to
Wang. As such, and because they bore names that mimicked family rela-
tionships, they were to be considered as family members, he said, echoing
Xiao Yong and many other contemporaries.
Ascertaining that bondservants had to be treated and regarded as one’s
own children or family members was obviously the expression of a strong
paternalistic rhetoric. Such views were intended to provide justification
for subordinating others to one’s own authority. In regard to Confucian
orthodoxy, which emphasized the virtue of humanity, and to the law,
which should have prevented personal servitude, such commonplace

74 Wang Mengji, Wang Mengji jiaxun, 1b.

75 The author distinguishes “officials” (guan 官) and “scholars” (shi 士). This probably
corresponds to the distinction between, on the one hand, officials who possessed and used
bondservants in their public career, and, on the other hand, scholars inside or outside
officialdom who possessed private bondservants.
 acting as master and bondservant 259

rhetoric was motivated by the need to present acceptable arguments. Such

was probably the main goal of narratives that frequently recounted one’s
human attitude toward bondservants. Let us cite, for example, another
passage written by Zhang Lüxiang’s:
When Zeng Wenke was travelling north to take the metropolitan examina-
tion, a household servant in charge of guiding his donkey followed him. One
night they encountered heavy snow and extreme cold; [Zeng] found it natu-
ral to propose that the servant share his room. But his companions refused
it. [Zeng] thus replied: “He is also the son of a man. How could we suffer to
leave him alone in the cold?” Such were his outstanding virtues.76
Of course, if Zeng’s attitude toward his servant was the expression of an
“outstanding virtue” (houde 厚德), this also infers that such concern was
unusual or exceptional.
However, narratives of commiseration for bondservants’ conditions and
allegories of master-bondservant harmony appear frequently in Ming writ-
ings. But was it simply pragmatic concern for inferiors who were expected
to remain safe enough to provide efficient service? Was it purely paternal-
istic rhetoric intended only to provide acceptable justifications for per-
sonal domination? In most cases, it probably was. Yet, in an innovative
and persuasive article, the Taiwanese scholar Wu Zhenhan has provided
enough evidences to suggest that, in the Ming, concern for bondservants
was to some limited extent sincere. Wu has notably demonstrated that
such concern was not simply motivated by paternalism or compassion,
but by the emergent conviction that bondservants were of the same
nature as their masters and that they were also human. Such ideas mainly
developed under the popular influence of Wang Yangming’s (1472–1529)
individualistic and equalitarian idea of innate knowledge, according to
which all human beings shared the same potentiality to access wisdom.77
This also reflects in other social trends that are the hallmark of late Ming
society, such as the development of charitable activities78 or the spread
of women’s literacy and education. Nevertheless, such ideas probably
attracted only a small (and not quantifiable) minority of masters and had
a very limited influence on ordinary practices regarding servitude. But

76 Zhang Lüxiang, Yangyuan xiansheng quanji, juan 44, 1287.

77 Ivanhoe, Ethics in the Confucian Tradition, 121–134, passim. It should nonetheless be
added that such conceptions and attitudes were prepared long before Wang Yangming,
notably by Song-period thinkers like Zhang Zai (1020–1077), for example; see Ivanhoe,
op. cit., 22.
78 See Joanna Handlin Smith, The Art of Doing Good.
260 claude chevaleyre

they instilled some substance into the theoretical principle of reciprocity

that was supposed to govern master-bondservant relationships. Moreover,
they introduced a certain nuance into the usual prism through which
bond-servitude has normally been analyzed. In other words, the principle
of reciprocity was not completely meaningless in theory, nor was it totally
insignificant in practice.

Reciprocity in Practice in Late Ming Judicial Casebooks

New trends in legal history have amply demonstrated how fruitful the
study of judicial archives can be in getting closer to economic and social
realities. In this regard, reading late Ming judicial cases turns out to be a
particularly fruitful approach to explore the nature of the master-bond-
servant bond and examine the validity of the principle of reciprocity. I
will introduce a few examples here, beginning with a case from the Brief
Account of Judgments in Songjiang, an anthology of court opinions by Mao
Yilu (?–1629), the prefectural judge of Songjiang between 1606 and 1610.79
As a prefectural judge, Mao Yilu was in charge of handling ordinary judi-
cial procedures of revision and appeal transmitted by superior provincial
authorities. On the frontline of judicial operations, he was also frequently
consulted in cases issued from other jurisdictions because of his reputa-
tion as a judicial expert. But he was not simply a legal technician who
mechanically applied the codified penal law. When confronted with what
we would name civil matters, he had some latitude in pronouncing judg-
ments in a manner closer to civil arbitration than to legal punishment. In
so doing, he had to find in the end the best balance between sentiments
and law80 in order to get the compliance of the parties, to ease minds, pre-
vent further prosecutions and, ultimately, restore social harmony. Such
was sometimes the case in affairs involving bondservants, close to one-
fourth of the cases collected in the Brief Account.81

79 Mao Yilu, Yunjian yanlüe. See Pierre-Etienne Will, Official Handbooks and Antholo-
gies of Imperial China (forthcoming), “Anthologies of Official Documents by Prefects and
Magistrates,” 6.3; see also Jiang Yonglin and Wu Yanhong, “Satisfying both Sentiment and
Law.”
80 For a complete presentation of the legal proceedings handled by Mao Yilu and the
requirement to balance sentiments and law, see Jiang Yonglin and Wu Yanhong, “Satisfy-
ing both Sentiment and Law.”
81  Wu Yanhong counted 42 cases involving bondservants out of a total of 173; see
Wu Yanhong, “Dui Mingdai beiyou falü diwei de kaocha,” 392.
 acting as master and bondservant 261

The Brief Account is not composed of full verdicts enunciating the pre-
cise conditions of the affairs and the final sentence. The pieces collected
concentrate instead on the motivations of the sentences. Thus, they are
invaluable for legal historians trying to penetrate the legal reasoning of
Ming magistrates. On the other hand, their sketchy nature makes such
documents somewhat difficult to read and forces the reader to specu-
late about circumstances and incriminations. Yet, it includes some very
informative pieces regarding bondservants, such as case no. 10,82 a rela-
tively complex affair opposing the descendants of a master to those of his
bondservant.

The Quarrel among the Sun from Qingpu

Some sixty years before the case was brought to Mao Yilu’s attention—
that is, in the late 1540s—a young boy named Sun Cheng was sold (yushen
鬻身) to Sun De’s ancestors.83 As time went by, Sun Cheng eventually
established his own household (zili menhu 自立門户) and had two sons.
In a classic configuration, he first adopted a son he named Sun Qiao before
having a natural, though probably unexpected son whom he named Sun
Zong. The prosecution started up years after Sun Cheng’s death, when Sun
Qiao fell critically ill. As the two brothers had already determined sepa-
rately, Zong feared that he might not get control over Qiao and so divert
his property. Thus, in order to achieve his goal, Zong conspired with the
descendant of their father’s master, who made an accusation against Qiao.
It is not clear, however, how such a strategy was supposed to enable Zong
to get hold of his brother’s property. What is nonetheless clear is that the
crux of the case was to decide whether Sun De, the master’s descendant,
was himself the master of Sun Qiao, the adoptive son of a bondservant.
At the first hearing, the Qingpu magistrate concluded that Sun De was
not Sun Qiao’s master. However, considering that the old contract of his
father was still extant, he still sentenced Qiao to pay two ounces to Sun
De, so as to “demonstrate somehow the significance of the rites.” Yet he
also ordered Zong to pay half of the fine, arguing that as Zong was the

82 Mao Yilu, Yunjian yanlüe, juan 1, 415.

83 It should be underlined here that the masters and their bondservants bore the same
surname. However, no element in the case can help understand whether Sun Cheng was
simply a distant relative of his master’s family, if he had had to change his surname, or if
he had been officially “adopted” in order to conceal the fact that he had been bought as
a bondservant.
262 claude chevaleyre

natural son of the former bondservant, and, despite the fact that he was
on the side of the accuser, “the law could absolutely not spare [him] and
solely bear on Qiao.” As one might guess, Zong was not satisfied with the
sentence. Nor was Sun De, who was in a state of dire poverty and had not
received the two ounces. Both being driven by cupidity, the previously
allied Sun De and Sun Zong brought new charges with the help of an
unscrupulous intermediary. The content of the prosecution that followed
and the nature of the accusations are, again, not clear. What is clear is
that Mao Yilu regarded their attempts as vain manipulations unlikely to
deceive a sagacious prefectural judge. To the demands of Sun De, who
wanted to be considered as Qiao’s master in order to have authority over
his property, Mao Yilu answered: “Trying to keep people as bondservants
generation after generation with the help of a decades-old damaged piece
of paper is both greedy and stupid.” And as if this statement were not
sufficient to support his argumentation, he added: “If there is actually a
connection between Zong and De, when Zong’s father died and when his
mother remarried, how did it happen that [De] did not even mention
[this relationship]; and he has now uselessly forged a contract!?”84 For the
judge, Sun De was only “telling ghost [stories] and dreams!”
Finally, considering that there had been no material loss, Mao Yilu only
reprimanded Sun De and likewise Sun Zong for his revolting behavior. Yet,
the sentence itself is less important than the enlightening (and suppos-
edly enlightened) reasoning of the magistrate and prefectural judge. First,
it is worth noting that neither the district nor the prefectural jurisdiction
raised objections against the possession of bondservants from ordinary
birth. The document says nothing about the master’s family status, nor
does it explain the conditions that led Sun Cheng to sell himself or to be
sold as bondservant. Yet, reading such a case and similar examples, one
soon comes to the conclusion that in the late Ming the voluntary selling
of persons was regarded as a normal practice, though it was contrary to
the spirit of early Ming institutions. And if Mao Yilu rejected the validity
of the contract, his motivations were elsewhere. It is also striking to notice
that bondservants often adopted sons. If the reasons for such adoptions
are rarely mentioned, and if sometimes they clearly result from simple
pragmatism,85 they may also denote a strong will among bondservants to

84 It remains unclear whether the contract was authentic or not, since it is once
described as an outdated old piece of paper and then as a forged contract; see below.
85 Mao Yilu, Yunjian yanlüe, juan 7, 512.
 acting as master and bondservant 263

share the same ritual standards as ordinary people. Adopting a son was
a pressing need for men who could not father a son and who wanted to
perpetuate their lineage and fulfill their ritual obligations towards their
ancestors. This could also indicate that bondservants did not consider
themselves as totally excluded from the society they lived in, and that
they did not necessarily regard their condition as permanently, perpetu-
ally and hereditarily binding.
Above all, this case confirms that the principles of solidarity and reci-
procity underlying the master-bondservant relationship were not pure
abstractions. Without concreteness and proper behaviors, the ties that
bound them to each other could be considered as relaxed or broken.
Before the courts (probably more likely in ordinary, civil cases than in
penal affairs), claiming one’s authority over a bondservant with an old
contract in hand was not sufficient. One had to prove that one had always
acted as a master and never let the link lapse. Thus, when adjudicating
cases regarding doubtful bondservant identities, magistrates often resorted
to the same argument. If one could show that master and bondservant
had established separate households for long enough without a protest
on the part of the master, the reality of the relationship could be denied;
and so could the formal and moral link that gave a man authority over
another one. Indeed, it is not for nothing that, even before recounting the
sequence of events, Mao Yilu expounds two facts: (1) sixty years ago, the
young Sun Cheng was bought by Sun De’s ancestors; and (2) since that
time, “seasons have succeeded one another” and master and bondservant
“had established separate households a long time ago.” When reading the
first two sentences of the judgment, one can immediately understand the
crux of the affair and the way Mao Yilu would handle it.
Then, what about the contract and its legal value? In the present case,
one cannot ascertain whether the contract was forged or not. It simply
seems that the district magistrate who handled the case on first hearing
had no reason to consider it was a fake (“Sun Cheng’s old contract is still
extant,” he said), while Mao Yilu barely mentions an old piece of damaged
paper before regarding it as forgery. Nevertheless, as real as the document
might be, a contract was not necessarily a legally enforceable agreement.
It was the embodiment of a private agreement and the formalized expres-
sion of a mutual commitment; it could be presented in court and its
provisions could actually be held against a bondservant who refused to
fulfill his obligations. Could it be proved that one of the parties involved
had broken the commitments, the contract would be regarded as invalid
and totally empty words. In other words, if the reality it was supposed to
264 claude chevaleyre

guarantee had evaporated, the contract was no more than an obsolete

“old piece of paper.” And such was the case in Mao Yilu’s eyes. Not only
had the litigants established separate households for a long time, but the
master’s family had also not gotten involved in the funeral and wedding
rituals of their former bondservants. Whatever was written in the contract
and whatever its authenticity, for the two magistrates who handled the
case, the relationship could not be maintained because it no longer rested
on a tangible reality.
Whatever the precise sentence pronounced by Mao Yilu, he might have
followed the initial judgment of the district magistrate and ordered the
descendants of the bondservant to pay a sum of money to the descendant
of the master in order to “demonstrate the signification of the rites.” If the
termination of the solidarity between master and bondservant was mani-
fest in the eyes of both officials, the main problem was that it had never
been formally established. In consequence, the district magistrate—whose
first mission regarding civil matters was to ease minds and prevent fur-
ther prosecutions—tried to pronounce an arbitration both parties would
abide by. In this regard, the sum of money he ordered the two brothers
to pay to the master’s descendant was mostly symbolic.86 Materially, it
was substantial enough to be regarded as a correct compensation by Sun
De, but not to be considered as the price of a regular redeeming. It did
not mean that Sun Zong and Sun Qiao were bondservants who redeemed
themselves by paying two ounces. Symbolically, the sum was meant to
demonstrate “the significance of the rites.” On the one hand, it broke the
bond formally and definitively, but on the other, it was meant to remind
that as descendants of bondservants, Zong and Qiao remained hierarchi-
cally inferior to the family of their father’s master. If the material bond
was broken, moral and hierarchical traces still remained. But for the two
brothers, such diffuse discrimination and the payment of two ounces were
worth the cost for future peace of mind and formal emancipation. Had
everyone acted without selfdom motives, the initial arbitration would

86 As a comparison, in case no. 15 of Yunjian yanlüe, a certain Gao Zhong, who was
“undoubtedly” a bondservant (even though he was in fact the adoptive son of a bondser-
vant), was finally authorized to redeem himself. Gao Zhong had shown himself to be so
reluctant to remain as a bondservant that his master, who in a first move placed a com-
plaint against him, finally accepted to let him go. What could he do with someone who
refused obstinately to comply with his orders? In consequence, the master entered into an
agreement with the bondservant’s clansmen and accepted a price of 40 ounces (more than
1 kilogram of silver) to redeem Gao Zhong; that is to say, forty times the “compensation”
Sun De received for each son of his family bondservants.
 acting as master and bondservant 265

have morally and materially satisfied all the parties. It is precisely because
it was not the case that Mao Yilu showed so much anger against Sun Zong
and Sun De: cupidity was not morally acceptable.

The Jin versus Wang Case

As another example of the reprobation against the pursuit of self-interest

and the demand for concreteness in master-bondservant relationship, let
me refer to case no. 96 in Draft Court Opinions from Cases Personally Tried
as Inspector of the Wu Region, another late Ming collection, this one of
cases adjudicated by Su-Song Regional Inspector Qi Biaojia (1602–1645).87
As a Regional Inspector, Qi Biaojia held a key position in the process of
judicial appeal and review. Some time around 1635, one Wang Qing, his
accusers and all the written pieces related to a case of forged homicide
were transmitted from Kunshan district to Suzhou and submitted to Qi’s
expertise. Examining the ins and outs, he stated:
It has been established that Jin Cheng, the grandfather of Jin Zhong and
Jin Xiao, long ago sold his own son, Jin Peng, to the Wangs. Since then, [Jin
Peng] had always served [the Wangs] well. After the successive deaths of
every member with official status, the [Wang] family entered a downturn.
As a consequence, [Jin] Zhong and [Jin] Xiao left and established them-
selves elsewhere, just as if they ignored being of hereditary bondservant
status. Everyone since then concealed the fact that they were bondservants
of the Wangs, and this is the reason why Xu Shengcai accepted to have
[his daughter] married [with one of the Jin brothers]. Yet, in 1627, a Wang
clansman, Wang Qing, blamed them for having broken their ritual duties,
and had an altercation with them. Thus, Jin Xiao, who was also the most
cunning, unexpectedly seized the occasion and accused [Wang Qing] with
the hope of washing away the name of “bondservant.” That is where all the
judicial troubles began.
In the continuation of the document, Qi Biaojia recognizes that the Wangs’
claim of authority over the Jin brothers was legitimate. It was not, he said,
an empty claim invented to counter the Jins’ accusations. Moreover, not
only did Jin Xiao himself accuse Wang Qing, but he also led his neph-
ews (Ya and Yun) to accuse the Wangs of having murdered his brother

87 Qi Biaojia, An Wu qinshen xigao, 517–518. See Handlin Smith, The Art of doing Good,
248–278, passim; and see Will, Official Handbooks and Anthologies of Imperial China,
“Anthologies of Judicial Sentences and Model Cases,” 4.1.5.
266 claude chevaleyre

(Jin Zhong, who had died in the meantime). But the sagacious Regional
inspector was not fooled by their tricks:
However skillful their counter-attack ploy, they have overlooked the fact that
the distinction between master and bondservant still remains. According to
the principles and according to the law, they naturally should be sentenced
to return and serve [the Wangs]. However, for having falsely accused people
of homicide, Jin Ya has already been severely chastised. For the sake of the
Wangs, and considering that this bondservant has proved so very rebellious,
returning him to serve in their house would be like opening the doors to a
bandit, and it must be feared that one day some domestic tragedy might
occur. It is not necessary to have him handle housework by himself in order
to rectify the distinction between master and bondservant.
As to the fair judgment I am pronouncing today onward, Jin Xiao and his
nephews shall never place accusations against their master anymore. How-
ever, considering that Xiao was likely to arouse people’s attention because
he claimed that the Wangs wanted to take advantage of his wealth, there is
[only] one way for the Wangs to clarify their motivations: That the persons
remain as persons belonging to them, and the property remain as property
belonging to the Jins. The Jins are authorized to go to the district magis-
trate’s office and get a certificate that will guarantee them the perpetual
property of all the goods acquired by Jin Zhong and Jin Xiao [and presently]
in the Jin family. As literati with dignity, how would the Wangs care for such
unrighteous goods? In so doing, the reasons Xiao had to cause trouble and
go to court will have been suppressed. [. . .]
What should be noted in this case is the care with which Qi Biaojia
attempts to determine the status of each person before he tackles the
judicial issue. In the same manner as Mao Yilu, he examines who was
sold to whom and if there had been a rupture in the relationship between
master and bondservants. In this particular case, there was no doubt. The
Jins were undoubtedly the hereditary bondservants (shipu 世僕) of the
Wangs and the bond had never been formally interrupted until the Wang
family fortunes entered a financial downturn. Probably out of resources,
the Wangs were constrained to let them leave. The Jins thus enjoyed
a state of de facto emancipation and eventually acquired some wealth.
When the Wangs attempted to exercise their authority over their former
bondservants (and over their newly acquired property), the latter refused
to comply. And this was the source of all the subsequent prosecution that
Qi Biaojia was in charge of adjudicating.
What is striking in such a case is the discrepancy between legal practice
and theory in late Ming judicial proceedings. The Code clearly stated that
if a bondservant ever accused his master he would be subjected to caning.
Moreover, if the accusation proved to be false, he could be sentenced to
 acting as master and bondservant 267

death by strangulation.88 This provision did not exempt the magistrate

from investigating the case, but it offered the accused master quasi-total
immunity, whether or not he was guilty of something. As there was no
doubt regarding the status of Jin Xiao and his nephews, Qi should logically
have rejected the accusations and condemned the bondservants. How-
ever, he did not invoke the law, and, rather than pronounce a simple and
expeditious judgment, he chose to develop more complex considerations.
Given that he mainly aimed at demonstrating the faults of the bondser-
vants, one might deem that, in the end, he respected the spirit of the law;
but on the other hand one might also admit that he did not apply the
letter of the law. No doubt, in his eyes there were more urgent and more
essential problems to deal with, despite the offence committed by bond-
servants in accusing their master. He knew that reaching the appropriate
balance between moral standards and economic considerations was more
likely to ease minds than a rough application of the law. The commercial-
ized society of the seventeenth century was different from the early Ming
social order for which the Great Ming Code was originally designed. The
quarrel over the Jins’ status was in fact a pretext for concealing a some-
what banal—though very delicate at that time—problem of property.
Considering that the Jins were without doubt hereditary bondservants
(shipu 世僕) belonging to the Wangs and in order to satisfy both prin-
ciples and law (juli jufa 據理據法), Qi Biaojia could not allow the Jins
to “wash away the name of bondservant” (xituo numing 洗脫奴名); he
had to “rectify the distinction between master and bondservants” (zheng
zhunu zhi fen 正主奴之分). Yet, he also acknowledged that over a period
of time, the bond had loosened and no longer rested on concrete grounds.
In his eyes, this was not insignificant. Thus, he had to clarify whether the
Wangs were solely motivated by legitimate moral claims or if they only
“wanted to take advantage of [the bondservants’ newly acquired] wealth.”
If such was indeed the case, his skillful judgment would teach them a
good lesson. On the one hand, the judgment would likely satisfy their
moral requirements by rectifying social relationships, and the person of
Jin Xiao would remain their property. But on the other hand, it was for-
mulated so as to thwart plans only motivated by self-interest. The Jins
could definitely secure (shishou 世守) the property they had acquired,
and they would even get an official certificate to enforce the decision. As
a consequence, the Wangs could only show themselves satisfied with the

88 Law no. 360; see Gao Ju, Da Minglü jijie fuli, 22:30a–31b.
268 claude chevaleyre

judgment. What are a few “unrighteous goods” compared to the rectifica-

tion of social hierarchy and harmony, Qi asked? Showing any dissatisfac-
tion with the decision would immediately expose their cupidity. And this
would probably have other unfavorable consequences.
One point cannot be clearly elucidated in this case: Did the judgment
imply the dissolution of the hereditary character of the tie binding the Jins
to the Wangs? The situation had gone so far that the bond remained only
nominally. Xiao and his nephews showed themselves so rebellious that
maintaining them in their master’s house would be like “inviting a bandit
in one’s house,” Qi said. Then how was the bond supposed to be perpetu-
ated generation after generation if it was only nominal and if it rested
on no concrete grounds? Qi also recognizes that the Jins constituted a
“household” ( Jin hu 金戶) to which was granted the exclusive use of its
property. Yet, bondservants had no personal households and were usually
registered under their masters’ names. In other words, how could such a
judgment be reconciled in practice with the status of hereditary bondser-
vant? Did it not objectively imply the end of hereditary ownership?
In fact, Qi had probably another reason to try to invalidate the heredi-
tary character of the bond. One of the Jin brothers, probably Jin Zhong,
had married the daughter of Xu Shengcai, a commoner who ignored the
real status of his son-in-law. Since a woman’s status depended on her hus-
band’s, when married to Jin Zhong the daughter of Xu Shengcai would
then have become a bondservant herself. This is precisely the reason why
the law prohibited weddings between bondservants and commoners’
daughters.89 And this was surely something a father did not want for his
daughter—an anomaly Qi Biaojia was very keen to rectify. By restoring
the bond only nominally, and, while recognizing that the Jins constituted
a household, the Regional inspector rectified both the status of the bond-
servants and that of Xu’s daughter.
Finally, it is worth noting that Jin Xiao and his nephews considered that
a judicial offensive against their master’s family could potentially turn to
their advantage. Despite their status of hereditary bondservants, they did
not consider their condition as totally irrevocable, nor did they regard

89 Law no. 121, op. cit. In a Yuan (1277–1367) encyclopedia for daily use that was very
popular in the Ming, weddings between ordinary and mean people are depicted as unions
between entities of different nature that were to be dissolved: “Every person has a mate
who is to be of the same kind. Considering that commoners and degraded [people] are
different, how could they intermarry? Would transgression occur, [the mates] should be
separated and [the situation] rectified;” Huang Zongli (?), Jujia biyong shilei quanji, 324.
 acting as master and bondservant 269

their actions as bound to fail. We will never know on what their calcula-
tions rested. But one thing can be taken for certain: At that time, bond-
servants felt less constrained by the law or by their status than we usually
think. The picture emerging from legal cases does not contradict what
we already know about Ming bond-servitude, but it is likely to modify
the perception we usually have of a rigid and uniform system of personal
domination. Of course, Chinese bond-servitude was discriminatory, and
it was the result of economic pressures and domination. But it was not
the exclusively one-way system of exploitation described with insistence
by some historians. It was an unequal game of powers, but in the bal-
ance, bondservants were not totally vulnerable and powerless. When the
stronger let his attention wander, the situation was likely to turn to the
weaker’s advantage. If a master let his power weaken, this was likely to
reinforce the bondservant’s position, and, at least in extreme cases, lead
to the dissolution of the bond. We should naturally be cautious when
making generalizations based on a few judicial cases; in practice, masters
probably were not worried by the same moral considerations as magis-
trates, and it is clear that bondservants did not hesitate to refer matters
to courts and make use of every possible judicial trick to get out of their
situation. And, in court, social hierarchy and moral domination on the
one hand, and economic domination on the other hand, were two dif-
ferent and hierarchized things. Reciprocity was not an empty word, and
economic domination could only be the result of a pure moral domina-
tion that, in turn, implied obligations on the part of the master. The pres-
ervation of social order depended on harmonizing names and realities. On
such a scale of values, material considerations were essentially incidental.

Conclusion

Among the various subjects that this chapter has not dealt with, the most
telling is probably that of quantitative data. When exploring any form of
servitude, historians focus first on two essential preliminary aspects: legal
status and detailed figures. In the case of late Ming China, the main prob-
lem lies in the fact that both data remain incomplete. Ming law did not
give a precise definition of social status and this is the reason why the
legal status of bondservants is usually reduced to the discrimination they
suffered and to a bare series of legal incapacities. And this is probably
the reason why the famous historian Wu Han (1909–1969) could affirm in
1961 that, in the Ming a bondservant was private property with no more
270 claude chevaleyre

value than an ox or a horse, considering that the Code prescribed quite

the same punishment for sacrificing an ox and for killing a bondservant.90
Hopefully, numerous subsequent studies have demonstrated first that
one could not simply regard the seriousness of two crimes or offences as
identical only on the basis that they were liable to the same punishment,
and, second, that bondservants were not simply dehumanized production
tools or chattel slaves.
As for figures, it must be acknowledged that, in the case of Ming China,
data are rare, most figures being insufficient and unreliable. In a series of
conferences held at Collège de France, Timothy Brook has shown how
critical the lack of data on prices could be for historians of the Ming
period.91 But such a statement is also true when considering population
series, especially when considering the distribution of social groups. Ming
writers and administrators rarely noted figures. But the task of evaluat-
ing the number of bondservants would be even more difficult than eval-
uating any group of commoners because many cases of bond-servitude
were disguised and hidden behind marriages or adoptions, as mentioned
above; and also because many made themselves bondservants in order
to evade taxes, subsequently disappearing from population registers. So,
what figure shall we accept? How are we to interpret the fact that at least
one bondservant was involved in 24% of the cases published in the Yun-
jian yanlüe? Shall we rely on the recurring laments of late Ming writers
claiming that great clans and families in the lower Yangzi regions had
“hundreds,” “thousands,” and even “tens of thousands” of bondservants?
Such numbers are less concrete than rhetorical, and they don’t say a word
about regional variations.92 Let’s assume that Ping-ti Ho was right when
he ascertained that mean people amounted to approximately 1% of the
Ming population.93 This would induce a total of mean people contained
between 700,000 and 2 million around 1600, among which, probably, a
great majority were bondservants.94 This is a plausible figure, but how far

90 Wu Han, Dengxia ji, 76.

91  Brook, “Valeurs et prix dans la Chine des Ming.”
92 Most of the documents presented in this paper focus on the Jiangnan region, that
is to say the prefectures of the rich and developed Yangzi River delta. As a consequence,
I shall warn the reader against generalizations on China as a whole. If bondservants had
the same legal status all over the empire in the seventeenth century, practices may have
greatly varied from one part of the realm to the other.
93 Ho, The Ladder of Success, 19.
94 Depending on ways of calculating, population figures for 1600 range from 66 to 230
million. See Brook, The Troubled Empire, 42–45.
 acting as master and bondservant 271

is it from the reality? This remains a pending question, as must the ques-
tion of how long-term trends evolved.
Some scholars have advocated a model of general decrease in the use
of bondservants through the Ming.95 Yet, one cannot ignore the recurring
narratives of late Ming and early Qing writers who lamented the increas-
ing number of bondservants all throughout the Ming.96 In the same man-
ner, shall we accept that late Ming insurrections were such a social trauma
that they lead to the quasi disappearance of bondservants as soon as the
Kangxi era,97 or just after the Yongzheng reforms?98 Or shall we simply
trust the numerous imperial archives that remain to testify that crimes or
offences committed by or against bondservants kept judicial authorities
busy at least until the beginning of the nineteenth century?
What can be ascertained is that bond-servitude and trafficking in
human beings persisted long after the fall of the Ming and probably after
the establishment of the Republic of China, some 250 years later.99 The
present chapter thus attempted to demonstrate that China, this civilized,
Confucian and too often atemporal historical object, was also part of the
world history of human bondage. As such, it should less be regarded as
exotic than as another configuration worthy of attention to shed light on
the universal experience of human bondage. In many respects, most of its
characteristics, though encapsulated in so-called Confucian values and in
specifically Chinese practices, can be fruitfully compared with the char-
acteristics of bondage practices throughout the world. As in many places
in the world, voices were sometimes heard to denounce the evils of the

95 Elvin, The Pattern of the Chinese Past, 235. However, Elvin’s analysis focuses on the
progressive disappearance of the “manorial order with serfdom and a serf-like tenancy,”
in which bondservants were included and not totally differentiated from other forms of
“serfs” and “tenants.” For him, this structure “continued to dominate the countryside,
though with diminishing vigour as time passed by,” to finally disappear in the course of
the eighteenth century.
96 Xie Guozhen estimated that bond-servitude had quite disappeared after the Song
dynasty, before a new period of steady growth in the Ming. See Xie Guozhen, “Mingji
nubian kao,” 195.
97 See the quotation of Shengguo jiwen stating that “under Kangxi, great families didn’t
dare to raise bondservants anymore,” in Xie Guozhen, “Mingji nubian kao,” 211.
98 Ho, The Ladder of Success, 19.
99 If Chinese bond-servitude can in no way be compared with forms of servitude based
on the traffic in human beings, illegal human trade networks, however, existed. See, for
example, the early Qing case adjudicated by Yu Chenglong, who dismantled a trade in chil-
dren between Guangdong and Guangxi provinces. More than 200 boys and girls had been
sold for “entertainment” by a network involving some 80 accomplices. See Yu Chenglong,
Yu Chenglong pandu jinghua, 37–39. See also the studies by James Watson on slavery rem-
nants in Guangdong under the Republic; in Watson, “Transactions in People”, 223–250.
272 claude chevaleyre

servile institution and to call for more humane treatments and benevo-
lence. But, in fine, and despite limits imposed by the principle of reciproc-
ity and the requirement to make social status match realities, Confucian
values—as many other civilized or humanist doctrines—did well with the
servile institution.
Considering that time has come to reintegrate the Chinese experience
into the world history of bondage, focusing on the late Ming dynasty
seemed a good starting point as many of the characteristics of the Chi-
nese servile institution as it was denounced—both by Western observers
and by Chinese modernists and revolutionaries—at the end the imperial
era were shaped at that time. To decide whether the Chinese servile insti-
tution must be qualified slavery, bond-servitude, or whatever else would
seem appropriate is of secondary importance. What is important is that
the Chinese experience is no less informative than other experiences to
understand how universally shared institutions such as slavery operate.
 Public Works and the Question of Unfree Labour

Chitra Joshi

Studies on Indian labour till recently were framed within a teleological

logic which associated capitalist industrialization with notions of free-
dom. Within such frames, forms of coercion and unfreedom appeared as
signs of the incompleteness of processes of modernization and transfor-
mation. Discussions around bondage and slavery were important to writ-
ings on agricultural labour but the focus of labour history was primarily
on factory industries. Within modern industries state legislation was seen
as a move from punitive to protective measures: laws limiting working
hours, regulating wages legalizing trade unions marked this shift.1 State
intervention was thus seen as formalizing, legalizing the network of indus-
trial relations. The ways in which state legislation in fact sanctioned the
private coercive powers of capital over labour were not studied. Recent
writings on law and labour question earlier frameworks in two ways: One,
they push the origins of labour policy to the late eighteenth and early
nineteenth century;2 two, they move beyond the realm of the formal and
suggest how the state was deeply implicated in creating a network of
informal systems of control and coercion which lay outside the field of
law. Coercive, ‘unfree’ forms were actively created by the state and were
not residual or a product of the absence of laws.3
The first half of the nineteenth century was a period when the East India
Company was in the process of expansion and consolidation of its empire.
Mapping, exploring and building of road networks to allow a rapid move-
ment of troops was part of this process.4 Some of the major road proj-
ects, the Grand Trunk Road extending over fifteen hundred miles from
Calcutta to the North-West, the Agra-Bombay road extending over 700 miles,

1 On this shift see Anderson, “Work Construed”; Anderson, “India, 1858–1930”, in Hay
and Craven (eds.), Masters, 402–54.
2 Ahuja, “The Origins”.
3 Arguing this very forcefully, Prabhu Mohapatra points to various legislative measures
that gave additional private powers to employers for instance, Act. VI of 1865 gave addi-
tional penal powers to planters to discipline labour. Mohapatra, “Regulated Informality”.
4 On the importance given to road building by administrators like Bentinck and
Auckland see Singha, A Despotism, 230–32, 258–61; Sharma, Famine, 160–71.
274 chitra joshi

the Delhi—Allahabad, road (around six hundred miles) the Western Road
in Madras Presidency, the Deccan Road from Mirzapore to Jabalpur the
Hindustan-Tibet road, were undertaken in this period.5 Projects like the
Western Road employed on an average 1598 labourers per day, running
up a total of around 40,000 in some months; the Hindustan-Tibet road
between Dugshai and Simla employed on an average 1747 labourers per
mile (a total of 86000 labourers) and between Simla and Narkunda, 1965
workers per mile.6
Who worked on the roads? How were workers mobilized for work on
the roads? Labour on the roads embodied some of the contradictions of
modernity. Roads, in a sense, symbolized modernity. Yet they were con-
structed by forms of labour that could be not legitimated through the
rhetoric of modernity. How do we understand this?
Work on large public works had certain imperatives and these are impor-
tant to understanding the strategies of labour mobilization and discipline.
The construction of roads, canals and other public works required large
bodies of labour at particular sites for a fixed time period. The demand for
labour was a spatially shifting one, moving from one construction site to
another. The demand was temporally bounded: once a project was com-
plete the demand fell. This created problems. It was difficult to get an
adequate and stable supply when it was needed, and a spurt of demand in
one place could push up wages. So officials on public works had to organ-
ise a supply that could meet the concentrated demand of labour without
raising wage costs. Obtaining local labour in the different places that the
roads passed through was particularly difficult. During plowing, sowing
and harvest times field labourers worked on the land. They could not be
attracted to public works without offers of high wages.7 Moreover landed

5 Contrary to colonial official claims however it is clear from recent studies that road
works in the nineteenth century did not always entail the opening up of completely
new and uncharted routes but the re-making of old in accordance with the new trans-
port requirements. See for instance Ravi Ahuja’s important recent publication, Pathways,
119–51.
6 Report on the Western road by W.K. Wurder, 30 October 1839, Home Judicial Proced-
ings (Progs), 1 March 1841 (National Archives of India, hence NAI); Briggs, Superintendent,
Hindustan Tibet road to Elliot, Secretary, Foreign Department, Foreign Consultations,
7 March 1851, no. 196 (NAI).
7 In the 1850s, Public Works officials in Madras Presidency pointed out that except
in districts where traditional castes associated with earthwork or castes, considered ritu-
ally impure within the Brahamanical order—like Pallars or Pariahs for instance—were
abundant, it was impossible to get labour to work on the roads. To resolve this problem
they proposed the idea of forming a ‘Permanent Corps of coolies’ for work on roads. See
 public works and the question of unfree labour 275

elites were often apprehensive of roads that absorbed village land, and cut
through cultivated tracts. They were interested in smaller local roads that
connected the villages to the grain markets, not highways. Inter-regional
roads, in fact, undercut the authority of landed elites. In many regions
local lords resisted the projects of public works departments.8
Historically, in different conjunctures, several different forms of coer-
cive labour were employed to meet the demands of road building. Three
kinds of labour commonly employed in road works in the nineteenth cen-
tury were begar,9 convict and famine labour. If in the early decades of
the nineteenth century, convict labour was predominant in India, other
forms were more important later. Canal and road building projects in the
late nineteenth century recruited large numbers of workers from famine
stricken areas. Public works in many areas became synonymous with
famine relief.10 The construction of irrigation works and roads proceeded
most rapidly during times of famines. Starving peasants were paid famine
wages to build the public infrastructure that became one of the signs of
modern India.11 What explains the shift from one form to another?

Tribute Labour

On projects like the Hindustan Tibet road undertaken in the 1850s, the
demand for labour was much higher than roads in plain areas. At any
point of time there were from 1700 to 2000 workers on each mile of the
road. A large proportion of the labour employed on the Hindustan Tibet
road was categorized officially as ‘unpaid’ labour. A 132-mile stretch
of road on the route between Kalka and Chini employed around 1,164,644
workers in the year, of which nearly 27 per cent (312598) was ‘unpaid’. In
the official logic, labour classified as ‘unpaid’ was considered as part of
the tribute hill states were obliged to pay for the protection they received
from the colonial government. In the official rationale, tribute labour sup-
plied by Hill states was not like ‘begar’ or the customary form of unpaid

Capt W.H. Horsley, to Capt J.H. Bell, 29 September 1853, Selections from Government
Records, Madras, No. VI (Government Press: Madras 1855), 1–4.
8 See for instance R. Ahuja, “Opening up”.
9 Begar was a customary form of unpaid labour demanded by dominant groups in a
region.
10 In the famine of 1838, labour from famine affected areas was employed on road works
in many areas. See Sharma, Famine, 161–66.
11 See for instance, Henvey, A Narrative, 99–128; CER Girdlestone, Report, 47–50, 99.
276 chitra joshi

labour. The Hill states were expected to give revenue remission to those
who were working on the roads. In official perception the injustices of
this system were mitigated by its longer term benefits. Road work could
ultimately do away with what in official eyes was the worst form of the
system—the use of coolies for porterage. Roads would allow a freer move-
ment of animal and cart traffic.
In practice, tribute labour on mountain roads intensified the practice
of begar.12 While forms of ‘begar’ labour were customarily employed in
building village roads, the use of begar in public works was in many ways
distinct from its customary use. Work on public roads was located away
from the village. Often it meant working dangerously on treacherous ter-
rain and staying out in the open without shelter for days. Begar on distant
roads was doubly exacting because workers had no access to subsistence
resources from the village: they had to buy grain at high prices from the
market. Zamindars from Kulu protested against labouring on roads:
as our interests suffer by this order we presume humbly to approach your
lordship with this petition we are poor men, and our country is ill populated
we have roads of our own not less than one hundred and eleven kose long
to look after. . . . we are charged with the construction and maintenance in
thorough repair of this long line of road. This duty leaves us little or no
leisure. Ordinarily we live in separate families with no more than one adult
male to a family. Our agricultural occupation and our public duties on the
roads and for conveyances of baggage are got through with difficulty. If in
addition we are obliged to work on the Grand road our interests will suffer.
As a people we are remarkable for attachment to our own hills. From choice
we have never consented to leave them. The Government orders we obeyed
without a murmur. But now that your lordship has honoured our country
with your presence we account the present a favourable opportunity for
settling our grievance before your lordship’s consideration.13
Contrary to the directives of the Governor General the hill roads project
could not be executed on the basis of tribute labour alone.14 Yet the use

12 The use of the bethi labour in Orissa in the nineteenth century is somewhat similar.
See Ahuja, “Opening Up the Country”, 73–130.
13 Petition of Zamindars of Koolloo to Gov Gen of India, Public Works Department, Hill
Road Papers, Hindustan Tibet Road Papers, 6 June 1851, no. 29 (NAI).
14 Note for instance a minute of the Governor General in 1850, ‘I acquaint the Super­
intendent of Hill states that without further information I decline to give my sanction to
the measure he has proposed, for employing a large number of hired labour in the con-
struction of the new road now in progress from Kalka through the Hills. The application
of the Superintendent involves a wide departure from the conditions on which I gave my
assent to the commencement of this work, and entirely alters its direction.’ Foreign Dept
(Political), 13 September 1850 (NAI).
 public works and the question of unfree labour 277

of tribute labour from Hill states served two functions: it cut total costs
involved; it also tended to push down the market rate for labour. The
official rate of wages on the Tibet road was fixed at 2–2 1/2 as. per day.
However, there were areas where workers could not be mobilized even at
double the rate. On regions bordering on Ladakh for instance, road build-
ing authorities were anxious: ‘I doubt very much whether a single Lad-
hakhee could be got to work on the road for 4 as. a day.’15 Apart from the
question of costs, there was also the issue of moral legitimacy. The state
was complicit in sanctioning a system of forced labour within Hill states
without being directly implicated in deploying a system of forced labour.
The terms within which discourse on such forms of tribute labour was
framed is different as we shall see from discussions on convict labour.
Unlike convict labour which bore the marks of its unfreedom, force in
the case of tribute labour is masked: the coercive mechanisms through
which labour is deployed from the supplying region is not recognized. In
the case of both forms however, the idea of public good provides a basis
through they are legitimated within colonial discourse.

Convict Labour

Some of the important road projects undertaken in the 1830s and 1840s
covering vast stretches employed large bodies of convict labour. The
construction of public works in India in this period was supervised by
the Military Board and the practice of using coercive forms of labour for
works commissioned by military authorities was common.
Military engineers supervising road works had anywhere between 1000
to 3000 prisoners working under them. On the road between Delhi and
Allahabad (approximately 600 miles) around 4686 prisoners distributed
over 15 gangs, were employed.16 Usually convicts were employed on work

15  WM Hay, Superintendent Hill States to Secretary, Government of India, 1 Nov. 1854,
Hindustan Tibet Road Papers, Public Works Department, 19 January, 1855, no. 512–516
(NAI).
16 A gang consisted of some 300 prisoners who were sub-divided into groups of 20–30 to
do specific tasks. A hierarchy of guards: one burkandauze (an armed guard, usually with a
matchlock) to every three prisoners and four duffadars (originally a person in charge of a
small body of troops) over every 100 burkandauzes kept watch while the convicts worked.
A gang of 300 prisoners had over 90 burkandauzes guarding them. The heavy work of
digging and raising of embankments was done by large gangs of prisoners. For purposes
of surveillance large gangs were preferred. For the work of metalling and making bridges,
smaller groups had to be employed which made guarding more difficult. I.G. Drummond,
278 chitra joshi

which required large bodies of workers—like digging and raising embank-

ments. For metalling roads or making drains and bridges which required
smaller parties of workers, hired labour was preferred.17 In the Western
road in Madras Presidency, on the basis of a daily average of 1598 convicts,
the total number of convict days involved peaked to over 40000 in the
rainy month of July.18
Convicts doing earthwork had to lift a minimum of 50 cubic feet of
earth by midday. Usually it took much longer and even by ‘straining every
nerve’ it was 2–3 pm by the time they finished.19 Superintendents in a
rush to complete projects and corrupt guards made heavier demands on
convicts.20 Medical officers enquiring into the reports of mortality on the
roads drew direct connections between burden of work and mortality on
the roads.
An important feature characterizing the convict labour regime on the
roads was the elaborate system of surveillance and control. On the Delhi—
Allahabad road for instance, there were a total of 1338 guards watching
4686 prisoners. There were in all 15 gangs of prisoners working on this
road. Each gang had over 300 prisoners with around 90—burkandauzes
overlooking them.21 Gangs were further subdivided into smaller parties of

Superintendent, Delhi and Allahabad to Capt Gowan Officiating Secretary, Military Board
(MB), 21 March 1834, MB Proceedings 27 June 1834 (NAI).
17 ‘All that belongs to the marking out and metalling of roads, and the construction
of drains and bridges should be done by hired labour. It is necessary to employ small
parties for executing works of the description mentioned and it appears that prisoners
cannot with security be so unless by increasing the strength of guards at an expense that
would perhaps exceed the value of their work. Nicolson, to Debude, MB Proceedings,
25 November, 1836 (NAI).
18 Report on Western Road by W.K. Wurder, 30 Oct 1839, Home Judicial Proceedings,
1 March 1841 (NAI).
19 Drummond, Superintendent, Delhi and Allahabad to Capt Gowan Officiating Magis-
trate, 21 March 1834, MB Proceedings, 27 June, 1834 (NAI).
20 Burkandauzes tended to be partisan forcing more work on some and letting off oth-
ers. Reports like the following were common ‘. . . one of the convicts named Golamully
made a complaint to me stating that his basket was loaded with much more earth than
the rest wore, as he had not given the usual present.’ Jenkinson, to Thomson, 21 February
1833, Judicial Criminal Proceedings, 15 March, 1833 (NAI).
21 Drummond, Superintendent Delhi and Allahabad road to Gowan, Offg Secy Mili-
tary Board, 21 Mar, 1834, MB Progs, 27 June 1834. In selecting burkandauzes, caste was an
important issue. Prisoners from an upper caste background—Brahmins, Thakurs—were
not placed under the supervision of burkandauzes from a lower caste background. MB Pro-
ceedings, 11 November, 1834 (NAI). Burkandauze: armed guard usually having a musket.
 public works and the question of unfree labour 279

20–30 prisoners with assigned tasks to complete. In addition there were

around four duffadars and two jamadars22 over every 100 burkandauzes.23
Convicts had to work in fetters in the day and at night groups of around
twenty were bound together by a chain which ran through the centre
link of each individual leg chain. Although prisoners were supposed to be
relieved of their irons in case of extreme sickness, this was rarely done. A
doctor reporting on the state of convicts working in road gangs in India
in the 1830s was surprised to find: ‘. . . that the sick [convicts] even in the
last stages of disease are kept in fetters . . . in every instance even when
the prisoners were reduced to the appearance of a skeleton and unable
to move from side to side the irons had not been struck off . . .’24 On the
Burdwan road, prisoners attacked by cholera, remained chained all night.
A doctor visiting them was almost driven back by the insufferable stench
of excreta emanating from the tent.25
Surveillance and repressive structures of control are not peculiar to
the convict labour system. In plantations and mines, managers closely
regulated the body of the worker. The space of work in mines and plan-
tations was bounded. Coolies trying to escape from plantations were
often flogged sometimes to the point of death.26 The powers of planters
extended beyond the workspace to the ‘coolie lines’: they regulated and
controlled the private lives of workers. Mines were known for the severity
of measures against theft of mine property. Miners songs in the Kolar gold
mines in South India, recorded the extremes to which checking against
pilferage went: they lampooned mine authorities for forcing them to bend
over like cattle to for signs of hidden gold in their orifices.27
In contrast to mine or plantations workers who are kept under surveil-
lance within a circumscribed work space, convicts on road work have to

22 Duffadar: originally a military word for a person in charge with a small contingent of
troops. Jamadar: used usually for the leader of a group of individuals.
23 Thomson, Supdt roads to Goween, Officiating Secretary, MB, MB Progs, 11 November,
1834.
24 MB Progs, 2 August, 1836 (NAI).
25 Report of the Committee on Convict Labour, MB Progs, 2 August, 1836 (NAI).
26 See for instance, Rana P. Behal, and Prabhu P. Mohapatra, “ ‘Tea and Money versus
Human Life’: The Rise and Fall of the Indenture System in the Assam Tea Plantations
1840–1908”, Journal of Peasant Studies, 19 (1992): 142–72.
27 It was suggested that women should be made to jump up when they left the mine
premises, to check if gold was hidden in their private parts. Nair, Miners, 52. On systems of
surveillance and control in South African mines see: Worger, South Africa’s, 110–46; Crush,
“Power”.
280 chitra joshi

be on the move. The paradoxical character of the convict labour regime

lies in its creation of a labour force that it is on the one hand fixed and
immobile and on the other flexible and mobile: the location of work is a
continuously shifting one. This mobility enhances the need for fettering.
Unlike other coercive systems in which labour mobility is controlled and
regulated through the burden of debt or penal contracts, convicts workers
are physically immobilized.28

The Logic of a Fettered Regime

Officials in the Military Board engaged in the building of roads had to

continuously engage in a dialogue with critics of convict labour. What
were the terms of this debate? What accounts for the use of such a coer-
cive system of labour? Crucial to the discussion on convict labour in the
early nineteenth century India as in debates over free/unfree labour else-
where was the issue of costs.29 Did employing convict labour work out
cheaper than hired labour? What were the calculations involved? In offi-
cial discussions on the relative advantages of convict versus other forms
of labour, three issues are raised: the difference between average wage
rates for hired labour and costs of employing convicts; the costs of keep-
ing prisoners in jail and on the roads; the productive efficiency of convicts
and hired labour.
Military engineers executing vast road-building projects were insistent
that prisoners could provide labour which was in the long run more cost
efficient than hired labour. They responded sharply to critics of the sys-
tem, questioning the basis of their evidence.
In regions like Bengal, in India, lines were sharply drawn between
the military and civil authorities over the question of employing convict
labour. Here the high mortality among convicts on road work in the 1830s,
led to mounting attacks on the system.30 Judicial officials gave evidence
to prove that it was in the long run more expensive to employ convicts

28 On the question of mobility and ‘forced’ labour see also, van der Linden, “Forced
Labour” in Brass and van der Linden (eds) Free and Unfree Labour, 351–62.
29 On the debate on this in the transatlantic context, see Drescher, “Free Labor” in
Engerman ed. Terms 50–86.
30 In Bengal Presidency the average mortality among prisoners working away from the
prison, under executive officers, was around 7.3 per cent in the period between 1833–36, as
against 3.38 per cent in North-West Provinces. Report of the Prison Discipline Committee,
1838, Home Legislative Progs, September, 1838, no. 47–48, (NAI).
 public works and the question of unfree labour 281

on road work. According to estimates drawn up by the Prison Discipline

Committee set up in 1838, the extra costs involved in employing convicts
to do road work was around Rs.2 a month, but the same work could be
done by contract labour at 2/3rds and in some places at 1/2 the price.31
Besides, convicts they argued worked fewer days a month: on an aver-
age, around 21 days, and this had to be taken into account in calculating
costs.32
Military engineers were not convinced by these arguments. Irvine, a
Superintending Engineer in Central Provinces, asserted that the estimates
of the Prison Discipline Committee had no general validity, derived as
they were from limited, unreliable and questionable evidence. They were
in fact based only on the experience of Captain Murray in one single area
in Bengal, in Noakhali. Ridiculing the estimates drawn by Murray, Irvine
wrote:
Captain Murray may probably have got a ditch round his compound at the
rate of 13 as. per 1000 cubic feet, but this is very different kind of work from
making a road 30 ft wide . . . In no part of Bengal could work of this kind
done by hired labour on contract at the rate of 13 as per 1000 cubic ft. Unfor-
tunately Capt Murray has done no road work by hired labour so that I am
unable to bring forth any of his Bills to show what 1000 cubic ft of road
would really cost at Noacally.33
He went to absurd lengths to argue that fetters did not really restrict
the convict’s ability to dig earth or move as fast on level ground as free
labourers.
In regions like Bombay, those arguing for convict labour system claimed
that absolute costs of hired labour in the Province were higher than other
regions. In Salsette, in Bombay Presidency for instance, military officials
responded to the objections of civil authorities arguing: ‘. . . labour here . . .
is so much dearer than in the Upper Provinces we cannot here hire labour
under Rs 5 a month’.34 Military engineers in Bombay emphasized on the
specific local situation to justify the need for employing convicts at a time
when the Bengal Government was trying to legislate against their use.

31  Prison Discipline Committee Report, 57.

32 Convicts did not work on Sundays and other local holidays.
33 Irvine to Military Board, MB Progs, 8 December, 1840 (NAI).
34 R Foster to IP Willoughby, Secretary to Government, Bombay, 9 September, 1839,
Home Judicial Progs. 13 January, 1840 (NAI), Irvine, Superintendent Engineer, to Military
Board, MB Progs, 8 December, 1840 (NAI).
282 chitra joshi

Others asserted the advantages of putting convicts on road work by

comparing the relative costs of keeping them in jail. Estimates drawn up
by the Revenue Department in Madras for instance showed that employing
convicts on road work cost the state around 13 per cent more than hired
labour. This relative disadvantage, Madras officials argued, was more than
compensated by the savings in prison expenses. Keeping them in prison
was more than three times the cost of employing them on roadwork.35
The question of costs was also linked up with the issues of labour
discipline and efficiency. Critics of the convict labour system tended to
argue that ‘free’/ hired labour was characterized by a motivation to labour,
absent in convict labour.36
Morever, the wearing of fetters, they argued, restricted the mobility
and productivity of convict labour. Advocates of the system countered
assertions about the productive efficiency of ‘free’ labour by pointing to
the problems of managing and disciplining hired workers. Hired labour
had to be mobilised through the contract system and this always had its
risks. The majority within the Military Board were opposed to giving the
contract system a trial. Local contractors, they were convinced, used poor
quality material and did not fulfill terms of contract. Although regulations
drafted in 1814 and 1816 gave employers the right to file complaints against
workers who left before completing their contract, these were considered
inadequate.37 Employers continued to demand greater safeguards against
absconding workers.38
Convict workers could be employed directly under executive officers
without the mediation of contractors. Military officials like Vetch, involved
in the construction of the Burdwan–Benares road saw this as a chief advan-
tage: ‘To Executive officers labour by convicts afford much satisfaction.

35 Minute on Employment of Convict Labour, by Chief Secretary, 22 February, 1840.

Home Judicial Progs, 1 March, 1841 (NAI).
36 This assumption was important in some of the exchanges between defendants of
slavery and abolitionists in the late eighteenth and early nineteenth century. In reality
abolitionists found it difficult to sustain this argument, especially when sugar production
in the Caribbean collapsed in the aftermath of abolition. However these assumptions con-
tinued to underline debates over convict labour in India. On this debate see Drescher,
“Free Labor vs Slave Labor”, 50–86, S. Engerman, “Slavery and Emancipation in Compara-
tive Perspective: A Look at Some Recent Debates” Journal of Economic History, 46 (1986):
317–39.
37 The law of 1814 gave powers to Justices of Peace to issue warrants and conviction
notices against workers who left work in violation of terms of contract. The regulation of
1816 empowered the Justices of Peace to take action against workers attempting to organ-
ise in defense of higher wages. Legislative Dept Progs, 12 March, 1852, no. 47 (NAI).
38 Legislative Dept Progs, 12 March, 1852, no. 47 (NAI).
 public works and the question of unfree labour 283

Extension work by daily labourers requires a system of native contractors

who are in general an endless source of vexation to superintendents.’39
Moreover, convicts could be mobilized to work on distant jungle roads
over long stretches of time necessary for road work, and in places where
hired labour was difficult to find.40
However, if we closely examine the context in which it was used, it
is quite clear that costs alone do not provide a sufficient explanation.
In India, South Africa and in other countries the convict labour system
points to the connected histories of labour and punishment. The regime
of Bentinck the Governor – India in India, associated with grand road
building schemes in the 1830s, coincided with a period when utilitarian
ideas of legal reform were being deliberated. Imprisonment and new
modes of prison discipline were part of a programme of penal reforms
in this period. In the absence of a prison infra-structure and buildings,
local magistrates saw outdoor work for prisoners as a short term solu-
tion to the problem of prison space.41 A spurt in the demand for convict
labour coincided with a rise in the number of convictions in the 1830s
and an absence of prison infra-structure.42 But extra-mural work became
a contentious issue for advocates of penal reform. When the Prison Dis-
cipline Committee was appointed in 1836, there was a growing body of
opinion against extra-mural work. The Committee members saw outdoor
work particularly on roads at a distance from jails as a threat to a system
of prison discipline.43
To judicial officials, a continuation of the convict labour system was
dependent on its deterrent effect. Work in fetters on the roads was mean-
ingful if it shamed the criminal. But did it? Officials differed, turning the
arguments continuously around, but the theme of shame recurred in dif-
ferent ways. For some, fettered work on the roads meant a mode of public

39 Capt. G.A. Vetch, Superintendent, Burdwan roads to E.P. Gowan Acting Secy, Mili-
tary Board, MB Progs, 27 June, 1834 (NAI).
40 Lt Col Craigie, Military Board to Auckland, Governor, Bengal, 12 Aug., 1836, MB Progs,
12 Aug., 1836 (NAI).
41 Before the Law Commission put forward its proposals for penal reform, judicial offi-
cials were instructed by the Governor General to send prisoners with sentences exceeding
three months to labour on public works. Macsween to Hunter, Judicial Criminal Progs,
30 Dec. 1833 (West Bengal State Archives); See also David Arnold, ‘The Colonial Prison’,
Subaltern Studies VII (New Delhi: Oxford University Press, 1994), 148–87.
42 Prison Discipline Committee Report, 9; F.J. Mouat, Journal of The Statistical Society
of London, 25 (1862): 183.
43 Prison-Discipline Committee Report, 61. On the changing ideology of punishment
see, Garland, Punishment, 111–55.
284 chitra joshi

shaming—the fettered body was not hidden away within a prison; it was
visible, made the object of public gaze and censure. Upper caste convicts,
in fact, were doubly shamed: first being exposed in irons and then being
seen working with other ‘lowly’ folk. Other officials saw no straightfor-
ward connection between caste and notions of stigma. ‘Natives’, they felt,
were devoid of any sense of shame.44 Besides, reports on prisons like the
one in 1838, point to a preference for muscular, outdoor over indoor work
among male prisoners. Spinning and grinding flour were seen as feminine
activities and therefore more demeaning.45 This was seen as undermin-
ing the deterrent value of extra-mural work. Once outside, masculinity
was exercised and displayed in a variety of ways. At the worksite, it was
often the weaker ones who were forced to do the strenuous tasks, while
the stronger and macho bodies could bully supervisors and get away with
less. Within a seemingly contradictory logic, masculine attributes were
reaffirmed by demonstrating the ability to do the least muscular activity.
The judicial critique of convict labour in India developed in the context
of a larger conflict between civil and military authorities over questions
of governance. The use of openly coercive methods conflicted with lib-
eral ideas of rule and notions of legitimacy important to the colonial state
in the nineteenth century. These tensions between the civil and military
authorities recur in discussions within the Committee on Convict Labour.
In the scathing indictment of the system by the Committee it is the lib-
eral, reformist discourse which has an edge. The voice of military officials
arguing for an intensification of control and surveillance is increasingly
marginalized.
By the 1850s the legitimacy of the system was corroding. The influence
of liberal and utilitarian, arguments against the use of such forms of labour
became more important. This was also a period marked by an expansion
of the prison establishment, and a questioning of the idea of road work
as a form of punishment among the prison authorities. As scarcities and
famines recurred with untiring regularity from the 1870s, the supply of

44 The appendix to the Report of the Prison Discipline Committee includes written
statements from different regions on the question of caste and punishment. Some offi-
cials were convinced that labour in irons was not considered ‘disgraceful’ by local people.
See for instance evidence of the Magistrate of Farukhabad, Prison Discipline Committee
Report, 223.
45 The Magistrate of Humeerpore noted: I do not think that labor on the roads in irons
is felt as a greater disgrace than laboring in irons within the Jail, at spinning thread or
grinding atta- perhaps not so much, for these occupations are looked down upon as fit
only for women.’ Prison Discipline Committee Report, 224.
 public works and the question of unfree labour 285

cheap famine labour also expanded. Yet there remained a current of opin-
ion within judicial discourse which favoured the employment of convicts
on public works.46
By the second half of the nineteenth century the idea of public good
that gave a moral sanction to the convict labour system was itself under
question. Supporters of the convict labour system operated with two sets
of assumptions: one, physical labour had a redemptive potential: it puri-
fies the body and redeems the individual; two, labour on public works is
doubly virtuous. It has a symbolic reciprocal function of giving back to the
public; of paying for the economic and moral costs of criminality. By the
late nineteenth century reformers were arguing against such assumptions.

Public Works and Coerced Labour

The colonial state was involved in distinct and different ways in the pro-
duction and reproduction of different forms of coercive labour. In tea
plantations in India state legislation formalized the penal powers of plant-
ers; it also sanctioned a wide range of ‘informal’ private powers of planters
over labourers. The wider system of debt peonage was made possible by
colonial legislations on contract. But state intervention in the working of
these systems was indirect: it created the legal framework which allowed
the allowed them to operate. In employing systems of tribute and convict
labour, the state was more directly implicated in the working of a repres-
sive regime of labour.47 State strategies of mobilizing tribute labour pro-
vided a basis for the intensification of begar in Hill states.
The specific historical context in which the state came to play a more
interventionist role in regulating and disciplining labour is important.
State involvement in recruitment and disciplining strategies in the late
eighteenth and nineteenth century India took place against recurring
problems of creating and retaining a pliant labour force. In eighteenth cen-
tury Madras for instance, militarist forms of control over labour emerged

46 The Report of the Prison Conference of 1877 was firmly of the opinion that the sys-
tem of employment on public works was a valuable adjunct to the system of jail admin-
istration: ‘It affords a means by which over-crowding may be relieved and space made
available for the construction of separate sleeping accommodation for the main body of
prisoners. It is also an important extension of our available means for providing labour at
once penal and profitable.’ Home Judicial Progs, October 1882, nos. 121–152 (NAI).
47 Behal and Mohapatra, “Tea and money”; Mohapatra, “Assam and the West Indies” in
Hay and Craven (eds.) Magistrates, 455–80.
286 chitra joshi

against a background of East India Company wars against regional powers

and its struggles over scarce local labour resources. Ravi Ahuja elaborates
how this situation was complicated by ritualized modes of collective with-
drawal and desertion by labour prevalent in the region.48 Coercive labour
strategies could act as a regulating mechanism, putting a downward pres-
sure on the price of hired labour in the market. State intervention thus
shows the processes through which the labour market was segmented. It
was not just limits to labour mobility which created segmented pockets,
but state regulation created a segmentation of the market.
As with other coercive forms the use of convict labour allowed employ-
ers to regulate the labour market by pushing down prices. The very pres-
ence of convicts on construction sites, officials pointed out, could keep
a check on local wage rates.49 In the late nineteenth century when the
practice of extra-mural work for convicts was practically discontinued,
many officials still argued for convicts on public works. The Sirhind Canal
in Punjab in the 1880s was built largely with convict labour. The use of
convict labour ensured a steady supply of labour which was not sub-
ject to harvest time fluctuations. It could be secured in the harshest of
conditions; in times when hired labour was impossible to get. Note for
instance the satisfaction with which an official employed on the Sirhind
canal project reported: ‘On cold wet mornings it is practically impossible
to collect a large body of free labourers at short notice . . . whereas a single
message is sufficient to bring a jail full of prisoners at a moment of need
on to an endangered bund’.50 On dangerous and difficult stretches of the
Hindustan-Tibet road, convict labour was requisitioned.51 Tunnel building
in terrain where blasting was not possible required steady and concen-
trated labour for which convicts were preferred.

48 Ahuja, “Labour Unsettled”; Ahuja, “Labour Relations”; Lucassen, “The Brickmakers’

Strikes”.
49 Note for instance a statement by Inspector General of Prisons, Lahore: ‘this value
[of convict labour]is due, not only to the amount of work done by them, but also to the fact
that their presence tends to prevent the rates of local labour from rising excessively, and
that their numbers do not fluctuate with sowing and harvesting seasons.’ W.P. Dickson,
Report on the Extra Mural Employment of Prisoners on the Sirhind Canal at Rupar, Punjab
between 1869–81 (Lahore 1884), 11.
50 Sirhind Canal Report, p. 11.
51 In making a tunnel through hard rocky terrain, north of Simla, 10000 convicts were
employed. Report on Hindustan Tibet road, 1 Oct. 1852, Foreign Department (Foreign Con-
sultations), 22 December, 1852, no. 284 (NAI).
 public works and the question of unfree labour 287

The history of labour on public works brings out the close association
between coercive labour regimes and public works. Large bodies of labour
were required across vast spaces and over a sustained period of time.
In the official logic, convict labour, begar and other forms of impressed
labour could provide a compliant labour force which could do arduous
tasks at lower than market prices. Yet there were limits within which such
systems worked. It was not possible for Public Works authorities to draw
entirely on tribute labour. There were regions which refused to comply
with state demands. On difficult stretches of road other coercive labour
strategies were deployed. The convict labour regime was one of the ways
in which capital and the state try to deal with the problem of worker
‘insubordination’ but like all coercive forms it had a troubled existence.
Problems of escape and mortality threatened its basis.52 We need then to
look beyond binaries of freedom and unfreedom and unravel the tangled
pasts that lie behind these categories. The margins of closed and ‘unfree’
regimes were continuously pushed and remade, and it was in these mar-
gins that possibilities of transcendence and breakdown also lay.
The history of labour on public works in India brings out the tenu-
ous links between modernity and freedom. Roads- symbolic of unfettered
mobility—hide behind them the marks of fettering and unfreedom. I do
not wish to end with the suggestion that these coercive forms were pecu-
liar to colonial modernity. The specificities of colonial forms are often
developed through unproblematic contrasts with idealized models of
capitalism and modernity. The colonial appears in these constructs as an
imperfect articulation, or deformation of the pure form which unfolds in
the west.53
Recent writings suggest that even in the west the growth of the mar-
ket and capitalism was not synonymous with ideas of freedom. Steinfeld
and others powerfully demonstrate how relations between employers
and workers in the ‘modern’ factory continued to operate outside formal
languages of contract.54 Employers continued to use penal sanctions and
various coercive forms to subordinate labour into the late nineteenth cen-
tury. Regulations, often, formalised these coercive relations. We need to
recognize that coercion and violence is integral to the history of capital-
ism, and that it is inseparable from the history of modernity.

52 For more on this see “Fettered Bodies”.

53 See for instance, Chakrabarty, Rethinking.
54 Steinfeld, Coercion; Reddy, Money.
 References

Archives

GARF (State archives of the Russian Federation), fond 109 opis’ 3, delo 1885.

RGADA (Russian State archives of Ancient Acts).

fond 291 (glavnyi magistrat), opis’ 1, chast’ 4; opis’ 2.
fond 294 (manufaktor-kontora) opis’ 1, dela 385, 386, 567; opis’ 2, dela 135, 145, 168.
fond 342, Novoulozhennye komissiii, opis’ 1, dela 37, 64, 119, 120.
fond 396, opis’ 2, ch. 5 (1616–1732), cadaster Smolensk, and several other districts.
fond 615 (krepostnye knigi mestnyjh uchrezhdenii XVI–XVIII v), opis’ 1.
fond 1209 (prilozhenie arkhiv premikikh votchnikh del, 1565–1692), opis’ 1, ch. 1–3,
opis’ 2, ch. 1–2, opis’ 16–72.
fond 1239, Cadastres documents (Pistsovye knigi), opis’ 3, chast’ 17, 69–72, 74, 76, 86–87
(cadastres of Moscow district, 1674–1681).
fond 1273, opis’ 1.
fond 1287, Sheremetevy, opis’ 3 chast’ 1, delo 582; chast’ 2, dela 1283, 1298, 1391.

RGIA (Russian State Imperial Archives)

fond 1149 (department zakonov gosudarstvennogo soveta) opis’ 1, dela 64, 119, 120;
opis’ 2, dela 20, 44; opis’ 3, delo 125.
fond 1287, opis’ 39, dela 65, 72.

NA (Národní archiv Praha) (National archives of the Czech Republic, Prague) fond SM R
109/45

SOA Děčín (Státní oblastní archiv Litoměřice, pobočka Děčín) (State district archives
Litoměřice, depository Děčín, Czech Republic) fond Rodinný archiv Clam-Gallasů, His-
torická sbírka (HS)

TSGIAM (Central State Moscovite Archives)

fond 16 (Chancellery of the General-Governor), opis’ 1, dela 1293, 1328, 1352, 1801. Opis’ 2,
dela 192, 1618. 227, 938, 948, 976, 980, 986, 1573, 1571, 1588, 1589.
fond 32 (Moskovskoi gorodskoi Magistrat), op. 1 (1842–1866), dela 5, 13, 46, 118, 153, 284;
opis’ 2 (1823–1844) dela 31, 40, 82, 124, 146.
fond 54 (Moskovskoe gubernskoe upravlenie), 1783–1917, opis’ 1, dela 56, 284, 966, 1509,
1618.

Bibliography

Adamczyk, Dariusz. “Zwischen (Welt)Markt und endogenen Zwängen: Die Entwicklung

der Gutswirtschaft im Polen der frühen Neuzeit.” Zeitschrift für Weltgeschichte 7, 1
(2006): 77–86.
——. Zur Stellung Polens im modernen Weltsystem der Frühen Neuzeit. Hamburg: Kovac,
2001.
Ahmad, Abdussamad H. “Ethiopian Slave Exports at Matamma, Massawa and Tajura,
c. 1830 to 1885”. In Clarence-Smith (ed.), Economics of the Indian Ocean Slave Trade:
93–102.
290 references

Ahuja, Ravi. “ ‘Opening up the Country’? Patterns of Circulation and Politics of Communi-
cation in Early Colonial Orissa”. Studies in History 20 (2004): 73–130.
——. “Labour Relations in an Early Colonial Context: Madras, c. 1750–1800”. Modern Asian
Studies 36 (2002): 793–826.
——. “Labour Unsettled: Mobility and Protest in the Madras Region, 1750–1800”. The
Indian Economic and Social History Review 35 (1998): 381–404.
——. “The Origins of Colonial Labour Policy in Late Eighteenth-Century Madras”. Interna-
tional Review of Social History 44 (1999): 159–95.
——. Pathways of Empire: Circulation, ‘Public Works’ and Social Space in Colonial Orissa,
c. 1780–1914. New Delhi: Orient Blackswan, 2009.
Aitken, E.H. (comp.), Gazetteer of the Province of Sind. Karachi: Mercantile Steam Press,
1907.
Allen, Richard Blair. “The Constant Demand of the French: The Mascarene Slave Trade
and the Worlds of the Indian Ocean and Atlantic during the Eighteenth and Nineteenth
Centuries.” Journal of African History 49 (2008): 43–72.
——. “The Mascarene Slave-Trade and Labour Migration in the Indian Ocean during the
Eighteenth and Nineteenth Centuries.” In Campbell. Ed. Structure of Slavery: 33–50.
——. Slaves, Freedmen, and Indentured Laborers in Colonial Mauritius. Cambridge: Cam-
bridge University Press, 1999.
Allen, Robert. “Tracking the Agricultural Revolution in England”. The Economic History
Review, 52,2 (1999): 209–235.
——. Enclosures and the Yeoman. The Agricultural Development of the South Midlands,
1450–1850. Oxford: Clarendon Press, 1992.
Alpern Engel, Barbara. Women in Russia 1700–2000. Cambridge: Cambridge University
press, 2004.
Alpers, Edward A. “Flight to Freedom: Escape from Slavery among Bonded Africans in
the Alpers, Edward A. Indian Ocean World, c. 1750–1962.” In Campbell. Ed. Structure of
Slavery: 51–68.
——. Ivory and Slaves in East Central Africa: Changing Patterns of International Trade to the
Later 19th Century. Berkeley: University of California Press, 1975.
Anderson, Clare “The Bel Ombre Rebellion: Indian Convicts in Mauritius, 1815–53.” In
Campbell. Ed. Abolition and its Aftermath: 50–65.
——. Legible Bodies: Race, Criminality and Colonialism in South Asia. Oxford: Berg Publishers,
2004.
Anderson, Michael. “India, 1858–1930: The Illusion of Free Labor”. In Masters, Servants, and
Magistrates in Britain and the Empire, edited by Douglas Hay and Paul Craven, 402–54.
Chapell Hill: University of North Carolina Press, 2004.
——. “Work Construed: Ideological Origins of Labour Law in British India to 1918”. In Dalit
Movements and the Meanings of Labour in India, edited by Peter Robb, 87–150. Delhi:
Oxford University Press, 1993.
Anonymous “Gold in Madagascar.” Anti-Slavery Report (March and April 1890): 69–70.
——. Enquête sur les conseils de prud’hommes et les livrets d’ouvriers. Paris: Impr. Impérial,
1869.
——. Pétition pour le rétablissement d’une mercuriale adressée au conseil des prud’hommes
par les chefs d’atelier et ouvriers de la Fabrique. Lyon: imprimerie Boitel, 1840.
——. Recueil des usages du conseil de prud’hommes de la ville de Lyon pour les industries
de la soierie, de la passementerie, des tulles, de la teinture, de la confection et des cuirs et
peaux. Lyon: imprimerie A. Bonnaviat, 1891.
——. Usages du conseil des prud’hommes de Lyon pour les industries de la soierie. Lyon:
imprimerie A. Bonnaviat, 1872.
Anwar, Ghazar. “Attitudes towards Homosexuality.” In Bonnie Zimmerman. Ed. Encyclo-
pedia of Lesbian and Gay Histories and Cultures vol. 1. New York: Garland Publishing,
2000: 403–5.
 references 291

Ariès, Philippe. Centuries of Childhood: A Social History of Family Life. Trans. Robert
Baldick. New York: Vintage, 1962.
Arkhiv gosudarstvennogo soveta (archives of the state council), 5 vv. (Saint Petersburg
1869–1904): 4, section 1, ch. 2, 253–258.
Arnold, David. “The Colonial Prison”. In Subaltern Studies VII edited by David Arnold and
David Hardiman, 148–87. New Delhi: Oxford University Press, 1994.
Aston, Timothy (ed.). Crisis in Europe, 1560–1660. New York: Doubleday, 1967.
Aston, Trevor, and Philpin, Charles (eds.). The Brenner Debate: Agrarian Class Structure
and Economic Development in Pre-industrial Europe. Cambridge: Cambridge University
Press 1985.
Atleson, James. Values and Assumptions in American Labor Law. Amherst, MA: University
of Massachusetts Press, 1983.
Aubin, Gustav. Zur Geschichte des gutsherrlich-bäuerlichen Verhältnisses in Ostpreußen von
der Gründung des Ordensstaates bis zur Steinschen Reform. Leipzig: Duncker & Humblot,
1910.
Austen, Ralph A. “The 19th Century Islamic Slave Trade from East Africa (Swahili and Red
Sea Coasts): A Tentative Census.” In William Gervase Clarence-Smith. Ed. The Econom-
ics of the Indian Ocean Slave Trade in the Nineteenth Century. London: Frank Cass, 1989:
21–44.
Baak, Paul E. “About Enslaved Ex-Slaves, Uncaptured Contract Coolies and Unfreed Freed-
men: Some Notes about ‘Free’ and ‘Unfree’ Labour in the Context of Plantation Devel-
opment in Southwest India, Early Sixteenth Century-Mid 1990s,” Modern Asian Studies
33,1 (1999): 121–57.
Balfour, Edward. The Cyclopædia of India and of Eastern and Southern Asia: Commercial,
Industrial and Scientific, Products of the Mineral, Vegetable, and Animal Kingdoms, Useful
Arts and Manufactures vol. 3. London: Bernard Quaritch, 1885.
Banaji, Jairus. Theory as History. Essays on Modes of Production and Exploitation Historical
Materialism Series. Leiden: Brill, 2010.
Banton, Mandy. “The Colonial Office, 1820–1955: Constantly the Subject of Small Struggles”.
In Hay and Craven (eds.). Masters, Servants and Magistrates: 251–302.
Barraclough, Geoffrey. Ed. Times Atlas of World History. London: Times Books, 1978.
Bartlett, Roger. “Serfdom and State Power in Imperial Russia.” European History Quaterly,
33, 1 (2003): 29–64.
Batou, Jean. “Attitudes of State and Society towards Industrialization in the Nineteenth-
Century Third World: The Cases of Egypt and Paraguay.” Paper presented to the Third
World Economic History and Development Conference (University Of Manchester,
13–15 September 1991).
Baumgarten, Karl, and Bentzien, Ulrich. Hof und Wirtschaft der Ribnitzer Bauern. Edition
und Kommentar des Kloster-Inventariums von 1620. Berlin: Akademie, 1963.
Becker, Samuel. Nobility and Privilege in Late Imperial Russia. Dekalb: Northern Illinois
University Press, 1985.
Behal, Rana P. and Prabhu P. Mohapatra. “ ‘Tea and Money versus Human Life’: The Rise
and fall of the Indenture System in the Assam Tea Plantations 1840–1908”. Journal of
Peasant Studies 19 (1992): 142–72.
Benoît, Joseph. Confession d’un prolétaire. Paris: Editions sociales, 1968.
Benton, Laureen. Law in Colonial Context. Cambridge: Cambridge University press, 2001.
Bentzien, Ulrich. “Fortschritte und Fortschrittsträger der deutschen Landwirtschaft im
Spätfeudalismus.” Jahrbuch für Volkskunde und Kulturgeschichte 21, NF 6 (1978), 125–60.
Berends, R., A.H. Huussen jr., R. Mens, and R. de Windt. Arbeid ter Disciplinering en Bestraf-
fing. Veenhuizen als onvrije kolonie van de maatschappij van weldadigheid 1823–1859.
Zutphen: De Walburg Pers, 1984.
Berg, Maxine. The Age of Manufactures: Industry, Innovation and Work in Britain, 1700–1820.
London: Fontana, 1985.
292 references

Bezon, Jean. Dictionnaire général des tissus anciens et modernes. Ouvrage où sont indiquées
et classées toutes les espèces de tissus connues jusqu’à ce jour soit en France, soit à l’étran-
ger, notamment dans l’Inde, la Chine, etc., etc. avec l’explication abrégée des moyens de
fabrication et l’entente des matières, nature et apprêt, applicables à chaque tissu en parti-
culier, t. 1. Vol. 1. Lyon: Imprimerie T. Lepagnez, 1859.
Bezucha, Robert J. “Aspects du conflit des classes à Lyon, 1831–1834.” Le Mouvement social,
n°. 76 (September 1971 ): 5–26.
Biernacki, Richard. The Fabrication of Labor. Germany and Britain, 1640–1914. Berkeley: Uni-
versity of California Press, 1995.
Blanc, Louis. Histoire de dix ans: 1830–1840. Paris: Pagnerre, 1843.
Blanchard, Ian. Russia’s Age of Silver: Precious Metal Production and Economic Growth in the
eighteenth century. London: Routledge, 1989.
Bloch, Marc. “Serf de la glèbe. Histoire d’une expression toute faite.” Revue historique, 36
(1921): 220–242.
Blue, A.D. “Chinese Emigration and the Deck Passenger Trade.” Journal of the Hong Kong
Branch of the Royal Asiatic Society 10 (1970): 79–93.
Blum, Jerome. “Prices in Russia in the Sixteenth Century.” The Journal of Economic History,
16,2 (1956): 182–199.
——. “The Rise of Serfdom in Eastern Europe.” American Historical Review 62,4 (1957):
807–36.
——. Lords and Peasants in Russia from the Ninth through the Nineteenth Century. New
York: Sharpe, 1964.
Blussé, Johan Leonard. Strange company: Chinese settlers, mestizo women and the Dutch in
VOC Batavia. Dordrecht [etc.]: Foris, 1986.
Boelcke, Willi. Bauer und Gutsherr in der Oberlausitz. Ein Beitrag zu Wirtschafts-, Sozial-
und Rechtsgeschichte der ostelbischen Gutsherrschaft. Bautzen: Domowina, 1957.
Bois, Guy. La crise du féodalisme. Paris: Presses de Sciences-Po, 1980.
Bok Rae, Kim, “Nobi: A Korean System of Slavery” in Campbell. Ed. Structure of Slavery,
155–68.
Bonnassie, Pierre. From Slavery to Feudalism. Cambridge: Cambridge University Press 1991.
Boomgaard, Peter. “Human Capital, Slavery and Low Rates of Economic and Population
Growth in Indonesia, 1600–1910.” In Campbell. Ed. Structure of Slavery: 83–96.
Bosma, Ulbe. “European colonial soldiers in the Nineteenth-Century. Their role in white
global migration and settlerdom.” Journal of Global History 4,2 (July 2009): 317–36.
——. “The Discourse on Free Labour and the Forced Cultivation System: The Contra-
dictory Consequences of the Abolition of Slave Trade for Colonial Java 1811–1863.” In
Humanitarian Intervention and Changing Labor Relations. The Long-term Consequences
of the Abolition of the Slave Trade, edited by Marcel van der Linden, 385–416. Leiden:
Brill, 2011.
——. Karel Zaalberg. Journalist en strijder voor de Indo. Leiden: KITLV Press, 1997.
Bosma. Ulbe and Remco Raben. Being “Dutch” in the Indies. A History of Creolisation and
Empire, 1500–1920. Singapore and Athens: Singapore University Press and Ohio Univer-
sity Press, 2008.
Bourgin, Georges, et Bourgin, Hubert. Le Régime de l’Industrie en France de 1814 à 1830: Les
patrons, les ouvriers et l’Etat. Paris: A. Picard, 1912.
Bourgon, Jérôme. “Uncivil Dialogue: Law and Custom did not merge into Civil Law in the
Qing.” Late Imperial China 23.1 (2002): 50–90.
Bowring, John. “Historical Forms of Bondage in Siam” (1855). In Lasker, Bruno, Human
Bondage in Southeast Asia. Chapel Hill: University of North Carolina Press, 1950: 283–8.
Brain, J.B. “Indentured Indians who returned to India from Natal”—http://ancestry24.com/
natal-indians/ (15/11/10).
Brass, Tom, and van der Linden, Marcel. Free and Unfree Labour. The Debate Continues.
Bern: Peter Lang, 1997.
 references 293

Braudel, Fernand. Civilisation matérielle, économie et capitalisme. Paris: Armand Colin,

3 vol., 1979.
Brazell, Karen (trans). The Confessions of Lady Nijo. Stanford: Stanford University Press,
1973.
Breman, Jan. Koelies, planters en koloniale politiek. Het arbeidsregime op de grootlandbouw-
ondernemingen aan Sumatra ̕ s Oostkust in het begin van de twintigste eeuw. Dordrecht
[etc.]: Foris, 1987. Published in Englishs as Taming the coolie beast: plantation society and
colonial order in Southeast Asia. New York [etc.]: Oxford University Press, 1989.
——. Koloniaal profijt van onvrije arbeid. Het Preangerstelsel van gedwongen koffieteelt op
Java. Amsterdam: Amsterdam University Press, 2010.
——. Labour Bondage in West India. From Past to Present. New Delhi: Oxford University
Press, 2007.
Brenner, Robert. “Agrarian Class Structure and Economic Development in Pre-industrial
Europe.” Past and Present, 70 (1976): 30–74.
Brion Davis, David. The Problem of Slavery in the Western Culture. Ithaca: Cornell University
Press, 1966.
Brook, Timothy. “Valeurs et prix dans la Chine des Ming (1368–1644).” Series of papers
presented at Collège de France, Paris, France, October 6, 13, 20, 27, 2010.
——. The Troubled Empire: China in the Yuan and Ming Dynasties. Cambridge, London:
Harvard University Press, 2010.
Brower, Daniel. “Urbanization and Autocracy: Russian Urban Development in the First
Half of the Nineteenth century.” Russian review, 42, 4 (1983): 377–402.
Brugmans, I.J. De arbeidende klasse in Nederland in de de 19e eeuw (1813–1870). Utrecht-
Antwerpen: Aula-boeken, 1858; 3rd ed. first print 1929.
Brym, Robert. “A note on Raznochintsy.” Journal of Social History, 10 (1976–7): 354–359.
Burbank, Jane. Russian Peasants Go to Court: Legal Culture in the Countryside, 1906–1917.
Bloomington: Indiana University press, 2004.
Burnette Joyce. “An Investigation of the Female-Male Wage Gap during the Industrial
Revolution in Britain.” Economic History Review 50 (May 1997): 257–281.
Bush, Michael (ed.). Serfdom and Slavery: Studies in Legal Bondage. Manchester: Man-
chester University Press, 1996.
Cain, Peter J. and Hopkins, Anthony G. British Imperialism: Innovation and Expansion,
1688–1914. London: Longman, 1993.
Callet, R.P. Histoire des Rois. Tantaran’Ny Andriana. Trad. George S. Chapus et Emmanuel
Ratsimba. Tananarive: Editions de la Librairie de Madagascar, 1974.
Campbell, Cameron and Lee, James. “Free and Unfree Labour in Qing China: Emigra-
tion and Escape Among the Bannermen of Northeast China, 1789–1909.” History of the
Family, 6 (2001): 455–476.
Campbell, Gwyn (ed). The Structure of Slavery in Indian Ocean, Africa and Asia. London,
Portland: Frank Cass, 2004.
Campbell, Gwyn, Miers, Suzanne, and Miller Joseph. Children in Slavery Through the Ages.
Athens, Ohio: Ohio University Press, 2009.
Campbell, Gwyn. “An Industrial Experiment in Pre-colonial Madagascar, 1825–1861,” Journal
of Campbell, Gwyn. “Children and Slavery in the New World: A Review.” Slavery and
Abolition 27,2 (August 2006): 261–86.
——. “Disease, Cattle and Slaves: The Development of Trade between Natal and Madagas-
car, 1875–1904.” African Economic History 19 (1990–91): 105–16.
——. “Gold Mining and the French Takeover of Madagascar, 1883–1914.” African Economic
History 17 (1988): 1–28.
——. “Indians and Commerce in Madagascar, 1869–1896” University of the Witwatersrand.
African Studies Seminar Paper 345 (1993).
——. “Introduction: Abolition and its Aftermath in the Indian Ocean world.” In idem. Ed.
Abolition and its Aftermath: 1–28.
294 references

——. “Madagascar and Mozambique in the Slave Trade of the Western Indian Ocean
1800–1861.” In Clarence-Smith. Ed. Economics of the Indian Ocean Slave Trade: 166–93.
——. “Madagascar and the Slave Trade, 1810–1895.” Journal of African History 22,2 (1981):
203–27.
——. “Slavery and Fanompoana: The Structure of Forced Labour in Imerina (Madagascar),
1790–1861” Journal of African History 29,3 (1988), 463–86.
——. “The Role of the London Missionary Society in the Rise of the Merina Empire, 1810–
1861.” PhD. University of Wales, Swansea, 1985.
——. “The State and Pre-colonial Demographic History: The Case of Nineteenth Century
Madagascar.” Journal of African History, 31,3 (1991): 415–45.
——. An Economic History of Imperial Madagascar, 1750–1895. The Rise and Fall of an Island
Empire. Cambridge: Cambridge University Press, 2005.
——. Ed. Abolition and Its Aftermath in Indian Ocean Africa and Asia. London: Routledge,
2005.
——. Ed. The Structure of Slavery in Indian Ocean Africa and Asia. London: Frank Cass,
2004.
Campbell, Gwyn; Miers, Suzanne; and Miller, Joseph C. “Introduction.” In idem. Eds. Chil-
dren in Slavery Through the Ages: 1–15.
——. Eds. Children in Slavery Through the Ages. Athens: Ohio University Press, 2009.
Campbell, Perisa. Chinese Coolie Emigration to Countries within the British Empire: To Coun-
tries within the British Empire. London: Routledge, 1971.
Capela, José and Medeiros, Eduardo. O Tráfico de Escravos de Moçambique para as Ilhas do
Índico 1720–1902. Maputo: Departamento de História da Universidade Eduardo.
Carey, Peter B.R. The power of prophecy: prince Dipanagara and the end of an old order in
Java, 1785–1855. Leiden: KITLV, 2008.
Carey, Peter. “Waiting for the “Just King”: The Agrarian World of South-Central Java from
Giyanti (1755) to the Java War (1825–1830).” Modern Asian Studies 20,1 (1986): 59–137.
Carter, Marina and Gerbeau, Hubert. “Covert Slaves and Coveted Coolies in the Early Nine-
teenth Century Mascareignes.” Slavery and Abolition 9,3 (1988): 193–207.
Carter, Marina. Servants, Sidars and Settlers. Indians in Mauritius, 1834–1874. Delhi: Oxford
University Press, 1995.
Cartier, Michel “Travail et idéologie dans la Chine antique”, in Le travail et ses représenta-
tions. Paris: Éditions des archives contemporaines, 1984: 275–304.
Cayez, Pierre. Crises et croissance de l’industrie lyonnaise, 1850–1900. Paris: Éditions du
CNRS, 1980.
——. Métiers Jacquard et hauts fourneaux. Aux origines de l’industrie lyonnaise. Lyon: Pres-
ses universitaires de Lyon, 1978.
Čechura, Jaroslav. “Die Gutswirtschaft des Adels in Böhmen in der Epoche vor der Schlacht
am Weißen Berg.” Bohemia 36,1 (1995): 1–18.
——. Adelige Grundherrn als Unternehmer. Zur Struktur südböhmischer Dominien vor 1620.
Munich/Vienna: Oldenbourg, 2000.
——. “Dominium Smiřických—protokapitalistický podnikatelský velkostatek předbělo­
horských Čech.” Český časopis historický 90 (1992): 508–38.
Cerman, Markus. “Arbeitsrenten in den nordböhmischen Herrschaften Frýdlant und Libe-
rec im 18. Jahrhundert.” In Enderle-Burcel, Getrude et al. (eds.). “Discourses—Diskurse.”
Essays for Mikuláš Teich und Alice Teichová. Praha: Univerzita Karlova, 2008: 275–92.
——. “Gutsherrschaft vor dem ‘Weißen Berg’. Zur Verschärfung der Erbuntertänigkeit in
Nordböhmen 1380 bis 1620.” In Peters, Jan (ed.). Gutsherrschaftsgesellschaften im euro-
päischen Vergleich. Berlin: Akademie, 1997: 91–111.
——. “Venkovské společnosti a agrárně-dějepisecké tvoření modelů v nové perspektivě.
Srovnávací analýza středo- a východoevropských agrárních struktur od 14. do 17. století.”
Časopis Matice moravské 120,2 (2001): 337–95.
Chakrabarty, Dipesh. Rethinking Working-Class History. New Delhi: Oxford University
Press, 1989.
 references 295

Chandler, Alfred. Scale and Scope: The Dynamics of Industrial Capitalism. Cambridge:
Belknap Press of Harvard University Press, 1990.
Chandra, Bipan. “The Colonial Legacy.” In Bimal Jalan. Ed. The Indian Economy. Problems
and Perspectives. New York: Viking, 1992: 1–32.
Chandra, Siddharth. “The Role of Female Labor in the Late Colonial Netherlands Indies.”
Indonesia 74 (2002): 103–35.
Chanock, Martin. “South Africa, 1841–1924: Race, Contract and Coercion.” In Hay and Cra-
ven (eds.). Masters, Servants and Magistrates: 338–364.
Chatterjee, Indrani. “Abolition by Denial: The South Asian example.” In Campbell. Ed. Abo-
lition and its Aftermath: 150–68.
Chaudhury, Kirti N. The Trading World of Asia and the English East India Company, 1660–
1760. Cambridge: Cambridge University Press, 1978.
Chen Shoushi. “Ming Qing zhiji shiliao (1)” (Historical Materials on the Ming-Qing Transi-
tion—1). Guoxue yuekan, 2.3 (1927): 111–120.
——. “Ming Qing zhiji shiliao (2)” (Historical Materials on the Ming-Qing Transition—2).
Shixue yu dixue, 2 (1929): 1–15.
Chitty, Joseph. A Practical Treatise on the Law Relating to Apprentices and Journeymen, and
to Exercising Trades. London: W. Clarke & Sons, 1812.
Chrétien, Jean-Pierre. “Demography and ecology in East Africa at the end of the nine-
teenth century: an exceptional crisis?” Cahiers d’Etudes Africaines 27 (1–2), 105–6 (1987):
43–59.
Christiansen, Palle Ove. A manorial world. Lord, peasants and cultural distinctions on a
Danish estate, 1750–1980. Oslo: Scandinavian University Press, 1996.
Clarence-Smith, William Gervase. “Islam and the abolition of the slave trade in the Indian
Ocean.” In Campbell. Ed. Abolition and its Aftermath: 137–49.
——. “The Redemption of Child Slaves by Christian Missionaries in Central Africa, 1878–
1914.” In Campbell, Miers and Miller. Eds. Child Slaves in the Modern World: 252–79.
——. Ed. The Economics of the Indian Ocean Slave Trade in the Nineteenth Century. London:
Frank Cass, 1989.
——. Islam and the Abolition of Slavery. New York: Oxford University Press, 2006.
Clark, Gregory. “Productivity Growth Without Technical Change in European Agriculture
before 1850”, The Journal of Economic History, 47,2 (  june 1987): 419–432.
Cohen, Abner. “Cultural Strategies in the Organization of Trading Diasporas.” In Meillas-
soux, Claude (ed.). The Development of Indigenous Trades and Markets in West Africa.
London: Routledge, 1971: 266–278.
Collins, E.J.T. “Migrant Labour in British Agriculture in Nineteenth Century.” Economic
History Review, 29, 1 (1976): 38–59.
Condominas, Georges. Formes extrêmes de la dépendance. Contribution à l’étude de l’escla-
vage en Asie du Sud-Est. Paris: EHESS, 1998.
Coolie Budget Commission (translation by Robert van Niel). Living conditions of plantation
workers and peasants on Java in 1939–1940. Ithaca, New York: Modern Indonesia Project
Translation Series, 1956.
Cooper, Frederick. Plantation Slavery on the East Coast of Africa. Portsmouth, New Hamp-
shire: Heinemann, 1997.
Coquery-Vidrovitch, Catherine and Moniot, Henri. L’Afrique noire de 1800 à nos jours. Paris:
PUF, 1992.
Coquin, François-Xavier. La grande commission législative 1767–1768: les cahiers de doléan-
ces urbains, province de Moscou. Paris: Publications de la Sorbonne, 1972. 
Cornell, Laurel and Akira Hayami. “The Shumon Aratame Cho: Japan’s Population
Registers.” Journal of Family History 11, no. 4 (November 1986): 311–28
Cottereau, Alain. “Industrial Tribunals and the Establishment of a Kind of Common Law of
Labour in Nineteenth Century France”. In Steinmetz (ed.) Private Law and Social Inequa-
lity: 203–226.
296 references

——. “La désincorporation des métiers et leur transformation en “publics intermédiaires”:

Lyon et Elbeuf, 1790–1815”. in La France, malade du corporatisme? XVIIIe–XXe siècles, ed.
Steve Kaplan and Philippe Minard, 97–145. Paris: Belin, 2004.
——. “The fate of collective manufactures in the industrial world: The silk industry of
Lyons and London, 1800–1850.” in World of possibilities. Flexibility and Mass Production
in Western Industrialisation, 75–152. Cambridge: Cambridge University Press, 1997.
——. “Droit et bon droit. Un droit des ouvriers instauré, puis évincé par le droit du travail,
France, XIXe siècle.” Annales 57, 6 (2002): 1521–57.
——. “The Distinctiveness of Working-Class Cultures in France, 1848–1900.” in Working-
Class Formation. Nineteenth-Century Patterns in Western Europe and the United States,
ed. Ira Katznelson and Aristide R. Zolberg, 111–154. Princeton N.J.: Princeton University
Press, 1986.
Craft, Nicolas. British Economic Growth During the Industrial Revolution. Oxford: Clarendon
Press, 1985.
Crisp, Olga. Studies in Russian Economy before 1914. London: Basingstoke, 1976.
Cross Gary (ed.). Worktime and Industrialization: An International History. Philadelphia:
Temple University Press, 1988.
——. A Quest for Time. The Reduction of Work in Britain and France, 1840–1940. Berkeley:
University of California Press, 1989.
Crouzet, François. British Ascendant: Comparative Studies in Franco-British Economic His-
tory. Cambridge: Cambridge University Press, 1990.
Crummey, Robert. “Sources of Boyar Power in the Seventeenth Century.” Cahiers du monde
russe and soviétique, 34, 1–2 (1993): 107–118.
Crush, Jonathan. “Power and Surveillance on the South African Gold Mines”. Journal of
South African Studies 18 (1992): 825–44.
Cunningham, Hugh and Viazzo, Pier Paolo (eds.). Child Labour in Historical Perspective,
1800–1985. Historical Studies from Europe, Japan Florence: Unicef, 1996.
Curtin, Philip. Cross-Cultural Trade in World History. Cambridge: Cambridge University
Press, 1984.
Czap, Peter jr. “Peasant-class Courts and Peasant Customary Justice in Russia, 1815–1912.”
Journal of Social History, 1 (1967): 148–178.
Da Ming huidian (Collected Statutes of the Great Ming, dated 1587). Reprinted in vol. 617
of Wenyuan ge siku quanshu, Taipei: Taiwan Shangwu yinshuguan, 1986.
Da Qing Xuantong zhengji shilu (Veritable Records of the Xuantong Reign). Taipei: Huawen
shuju, 1968.
Das Gupta, Ranajit. “Structure of the Labour Market in Colonial India.” Economic and
Political Weekly 16, 44/46, (1981): 1781–3, 1785, 1787–9 + 1791, 1793, 1795–7, 1799, 1801, 1803,
1805–6.
Dasgupta, Keya. “Plantation labour in the Brahmaputra Valley: Regional Enclaves in a
Colonial Context.” In Campbell. Ed. Abolition and its Aftermath: 169–79.
Daunton, Martin. Progress and Poverty: An Economic and Social History of Britain 1700–1850.
Oxford: Oxford University Press, 1995.
Davis, Charles Ellis. The Master and Servant Act of 1867. London: Butterworths, 1868.
De Bandt, Jacques. “La filière comme méso-système.” In Traité d’économie industrielle by
Richard Arena and al., 232–238. Paris: Economica, 1991.
De Graaf, H.J. and Th.G.Th. Pigeaud. De eerste moslimse vorstendommen op Java. Studiën
over de staatkundige geschiedenis van de 15e en 16e eeuw. Verhandelingen van het Konink-
lijk Instituut voor Taal-, Land- en Volkenkunde 69.’s-Gravenhage: Martinus Nijhoff,
1974.
De Vries Jan. “The Industrial Revolution and the Industrious Revolution.” Journal of Eco-
nomic History, 54, 2 (1994): 249–270.
Deakin, Simon and Wilkinson, Frank. The Law of the Labour Market: Industrialization,
Employment, and Legal Evolution. Oxford: Oxford University Press, 2005.
Deakin, Simon. ‘‘Does the ‘Personal Employment Contract’ Provide a Basis for the Reunifica-
tion of Labour Law?” Industrial Law Journal, 36 (2007): 68–83.
 references 297

——. “Legal Origin, Juridical Form and Industrialisation in Historical Perspective: The
Case of the Employment Relationship and the Joint Stock Company”. Socio-Economic
Review, 7 (2009): 35–65.
——. “The Contract of Employment: a Study in Legal Evolution.” Historical Studies in
Industrial Relations, 11 (2001): 1–36.
Deane, Phyllis. “Capital Formation in Britain Before the Railway Age.” Economic Develop-
ment and Cultural Change, 9,3 (1961): 352–368.
Delaye, Karine. “Slavery and Colonial Representations in Indochina from the Second
Half of the Nineteenth to the Early Twentieth Century.” In Campbell. Ed. Structure of
Slavery: 129–42.
Dennison, Tracy. “Did Serfdom Matter? Russian Rural Society, 1750–1860.” Historical
Research, 79, 203 (2003): 74–89.
——. The Institutional Framework of Russian Serfdom. Cambridge: Cambridge University
Press, 2011.
Deutsch, Jan-Georg. “Notes on the Rise of Slavery & Social Change in Unyamwezi c. 1860–
1900.” In Henri Médard and Shane Doyle. Eds. Slavery in the Great Lakes Region of East
Africa. Oxford: James Currey, 2007: 76–110.
Dewerpe, Alain. Le monde du travail en France, 1800–1950. Paris: Armand Colin, 1989.
Dickson, W.P. Report on the Extra Mural Employment of Prisoners on the Sirhind Canal at
Rupar, Punjab between 1869–81. Lahore: Government Press, 1884.
Didry, Claude. Naissance de la convention collective, débats juridiques et luttes sociales en
France au début du 20ème siècle. Paris: EHESS, 2002.
Diepenhorst, P.A. De Nederlandsche arbeidswetgeving. 2 vols. Utrecht: s.n., 1921–1930.
Domanovszky, Alexander. “Zur Geschichte der Gutsherrschaft in Ungarn.” In Patzelt, Erna
(ed.). Wirtschaft und Kultur. Festschrift für A. Dopsch. Baden: Rohrer, 1938: 441–69.
Domar, Evsey and Machina, Michael. “On the Profitability of Russian Serfdom.” The Journal
of Economic History, 44, 4 (1984): 919–55.
Dorošenko, Vassily. “Der ostbaltische Herrenhof des 16.–18. Jahrhunderts als ‘Getreidefa-
brik’.” Guarducci, A. (ed.), Agricoltura e trasformazione dell’ambiente secoli XIII–XVIII.
Prato, Le Monnier, 1981.
Drescher, Seymour. “Free Labor vs Slave Labor: The British and Caribbean Cases”. In
Terms of Labor: Slavery, Serfdom, and Free Labor, edited by Stanley, L. Engerman, 50–86,
Stanford: Stanford University Press, 1999.
——. Capitalism and Antislavery. New York: Oxford University Press, 1987.
Drescher, Seymour and Stanley Engerman, eds. A Historical Guide to World Slavery. New
York: Oxford University Press, 1998.
Druzhinin, Nikolai. Gosudarstvennye krest’iane i reforma P.D. Kiseleva (State peasants and
Kiselev’s reforms). Moscow: Nauka, 1958.
Dubs, Homer H., trans. The Works of Hsüntze. London: Arthur Probsthain, 1928.
Duby, Georges. Les trois ordres ou l’imaginaire du féodalisme. Paris: Seuil, 1978.
Duchet, Michèle. Anthropologie et histoire au siècle des lumières. Paris: Albin Michel, 1971.
Dukes, Paul. Catherine the Great and the Russian Nobility. Cambridge: Cambridge Univer-
sity Press, 1967.
Dull, Jack L. Ed. Han Social Structure. T’ung-Tsu Ch’ü. Seattle, Washington, University of
Washington Press, 1969.
Eaton, Henry, “Cadasters and Censuses of Muscovy.” Slavic Review, 26, 1 (1967): 54–69.
Eerenbeemt, H.F.J.M. van den. “Van aalmoes naar werkverschaffing. Nieuwe ideeën over
de bestrijding van het pauperisme te ’s-Hertogenbosch bij de overgang van de 18e naar
de 19e eeuw.” Economisch- en social-historisch jaarboek. Bijdragen tot de Geschiedenis van
Nederland 33 (1970): 1–26.
Efremova, Natalia. Sudoustroistvo Rossii v XVIII–pervoi polovine IXIe v. (the juridical system
in Russia, 18th-first half of the 19th century). Moscow: Nauka, 1993 .
Eldredge, Niles (ed.). Time Frames: The Rethinking of Darwinian Evolution and the Theory of
Punctuated Equilibria. New York: Simon and Schuster, 1985: 82–115.
298 references

Eldredge, Niles and Gould, Stephen. “Punctuated Equilibria: an Alternative to Phyletic Grad-
ualism”. In Eldredge, Niles (ed.). Time Frames: The Rethinking of Darwinian Evolution and
the Theory of Punctuated Equilibria. New York: Simon and Schuster, 1985: 82–115.
Ellis, William. History of Madagascar. Comprising also the Progress of the Christian Mission
established in 1818; and an Authentic Account of the Recent Martyrdom of Rafaravavy; and
of the Persecution of the Native Christians. 2 vols. London: Fisher, Son, & Co., 1838.
Elson, R.E., ‘Sugar Factory Workers and the Emergence of “Free Labour” in Nineteenth
Century Java’, Modern Asian Studies 20, 1 (1986) 139–174.
—— Village Java Under the Cultivation System. Sydney: Allen and Unwin, 1994.
Eltis David. Coerced and Free Migration: Global Perspectives. Stanford: Stanford Univ. Press,
2002.
Elvin, Mark. The Pattern of the Chinese Past. Stanford: Stanford University Press, 1973.
Enders, Lieselott. “Frondienst in der Altmark. Analyse und Vergleich.” Jahrbuch für
Geschichte Mittel- und Ostdeutschlands 49 (2004): 83–147.
——. Die Prignitz. Geschichte einer kurmärkischen Landschaft. Potsdam: Verlag für Berlin-
Brandenburg, 2000.
——. Die Uckermark. Geschichte einer kurmärkischen Landschaft vom 12. bis zum 18. Jahr-
hundert. Weimar: Böhlau, 1992.
——. “Entwicklungsetappen der Gutswirtschaft vom Ende des 15. bis zum Beginn des 17.
Jahrhunderts, untersucht am Beispiel der Uckermark.” Jahrbuch für Geschichte des Feu-
dalismus 12 (1988): 119–65.
Engerman, Stanley, L. “Slavery and Emancipation in Comparative Perspective: A Look at
Some Recent Debates’. Journal of Economic History, 46 (1986): 317–39.
Engerman, Stanley (ed.). Terms of Labor. Slavery, Freedom and Free Labor. Stanford: Stan-
ford university press, 1999.
Eno, Omar A. “The Abolition of Slavery and the Aftermath Stigma: The Case of the Bantu/
Jareer People on the Benadir Coast of Southern Somalia.” In Campbell. Ed. Abolition
and its Aftermath: 83–93.
Fahmy, Khaled. All the Pasha’s men: Mehmed Ali, his army, and the making of modern Egypt.
Cambridge: Cambridge University Press, 1997.
Falconnet, Joachim. Indicateur annuaire de la fabrique de soie de Lyon pour l’année 1843,
contenant près de 4000 adresses spéciales. Première année. Lyon: Dumoulin Ronnet et
Sibuet, 1843.
——. Indicateur annuaire de la Fabrique d’étoffes de soie contenant les adresses des négo-
ciants-fabricants, des professions qui se rattachent à cette industrie, et des notices spéciales
à la Fabrique. Lyon: Veuve Ayné, 1849.
Farmer, Edward L. “The Status of the Person in Early Ming Law.” Journal of Chinese Studies
2.1 (1985): 73–80.
——. Zhu Yuanzhang and Early Ming Legislation: The Reordering of Chinese Society Follow-
ing the Era of Mongol Rule. Leiden: E.J. Brill, 1995.
Farris, William Wayne. Japan’s Medieval Population: Famine, Fertility, and Warfare in a
Transformative Age. Honolulu, HI: University of Hawai’i Press, 2006.
Fasseur, Cees. The politics of colonial exploitation: Java, the Dutch, and the cultivation sys-
tem. Ithaca, N.Y.: Cornell University, 1992.
Federico, Giovanni. Feeding the World. An Economic History of Agriculture, 1800–2000.
Princeton: Princeton University press, 2008.
Feeny, David. “The Decline of Property Rights in Man in Thailand, 1800–1913.” Journal of
Economic History 49, 2 (1989): 285–96. “The Demise of Corvée and Slavery in Thailand,
1782–1913.” In Klein. Ed. Breaking the Chains: 83–111.
Feigl, Helmuth. “Die Entwicklung der schlesischen Grundherrschaft unter den Habsbur-
gern (1526 bis 1742).” In Baumgart, Peter (ed.). Kontinuität und Wandel. Schlesien zwi-
schen Österreich und Preußen. Sigmaringen: Thorbecke, 1990: 135–65.
Feinstein, Charles, and Pollard, Sidney (eds.). Studies in Capital Formation in the United
Kingdom, 1750–1920. Oxford: Clarendon Press, 1988.
 references 299

Feinstein, Charles. “Capital Formation in Great Britain” in: Peter Mathias and Michael
Postan (eds), The Cambridge Economic History of Europe, vol. VII: The industrial Econ-
omies: Capital, Labor and Enterprise. Cambridge: Cambridge University Press, 1978:
28–94.
Fernyhough, Timothy. “Slavery and the Slave Trade in Southern Ethiopia in the 19th Cen-
tury.” In William Gervase Clarence-Smith. Ed. The Economics of the Indian Ocean Slave
Trade in the Nineteenth Century. London: Frank Cass, 1989: 103–30.
Festy, Octave. Le mouvement ouvrier au début de la monarchie de juillet (1830–1834). Paris:
Cornély, 1908.
Field, Daniel. The End of Serfdom: Nobility and Bureaucracy in Russia, 1855–1861. Cambridge
(Mass.): Harvard University Press, 1976.
Filliot, Jean-Michel. La traite des esclaves vers les Mascareignes au XVIIIe siècle. Paris, ORS-
TOM, 1974.
Finley, Moses. Ancient Slavery and Modern Ideology. New York: Viking Penguin, 1980.
Fitzpatrick, Brian. “Indentured Labour in Australia.” Labour History 9 (November 1965):
3–5.
Fogel, Robert William, and Engerman, Stanley, Time on the Cross. The Economics of Nego
American Slavery. Little: Brown, 1974.
Franck, Penelope. Rural Economic Development in Japan: From the Nineteenth Century to
the Pacific War. London; New York: Routledge, 2006.
Frank, Stephen, Crime, Cultural Conflict and Justice in Rural Russia, 1856–1914, Berkeley, Uni-
versity of California Press, 1999.
Fransen van de Putte. I.D. De regeling en uitbesteding van de suikercontracten op Java. Goes:
s.n., 1860.
Freeman, Joseph John and Johns, David. A Narrative of the Persecution of the Christians
in Madagascar with details of the Escape of the Six Christian Refugees now in England.
London: John Snow, 1840.
Fridenson, Patrick and Reynaud, Bénédicte. La France et le temps de travail (1814–2004).
Paris: Odile Jacob, 2004.
Frierson, Cathy. “I Must Always Answer to the Law . . .: Rules and Responses in the Reformed
Volost’ Court.” Slavonic and East European Review, 75 (1997): 308–334.
——. “Rural Justice in Public Opinion: The volost’ Court Debate.” Slavonic and East Euro-
pean Debate, 64,4 (1986): 526–545.
Fu Fengxiang, comp. Huang Ming zhaoling (Decrees and Proclamations of the August
Ming, dated 1539). Reprinted in vol. 58 of Siku quanshu cunmu congshu. Jinan: Qilu
shushe, 1996.
Fu Yiling. “Mingji nubian shiliao shebu.” Xieda xuebao 1 (1949): 163–170.
Fuller, William. Strategy and Power in Russia, 1600–1914. New York: Free Press, 1992.
Galenson, David. White Servitude in Colonial America: An Economic Analysis. Cambridge:
Cambridge University Press, 1981.
Gao Ju, ed. Da Minglü jijie fuli (The Great Ming Code with Commentaries and Substatutes,
dated 1610). Beijing: Xiuding falüguan, 1908.
Garden, Maurice. Lyon et les Lyonnais au XVIIIe siècle. Paris: Les Belles Lettres, 1970.
Garland, David. Punishment and Modern Society: A Study in Social Theory. Oxford: Claren-
don Press, 1990.
Gatrell, Peter. The Russian economy, 1850–1917. London: Batsford, 1986.
Géraud, Jean-François. “Les esclaves à l’épreuve de l’industrie: le passage à l’industrie
sucrière à Bourbon, 1810–1830.” In Maestri. Ed. Esclavage et abolitions: 295–308.
Gerbeau, Hubert. “Engagees and Coolies on Réunion Island, Slavery’s Masks and Freedom’s
Constraints.” In Piet C. Emmer. Ed. Colonialism and Migration: Indentured Labour before
and after Slavery. Dordrecht, The Netherlands: Martinus Nijhoff, 1986: 209–36.
Geremek, Bronisław. “Problem siły roboczej w Prusach w pierwszej połowie XV w.”
Przęglad Historyczny 48 (1957): 197–233.
Gershenkron, Alexander. Economic Backwardness in Historical Perspective. Cambridge:
Cambridge University press 1962.
300 references

——. Economic Backwardness in Historical Perspective. Cambridge (Mass.): Harvard Uni-

versity Press, 1962.
Girdlestone, CER. Report on Past Famines in the North-Western Provinces. Allahabad: Gov-
ernment Press, 1868.
Glavnyi General’nyi Shtab, Materialy dlia geografii I statistiki Rossii, sobrannye ofitserami
general’nogo shtaba (Materials for geography and statistics of Russia, collected by offi-
cials of military headquarter) 25 vv. Saint Petersburg, 1860–1868.
Godart, Justin. L’ouvrier en soie. Monographie du tisseur lyonnais. Étude historique, écono-
mique et sociale. Première partie, La réglementation du travail 1466–1791. Lyon et Paris:
Bernoux et Cumin et Arthur Rousseau, 1899.
Goldin, Claudia. “The Quiet Revolution That Transformed Women’s Employment, Educa-
tion, and Family”. American Economic Review, Papers and Proceedings 96 (May 2006):
1–21.
Goldstone, Jack A. “East and West in the Seventeenth century: Political Crises in Stuart
England, Ottoman Turkey, and Ming China.” Comparative Studies in Society and History,
30, 1 (1988): 103–142.
Goody, Jack. “Slavery in Time and Space.” In James L. Watson. Ed., Asian and African Sys-
tems of Slavery. Berkeley: University of California Press, 1980: 16–42.
Gordon, Alec. “Contract Labour in Rubber Plantations: Impact of Smallholders in Colonial
South-East Asia.” Economic and Political Weekly 36, 10 (2001): 847–49; 851–60.
Göttsch, Silke. “Alle für einen Mann . . .”. Leibeigene und Widerständigkeit in Schleswig-
Holstein im 18. Jahrhundert. Neumünster: Wachholtz, 1991.
Grantham, George and MacKinnon, Mary. Labour Market Evolution. London and New
York: Routledge, 1994.
Grantham, George, “Agricultural Supply during the Industrial Revolution: French Evidence
and European Implications.” The journal of Economic History, 49, 1 (1989): 43–72.
——. “Divisions of Labour: Agricultural Productivity and Occupational Specialization in
Pre-Industrial France”. The Economic History Review, 46, 3 (1993): 478–502.
Gregory, Paul. Before Command: an Economic History of Russia from Emancipation to the
First Five-Years plan. Princeton: Pricneton University press, 1994.
Griffiths, David. Hanes Madagascar, neu Grynodeb o Hanes yr ynys, ei Chynyrch, ei Mas-
nach, ac Ansawdd ei Thrigolion. Machynlleth: Richard Jones, 1843.
Gu Yanwu. Rizhi lu jishi (Records of Daily Knowledge, with Explanations, pref. dated
1695). Compiled and annotated by Huang Rucheng. Shanghai: Shanghai guji chuban-
she, 2006.
Guan Zhidao. Congxian weisu yi (Proposals for Safeguarding Customs following the Exam-
ple of the Ancients, dated 1602). Reprinted in vol. 88 of the zi section of the Siku quanshu
cunmu congshu. Jinan: Qilu shushe, 1995.
Guzowski, Piotr. “A Changing Economy: Models of Peasant Budgets in Fifteenth- and
Sixteenth-century Poland.” Continuity and Change 20,1 (2005): 9–25.
Hagen, William W. “How Mighty the Junkers? Peasant Rents and the Seigneurial Profits in
Sixteenth-century Brandenburg.” Past and Present 108 (1985): 80–116.
——. “Village life in East-Elbian Germany and Poland, 1400–1800: subjection, self-defence,
survival.” In Scott, Tom (ed.). The Peasantries of Europe. London/New York: Longman,
1998: 145–89.
——. Ordinary Prussians. Brandenburg Junkers and Villagers, 1500–1840. Cambridge: Cam-
bridge University Press, 2002.
——. “Capitalism in the Countryside in Early Modern Europe: Interpretations, Models,
Debates”. Agricultural History 62,1 (1988): 13–47.
Hall, John Whitney. “The bakuhan system.” In Early Modern Japan, vol. 4 of The Cambridge
History of Japan, edited by John Whitney Hall, 128–182. Cambridge, UK: Cambridge Uni-
versity Press, 1991.
Hallsworth, Sir Joseph and Davies, Rhys. Junior. The Working Life of Shop Assistants:
A Study of Conditions of Labour in the Distributive Trades. Manchester: The Authors,
1910.
 references 301

Halpern, Rick. “Solving the ‘Labour Problem’: Race, Work and the State in the Sugar Indus-
tries of Louisiana and Natal, 1870–1910.” Journal of Southern African Studies 30, 1 (March
2004): 19–40.
Hamano Kiyoshi. “Kinsei toshi no keizai kiki to jinkō: Kyōto Nishijin no jirei kara” [Early
modern urban economic crisis and population: From a case in Nishijin]. Kansai Univer-
sity, Keizai Ronshū [Essays in economics] 53, no. 3 (December 2003): 207–227.
Han Dacheng. Mingdai shehui jingji chutan (An Introduction to the Economy and Society
of the Ming Dynasty). Beijing: Renmin chubanshe, 1986.
Handlin Smith, Joanna. “Lü K’un Compromises with the Common People.” Ming Studies 1
(1975): 79–94.
——. The Art of Doing Good: Charity in Late Ming China. Berkeley, Los Angeles, London:
University of California Press, 2009.
Hansson, Anders. Chinese Outcast. Discrimination and Emancipation in Late Imperial China.
Leiden: Brill, 1996.
Harley, Knick. “British Industrialization Before 1841: Evidence of Slower Growth During the
Industrial Revolution.” The Journal of Economic History, 42, 2 (1982): 267–289.
Harnisch, Hartmut. “Die Gutsherrschaft in Brandenburg. Ergebnisse und Probleme.” Jahr-
buch für Wirtschaftsgeschichte 4 (1969): 117–47.
——. “Die Gutsherrschaft. Forschungsgeschichte, Entwicklungszusammenhänge und
Strukturelemente.” Jahrbuch für Geschichte des Feudalismus 9 (1985): 189–240.
——. “Probleme einer Periodisierung und regionalen Typisierung der Gutsherrschaft im
mitteleuropäischen Raum.” Jahrbuch für Geschichte des Feudalismus 10 (1986): 251–74.
——. Die Herrschaft Boitzenburg. Untersuchungen zur Entwicklung der sozialökonomischen
Struktur ländlicher Gebiete in der Mark Brandenburg vom 14. bis zum 19. Jahrhundert.
Weimar: Böhlau, 1968.
Harnisch, Hartmut, and Heitz, Gerhard. “Einleitung. Die Erforschung der Agrargeschichte
der Epoche des Übergangs vom Feudalismus zum Kapitalismus.” In Harnisch, Hartmut,
and Heitz, Gerhard (eds.). Deutsche Agrargeschichte des Spätfeudalismus. Berlin: Aka-
demie, 1986: 9–36.
Harris, Edward M. “Did Solon Abolish Debt-Bondage?” The Classical Quarterly 52, 2 (2002):
415–30.
Hastie, James. “Journal” (1817–18). In Antony Jully. Ed., “Le voyage de Tananarive en 1817.
Manuscrits de James Hastie.” Bulletin de l’Académie Malgache 2–3–4 (1903): 91–114, 173–
92, 245–69.
Hauser, William. Economic Institutional Change in Tokugawa Japan: Osaka and the Kinai
Cotton Trade. Cambridge: Cambridge University Press, 1974.
Hay, Douglas and Craven Paul (eds). Masters, Servants and Magistrates in Britain and the
Empire, 1562–1955. Chapel Hill and London: The University Of North Carolina Press,
2004.
Hay, Douglas and Rogers, Nicolas. English Society in the Eighteenth Century: Shuttles and
Swords. Oxford: Oxford University press, 1997.
Hay, Douglas. “Masters and Servants in England: Using the Law in the Eighteenth and
Nineteenth Century”. In Steinmetz, Private Law: 227–264.
Haya et al., [to Sugiyama Chōzaemon], “Issatsu no koto,” Letter of request, 8/1776, Sugi-
yama collection, no. S97, Kyoto University Museum Archives, Kyoto.
He Liangjun. Siyou zhai congshuo (Collected Thoughts from the Studio of the Four Fellows,
pref. dated 1570). Beijing: Zhonghua shuju, 1997.
Heerma van Voss, G.J.J. et al., Mr. C. Asser’s handleiding tot de beoefening van het Neder-
lands burgerlijk recht (Deventer: Kluwer, 2008).
Heitz, Gerhard. “Die sozialökonomische Struktur im ritterschaftlichen Bereich Mecklen-
burgs zu Beginn des 18. Jahrhunderts.” Beiträge zur Deutschen Wirtschafts- und Sozialge-
schichte des 18. und 19. Jahrhunderts. Berlin: Akademie, 1962: 1–80.
Hellie, Richard. Enserfment and Military Change in Muscovy. Chicago and London: Univer-
sity of Chicago Press, 1971.
302 references

Henning, Friedrich-Wilhelm. Dienste und Abgaben der Bauern im 18. Jahrhundert. Stuttgart:
Fischer, 1969.
——. Herrschaft und Bauernuntertänigkeit. Beiträge zur Geschichte der Herrschaftsverhält-
nisse in den ländlichen Bereichen Ostpreussens und des Fürstentums Paderborn vor 1800.
Würzburg: Holzner, 1964.
Henvey, F. A Narrative of the Drought and Famine which prevailed in the North-West Prov-
inces during the years 1868, 1869, and beginning of 1870. Allahabad: Government Press,
1871.
Hepple Bob (ed.) The Making of Labour Law in Europe. Cambridge, Cambridge University
Press, 2002.
Heydenrych, Hein. “Railway Development in Natal to 1895.” In Bill Guest and John M.
Sellers. Eds. Enterprise and Exploitation in a Victorian Economy. Aspects of the Economic
and Social History of Colonial Natal. Pietermaritzburg: University of Natal Press, 1985:
47–69.
Hill, Charles. “Pottage for Freeborn Englishmen: Attitudes to Wage Labour in the Sixteenth
and Seventeenth centuries”. In Feinstein, Charles (ed.). Socialism, Capitalism and Eco-
nomic Growth: Essays Presented to Maurice Dobb. Cambridge: Cambridge University
Press, 1967: 338–350.
Hirata, Lucie Cheng. “Free, Indentured, Enslaved: Chinese Prostitutes in Nineteenth-
Century America” Women in Latin America 5,1 (Autumn 1979): 3–29.
Hirdermeier, Manfred. “Was war das mescantsvo ? Zur rechtlichen und sozialen Verfas-
sung des unteren städtischen Standes in Russland”. Forschungen zur osteuropäschen
Geschichte, 36 (1985): 15–53.
——. Bürgertum und Stand in Russland, 1760–1860: Rechtliche Lage und soziale Struktur.
Cologne: Böhlau, 1986.
Ho Ping-ti. The Ladder of Success in Imperial China: Aspects of Social Mobility, 1368–1911. New
York, London: Columbia University Press, 1962.
Hoadley, Mason C. Towards a Feudal Mode of Production. West Java 1680–1800. Singapore:
Institute of Southeast Asian Studies, 1994.
Hobsbawm,  Eric. “The Crisis of the Seventeenth Century”. Past and Present, 6, 1 (1954):
44–65.
Hoch, Steven and Augustine, W. “The Tax Censuses and the Decline of the Serf Population
in Imperial Russia, 1833–1858.” Slavic Review, 38,3 (1979): 208–217.
Hoch, Steven. Serfdom and Social Control in Russia. Petrovskoe, a Village in Tambov. Chicago:
Chicago University Press, 1986.
Horváth, Pavel. “Der Charakter des Spätfeudalismus in der Slowakei.” Studia Historica Slo-
vaca 7 (1974): 78–101.
——. “Poddanská otázka na Slovensku v období tzv. druhého nevol’níctva.” Historické
stúdie 10 (1965): 7–21.
Houben, Vincent, Thomas Lindblad and others. Coolie Labour in Colonial Indonesia. A Study
of Labour Relations in the Outer Islands, c. 1900–1940. Wiesbaden: Harrassowitz, 1999.
Huang Zongli, comp(?). Jujia biyong shilei quanji (Complete Collection of Matters Neces-
sary for Household Use, ca. 1301). Reprinted in vol. 61 of Beijing tushuguan guji zhenben
congkan. Beijing: Shumu wenxian chubanshe, 1988.
Huang, Philip. The Peasant Family and Rural Development in Yangzi Delta, 1350–1988. Stan-
ford: Stanford University Press, 1990.
Huender, W. Overzicht van den economischen toestand der inheemsche bevolking van Java
en Madoera.’s-Gravenhage: Martinus Nijhoff, 1921.
Hui, Lu. “Les esclaves yi des Montagnes Fraîches: entre os blancs  et os noirs”, in Condomi-
nas (ed.), Formes extrêmes: 235–282.
Ibbetson, David. A Historical Introduction to the Law of Obligations. Oxford: Oxford Uni-
versity Press, 1999.
ILO. International Labour Conference, 2001. Papers and proceedings. Geneva: ILO press,
2002.
 references 303

Ingleson, John. ‘ “Bound Hand and Foot”: Railway Workers and the 1923 Strike in Java.’
Indonesia 31 (1982): 53–87.
——. “Life and Work in Colonial Cities: Harbour Workers in Java in the 1910s and 1920s.”
Modern Asian Studies 17, 3 (1983): 455–76.
Inglot, Stefan et al. Historia chłopów polskich. Tom I. Do upadku Rzeczypospołitej szlachec­
kiej. Warszawa: Ludowa spółdzielnia Wydawnicza, 1970.
Ingrao, Charles W. (ed.) State and Society in Early Modern Austria. West Lafayette, Ind.:
Purdue University Press, 1994: 154–80.
IPEC. Every Child Counts: New Global Estimates on Child Labour. Geneva: BIT, 2002.
Isaacman, Allen F. Mozambique: The Africanization of a European Institution, The Zambesi
Prazos, 1750–1902. Madison: The University of Wisconsin Press, 1972.
Ishii Ryōsuke. Shinpen Edo jidai manpitsu [New collection of essays on the Edo period].
Tokyo: Asahi Shinbunsha, 1979–88.
——. Shōnin [Merchants]. Tokyo: Akashi Shoten, 1991.
Ivanhoe, Philip J. Ethics in the Confucian Tradition: The Thought of Mengzi and Wang Yang-
ming. 2nd ed. Indianapolis and Cambridge: Hackett Pub., 2002.
Izydorczyk-Kamler, Anna. “Die Lohnarbeit auf dem Lande in Kleinpolen im 16. und in der
ersten Hälfte des 17. Jahrhunderts.” Studia historiae oeconomicae 20 (1993): 53–75.
Jain, Ravindra K. “Tamilian Labour and Malayan Plantations, 1840–1938.” Economic and
Political Weekly 28, 43 (1993): 2363–5, 2367–70.
Jaschok, Maria and Miers, Suzanne. Eds. Women and Chinese Patriarchy. Submission, Ser-
vitude and Escape. London: Zed Books, 1994.
Jennings, Eric. “Forced labour in Madagascar under Vichy, 1940–42. Autarky, forced labour
and resistance on the ‘Red Island’.” In Edward A. Alpers, Gwyn Campbell and Michael
Salman. Eds. Resisting Bondage in Indian Ocean Africa and Asia. London: Routledge,
2007: 60–8.
Jiang Yonglin and Wu Yanhong. “Satisfying both Sentiment and Law: Fairness-Centered
Judicial Reasoning as seen in Late Ming Casebooks.” In Thinking with Cases: Specialist
Knowledge in Chinese Cultural History, edited by Charlotte Furth, Judith T. Zeitlin, and
Hsiung Ping-chen, 31–61. Honolulu: University of Hawaii Press, 2007.
Jiang Yonglin. The Great Ming Code: Da Ming lü. Seattle: University of Washington Press,
2004.
Jirásek, Jiří. “Moravský venkov před Bilou horou.” Časopis Moravského musea—Acta musei
Moraviae, Vědy společenské—Scientiae sociales 48 (1963): 85–160.
Jones, Robert. “The Nobility and Russian Foreign Policy, 1560–1811.” Cahiers du monde russe
and soviétique, 34,1–2, (1993): 159–170.
——. The Emancipation of the Russian Nobility, 1762–1785. Princeton: Princeton University
Press, 1973.
Joshi, Chitra, “Fettered Bodies: Labouring on Public Works in Nineteenth Century India”.
In Labour Matters: Towards Global Histories, edited by Marcel van der Linden and
Prabhu Mohapatra, 3–21, New Delhi: Tulika Books, 2009.
Kaak, Heinrich. “Vom Erbzinsrecht zur Leibeigenschaft—Entstehung agrarischer Zwangs-
formen im frühneuzeitlichen Brandenburg.” Zeitschrift für Weltgeschichte 8,1 (2007):
71–103.
Kaak, Heinrich. Die Gutsherrschaft. Theoriegeschichtliche Untersuchungen zum Agrarwesen
im ostelbischen Raum. Berlin: de Gruyter, 1991.
——. Eigenwillige Bauern, ehrgeizige Amtmänner, distanzierte fürstliche Dorfherren. Ver-
mittelte Herrschaft im brandenburgischen Alt-Quilitz im 17. und 18. Jahrhundert. Berlin:
Berliner Wissenschafts-Verlag, 2010.
Kagawa Takayuki. “Kinsei shōnin no dōzoku soshiki” [The organization of early modern
merchant lineages]. In Shakai teki sho shūdan [Various social organizations], vol. 6
of Nihon no shakai shi [Social history of Japan] edited by Asao Naohiro et al., 173–80.
Tokyo: Iwanami Shoten, 1988.
304 references

Kahan, Arcadius. The Plow, the Hammer, and the Knout. Chicago: Chicago University Press,
1985.
Kahk, Juhan. Peasants and Lords in the Process of Transition from Feudalism to Capitalism
in the Baltics. Tallinn: Eesti Raamat, 1982.
Kahk, Juhan/Tarvel, Enn. An Economic History of the Baltic Countries. Stockholm, 1997.
Kahn-Freund, Otto. ‘The Personal Scope of English Labour Law, “Servant”,—“Employee”,—
“Workman” ’. Rivista di Diritto del Lavoro, 3 (1966): 508–524.
——. “Servants and Independent Contractors”. Modern Law Review 15 (1951): 504–509.
Kalma, J. Het arbeidscontract met den inlander. Groningen: Gebroeders Hoitsema, 1925.
Kamler, Marcin. Folwark szlachecki w Wielkopolsce w latach 1580–1655. Warszawa: Państ­
wowe wydawnictwo naukowe, 1976.
Kenwood, Albert G. and Lougheed, Alan L. The Growth of the International Economy 1820–
1990. London: Routledge, 1992.
Keppen, Petr’. Deviataia reviziia: issledovanie o chisle zhitelei v rossii v 1851 goda (the ninth
revision: research on the number of inhabitants in Russia in 1851). Saint Petersburg,
1857.
Khovanskii, Nikolai F. “Pomeshchiki I kret’iane Saratovskoi gubernii” (Landlords and peas-
ants of the Saratov province), in Materialy po krepostnomu pravu: Saratovskaia guberniia
(Saratov, 1911): 60–118.
Kim, Bok Rae. “Nobi: A Korean System of Slavery.” In Campbell. Ed. Structure of Slavery:
155–68.
——. “The Third Gender. Palace Eunuchs.” In Campbell, Miers and Miller. Eds, Children in
Slavery through the Ages: 135–52.
Kishimoto Mio. “Maojuan maokao susong yu Qingdai difang shehui” (Local society and
litigations over the prohibition against taking the examinations and entering official
career in the Qing dynasty). In Ming Qing falü yunzuo zhong de quanli yu wenhua (Power
and culture in Ming and Qing law), edited by Qiu Pengsheng and Chen Xiyuan, 145–173.
Taipei: Zhongyang yanjiuyuan, Lianjing chuban gongsi, 2009.
——. “Mindai no shakai shūdan to ‘sen’ no kannen” (Conceptions of Social Categories
and “Degradation” in the Ming). In Higashi Ajia kinsei toshi ni okeru shakaiteki ketsugō:
shomibun, shokaisō no sonzai keitai (Social Cohesion in Early Modern East Asian Towns),
edited by Inoue Tōru and Tsukada Takashi, 3–44. Osaka: Seibundō, 2005.
Kiss, István N. “Der Agrarcharakter der ungarischen Exporte vom 15. bis 18. Jahrhundert.”
Jahrbuch für Wirtschaftsgeschichte 1 (1978): 147–69.
Kjekshus, Helge. Ecology Control and Economic Development in East African History. Lon-
don: Heinemann, 1977.
Klein, Martin A. “Introduction: Modern European Expansion and Traditional Servitude in
Africa and Asia.” In idem. Ed. Breaking the Chains. Slavery, Bondage and Emancipation
in Modern Africa and Asia. Madison, Wisconsin: University of Wisconsin Press, 1993:
3–36.
——. “The Emancipation of Slaves in the Indian Ocean.” In Campbell. Ed. Abolition and
its Aftermath: 198–218.
——. Breaking the Chains. Slavery, Bondage and Emancipation in Modern Africa and Asia.
Madison: University Of Wisconsin Press, 1993.
Kleintjes, Ph. Het Staatsrecht van Nederlandsch-Indië. Beginselen en beschouwingen 2 vols.
Amsterdam: J.H. de Bussy, 1903.
Knight, G.R. “Peasant Labour and Capitalist Production in Late Colonial Indonesia:
The ‘Campaign’ at a North Java Sugar Factory, 1840–1879.” Journal of Southeast Asian
Studies 2 (1980): 245–65.
——. Colonial Production in Provincial Java: The Sugar Industry in Pekalongan-Tegal, 1800–
1942. Amsterdam: VU U.P., 1993.
Knittler, Herbert. “Between East and West. Lower Austria’s noble Grundherrschaft, 1550–
1750.” In Ingrao, C.W. (ed.). State and Society in Early Modern Austria. West Lafayette:
154–80.
 references 305

——. Nutzen, Renten, Erträge. Struktur und Entwicklung frühneuzeitlicher Feudaleinkom-

men in Niederösterreich. Vienna/Munich: Oldenbourg, 1989.
Ko, Dorothy. Teachers of the Inner Chambers: Women and Culture in Seventeenth Century
China. Stanford: Stanford University Press, 1994.
Kobata Atsushi. Okamoto son shi [History of Okamoto Village]. Imadatechō, Fukui: Oka-
moto Son Shi Kankōkai, 1956.
Kobayashi Takehiro. “Bakumatsu ishin ki Kyoto no toshi gyōsei” [City government in the
late Tokugawa and Restoration periods]. In Kindai Kyōto no kaizō [Remodeling early
modern Kyoto] edited by Itō Masuo, 3–15. Kyoto: Minerva Shobo, 2006.
Kohlbrugge, J.H.F. Is grondverhuur aan suikerfabrieken een zegen of een vloek voor den
Javaan? Haarlem: Tjeenk Willink, 1909.
Kolchin, Peter. Unfree Labour: American Slavery and Russian Serfdom. Cambridge (Mass.):
Harvard University Press, 1987.
Kopytoff, Igor and Miers, Suzanne. “African ‘Slavery’ as an Institution of Marginality.” In
idem. Eds. Slavery in Africa. Historical and Anthropological Perspectives. Madison: Uni-
versity of Wisconsin Press, 1977: 3–81.
Koval’chenko Ivan D. Russkoe krepostnoe krest’ianstvo v pervoi polovine XIX v. (The Russian
serf economy during the first half of the nineteenth century). Moscow: Nauka 1967.
Kriedte, Peter. Peasants, Landlords and Merchant Capitalists. Leamington Spa: Berg Pub-
lishers, 1983.
Kriedte, Peter; Medick, Hans, and Schlumbhom, Jürgen. Industrialization Before Industri-
alization. Rural Industry in the Genesis of Capitalism. Cambridge: Cambridge University
Press, 1982.
Kula, Witold. An Economic Theory of the Feudal System: Towards a Model of the Polish Econ-
omy, 1500–1800. London: NLB, 1976.
Kupers, E. Biografisch Woordenboek van het Socialisme en de Arbeidersbeweging in Neder-
land, http://www.iisg.nl/bwsa/bios/kupers.html.
——. Het vraagstuk van de poenale sanctie op de internationale arbeidsconferentie te Genève.
s.l.; s.n., November 1930.
Kutcher, Norman A. Mourning in Late Imperial China: Filial Piety and the State. Cambridge:
Cambridge University Press, 1999.
La Rue, George Michael, “The Brief Life of ‘Ali, the Orphan of Kordofan.” In Campbell,
Miers and Miller. Eds. Children in Slavery through the Ages: 71–87.
Lackner, Michael. “La portée des événements: Réflexions néo-confucéennes sur la ‘rectifi-
cation des noms’ (Entretiens 13.3).” Extrême-Orient—Extrême-Occident 15 (1993).
Landsberg, Helmut E. “Past Climates from Unexploited Written Sources.” In Robert I.
Rotberg and Theodore K. Rabb. Eds. Climate and History. Princeton: Princeton Univer-
sity Press, 1981: 51–62.
Lasker, Bruno. Human Bondage in Southeast Asia. Chapel Hill: University of North Carolina
Press, 1950.
Laslett, Peter, The World We Have Lost. Cambridge: Cambridge University press 1983.
Law, Hugh. “Slavery and Debt Bondage in Sarawak a Hundred Years Ago” (1848). In Bruno
Lasker. Human Bondage in Southeast Asia. Chapel Hill: University of North Carolina.
Le Roux, Rochelle. “The Evolution of the Contract of Employment in South Africa.” Indus-
trial Law Journal, 39 (2010): 139–165.
Lee, James and Feng, Wang. “Malthusian Models and Chinese Realities: The Chinese Demo-
graphic System, 1700–2000”. Population and Development Review, 25, 1 (1999): 33–65.
Lee, John. “Trade and Economy in Preindustrial East Asia, c. 1500–1800: East Asia in the
Age of Global Integration.” The Journal of Asian Studies, 58,1 (1999): 2–26.
Legge, James, trans. The Chinese Classics I: Confucian Analects, the Great Learning, and the
Doctrine of the Mean. London: Trübner & Co., 1861.
——. The Chinese Classics II: The Works of Mencius. London: Trübner & Co., 1861.
Leonard, Carol. Reform and Regicide: The Reign of Peter III of Russia. Bloomington, Ind.:
Indiana University Press, 1999.
306 references

Lequin, Yves. Les ouvriers de la région lyonnaise: 1848–1914. 1, La Formation de la classe

ouvrière régionale. Lyon: Presses universitaires de Lyon, 1977.
Lévy-Leboyer, Maurice, and Bourguignon, François. L’économie française au XIXe siècle.
Analyse macro-économique. Paris: Economica, 1985.
Liesegang, Gerhard. “A First Look at the Import and Export Trade of Mozambique,
1800–1914.” In Gerhard Liesegang, Helma Pasch and Adam Jones. Eds. Figuring African
Trade. Proceedings of the Symposium on the Quantification and Structure of the Import
and Export and Long Distance Trade in Africa 1800–1913. Berlin: Dietrich Reimer 1986:
452–523.
Lieven, Dominic. The Aristocracy in Europe, 1815–1914. New York: Columbia University
Press, 1993.
Lincoln, W. Bruce. The Great Reforms: Autocracy, Bureaucracy and the Politics of Change in
Imperial Russia. Dekalb: Northern Illinois University Press, 1990.
Lindert, Paul, and Williamson, Geoffrey. “English Workers Living Standards During the
Industrial Revolution: a New Look”. Economic History Review, XXXIV (1983): 1–25.
——. “Revising England’s Social Tables, 1688–1812”. Explorations in Economic History, XIX
(1982): 385–408.
Long Wenbin. Ming huiyao (Essentials on the Institutions of the Ming, dated 1887). Beijing:
Zhonghua shuju, 1956.
Longfellow, Linda. “The Second Serfdom in Bohemia: a Case Study of the Rožmberk estate.”
In Volgyes, Ivan (ed.) The Peasantry of Eastern Europe. Roots of Rural Transformations.
New York: Pergamon, 1978: 1–18.
Lourens Piet and Jan Lucassen. Lipsker op de Groninger tichelwerken. Een geschiedenis van
de Groningse steenindustrie met bijzondere nadruk op de Lipper trekarbeiders 1700–1900.
Groningen: Wolters-Noordhoff/Forsten, 1987.
Lovejoy, Paul E. Transformations in Slavery: A History of Slavery in Africa. Cambridge: Cam-
bridge University Press, 1983 (2nd edition 2000).
Lovejoy, Paul and Fayola, Toni. Pawnship, Slavery, and Colonialism in Africa. Asmara: Africa
World Press, 2003.
Lu Shiyi. Fushe jilüe (Brief Account of the Restoration Society). Reprinted in vol. 438 of
Xuxiu siku quanshu. Shanghai: Shanghai guji chubanshe, 1996.
Lubinski, Axel. “Everyday Work and Manorial Culture in Mecklenburg in the 18th and
19th Centuries—Some Problems of Research.” Sundberg, Kersten (ed.) Work and pro-
duction on manors in the Baltic Sea region 1700–1900. Stockholm: Nordiska museet, 2002:
145–64.
Lucassen, Jan and Lucassen, Leo (eds.). Migration, Migration History, History: Old Para-
digms and New Perspectives. Bern: Peter Lang, 1997.
Lucassen, Jan, Tine De Moor and Jan Luiten van Zanden. “The Return of the Guilds:
Towards a Global History of the Guilds in Pre-industrial Times.” International Review of
Social History 53,16 (2008): 5–18.
——. “A Multinational and its Labor Force: The Dutch East India Company, 1595–1795,”
International Labor and Working-Class History 66 (Fall 2004): 12–39.
——. “Labour and early modern economic development.” In A miracle mirrored: The Dutch
Republic in European perspective, edited by Karel Davids and Jan Lucassen, 367–409.
Cambridge [etc.]: Cambridge University Press, 1995.
——. “Mobilization of Labour in Early Modern Europe.” In Early Modern Capitalism. Eco-
nomic and Social Change in Europe, 1400–1800, edited by Maarten Prak, 161–174. London
[etc.]: Routledge Routledge 2001.
Lucassen, Jan. ‘The Brickmakers’ Strikes on the Ganges Canal in 1848–1849’, in: Interna-
tional Review of Social History 51 (2006) Supplement 14.
Luhmann, Niklas. Law as a Social System. Translated by Klaus A. Ziegert, edited by Fatima
Kastner, Richard Nobles, David Schiff and Rosamund Ziegert. Oxford: Oxford University
Press, 2004.
 references 307

Mac Leod, Christine. Inventing the Industrial Revolution. The English Patent System, 1660–
1800. Cambridge: Cambridge University Press, 1988.
Machado, Pedro. “A Forgotten Corner of the Indian Ocean: Gujarati Merchants, Portu-
guese India and the Mozambique Slave Trade, c. 1730–1830.” In Campbell. Ed. Structure
of Slavery: 17–32. Madagascar Times (8 December 1888).
Madariaga, Isabel de. “Catherine II and Serfs: a Reconsideration of Some Problems.” Slavo-
nic and East European Review, 52, 126 (1974): 34–62.
Maestri, Edmond. “Naissance et premiers développements d’un outil économique: le che-
min de fer de la Réunion.” In Claude Wanquet. Ed. Fragments pour une histoire des
économies et sociétés de plantation à la Réunion. St. Denis: Co-éditions du Service des
publications et du Centre de documentation et de recherche en histoire régionale de
l’Université de La Réunion, 1989: 285–303.
——. Ed. Esclavage et abolitions dans l’océan Indien 1723–1860. Systèmes esclavagistes et
abolitions dans les colonies de l’océan Indien. Paris: L’Harmattan, 2002.
Mager, Friedrich. Geschichte des Bauerntums und der Bodenkultur im Lande Mecklenburg.
Berlin: Akademie, 1955.
Man’kov, Aleksandr’. Razvitie krepostonogo prava v Rossii vo vtoroi polovine XVII veka (The
development of serfdom in Russia during the second half of the seventeenth century).
Leningrad: Nauka, 1962.
——. Tseny i ikh dvizhenie v russkom gosudarstve XVI veka (Prices and their dynamics in
sixteenth century Russian state). Moscow, Leningrad: AN SSSR: 1951.
Manning, Patrick. Slavery and African Life. Occidental, Oriental, and African Slave Trades.
Cambridge: Cambridge University Press, 1993.
Mao Yilu. Yunjian yanlüe (Brief Account of Judgments in Songjiang). In vol. 3 of Lidai panli
pandu. Edited by Yang Yifan and others. Beijing: Zhongguo shehui kexue chubanshe,
2005.
Maritch, Streten. Histoire du mouvement social sous le Second Empire à Lyon. Paris: Librairie
Arthur Rousseau, 1930.
Markovits, Claude. “Indian Merchant Networks Outside India in the Nineteenth and Twen-
tieth Centuries: A preliminary survey.” Modern Asian Studies, 33,4 (1999): 883–911.
——. The Global World of Indian Merchants 1750–1947: Traders of Sind from Bukhara to
Panama. Cambridge: Cambridge University Press, 2000.
Marrese, Michelle Lamarche. “The Enigma of Married Women’s Control of Property in
eighteenth-century Russia”. Russian Review, 58, 3 (1999): 380–395.
Marshall, Alfred. Principles of Economics. 8 éd. London: Macmillan and C° Ltd, 1920.
Martin, Virginia. Law and Custom in the Steppe: The Kazakh of the Middle Horde and Russian
Colonialism in the nineteenth century. Richmond UK: Curzon Press, 2001.
Materialy dlia istorii krepostnogo prava v Rossii. Izvlecheniia iz sekret’nykh otchetov minis-
terstva vnutrennykh del za 1836–1856 gg., (Materials for the history of serfdom in Russia.
Excerpts from the secret reports of the ministry of interior 1836–1856). Berlin, 1873.
Mathé, aîné. Les tisseurs de Lyon 1769–1900. Lyon: Chambre syndicale des tisseurs de Lyon,
A. Rey et Cie, 1900.
Mathee, Rudolph P. The Pursuit Of Pleasure: Drugs And Stimulants In Iranian History, 1500–
1900. Princeton: Princeton University Press, 2005.
Mather, Celia E. “Industrialization in the Tangeran Regency of West Java: Women Workers
and Islamic Patriarchy.” Bulletin of Concerned Asian Scholars 15,2 (1983): 2–17.
Maur, Eduard. “Das Gesinde in Böhmen in der frühen Neuzeit.” In Fauve-Chamoux, Antoi-
nette, and Fialová, Ludmila (eds.). Le phénomène de la domesticité en Europe, XVIe–XXe
siècles. Praha: Sociologický ústav AVČR, 1997: 75–94.
——. “Das Gesinde in Böhmen nach dem Soupis poddaných podle víry 1651.” In Cerman,
Markus, and Zeitlhofer, Hermann (eds.). Soziale Strukturen in Böhmen. Ein regiona-
ler Vergleich von Wirtschaft und Gesellschaft in Gutsherrschaften, 16.-19. Jahrhundert.
Munich/Vienna: Oldenbourg, 2002: 111–25.
308 references

——. “K demografickým aspektům tzv. druhého nevolnictví.” Historická demografie 8

(1983): 7–43.
——. “Poddanská otázka v předbelohorských Čechách.” Folia historica bohemica 11 (1987):
133–59.
Maybaum, Heinz. Die Entstehung der Gutsherrschaft im nordwestlichen Mecklenburg (Amt
Gadebusch und Amt Grevesmühlen). Stuttgart: Steiner, 1926.
McDermott, Joseph P. “Bondservants in the T’ai-hu Basin: A Case of Mistaken Identities.”
Journal of Asian Studies 40, 4 (1981): 675–701.
McKeown, Adam M. Melancholy Order. Asian Migration and the Globalization of Borders.
New York: Columbia University Press, 2008.
——. “Global Chinese Migration, 1850–1940.” Paper presented at the conference of the
International Society for the Study of Chinese Overseas (ISSCO) V, Helsignor, Denmark
(May 10–13, 2004).
McNeill, William Hardy. Plagues and Peoples. New York: Anchor Books, 1976.
Meier, Susan and Kopytoff, Igor (eds). Slavery in Africa: Historical and Anthropological Per-
spectives. Madison: University Of Wisconsin Press, 1977.
Meillassoux, Claude. Anthropologie de l’esclavage. Paris: PUF, 1986.
——. L’esclavage en Afrique précoloniale. Paris: Maspero, 1975.
——. The Anthropology of Slavery. The Womb of Iron and Gold. Chicago: University of Chi-
cago Press, 1991.
Mel’nitskii, Nikolai. Sbornik svedenii o voenno-uchebnykh zavedeniiakh v Rossii (Collection
of documents on the military-experimental units). 4 vv. Saint-Petersburg, 1857–1860.
Melton, Edgar. “Enlightened Seignorialism and Its Dilemmas in Serf Russia, 1750–1830.” The
Journal of Modern History 62, 4 (1990): 675–708.
——. “Gutsherrschaft in East Elbian Germany and in Livonia, 1500–1800: a critique of the
model.” Central European History 21 (1988): 315–49.
——. “Population Structure, the Market Economy, and the Transformation of Guts-
herrschaft in East Central Europe, 1650–1800: The Cases of Brandenburg and Bohemia.”
German History 16, 3 (1998): pp. 297–324.
——. “Proto-industrialization, Serf Agriculture, and Agrarian Social Structure. Two Estates
in 19th Century Russia.” Past and Present 115 (1987): 73–81.
——. “The Decline of Prussian Gutsherrschaft and the Rise of the Junker as Rural Patron.”
German History, 12 (1994): 334–50.
Mendels Franklin. “Proto-industrialization: The First Phase of the Industrialization
Process”. Journal of economic history, 32 (1972): 241–261.
——. “Des industries rurales à la proto-industrialisation: historique d’un changement de
perspective”. Annales ESC, 39 (1984): 977–1008. 
Merritt, Adrian. ‘ “Control” v “Economic Reality”: Defining the Contract of Employment”.
Australian Journal of Law and Society, 1 (1982): 105–124.
Middendorp, W. Twee achterlijke arbeidssystemen voor inboorlingen in Nederlandsch Oost-
Indië (Heerendienst en poenale sanctie). Haarlem: H.D. Tjeenk Willink & Zoon, s.a.
Miers, Suzanne. “Contemporary Forms of Slavery.” In Richard and Zachuermuk, On slav-
ery: 714–747.
Miers, Suzanne and Igor Kopytoff, eds., Slavery in Africa. Historical Anthropological Per-
spectives. Madison: The University of Wisconsin Press, 1977.
Miers, Suzanne and Klein, Martin A. Eds. Slavery in Colonial Africa. London: Frank Cass,
1998.
——. “Slavery and the Slave Trade in Saudi Arabia and the Arab States on the Persian Gulf,
1921–1963.” In Campbell. Ed. Campbell. Ed. Abolition and its Aftermath: 120–36. “Britain
and the Suppression of Slavery in Ethiopia.” Slavery and Abolition 18, 3 (1977): 257–88.
Míka, Alois, Poddany lid v Čechách v první polovině 16. století. Praha: Academia, 1960.
——. “Problém počátku nevolnictví v Čechach.” Československy časopis historický 5 (1957):
226–48.
 references 309

Mikulski, Krzystof, and Wroniszewski, Jan. “Das Vorwerk und die Wandlungen der wirt-
schaftlichen Konjunktur in den polnischen Ländern im 14.–17. Jahrhundert.” In Dygo,
Marian, Gawlas, Sławomir, and Grala, Hieronim (eds.). Ostmitteleuropa im. 14.–17. Jahr-
hundert—eine Region oder Region der Regionen? Warzsawa: DiG, 2003: 115–26.
Miller, Joseph C. “A Theme in Variations: A Historical Schema of Slaving in the Atlantic
and Indian Ocean Regions.” In Campbell. Ed. Structure of Slavery: 169–94.
Miller, Joseph Calder. Slavery and Slaving in World History: A Bibliography, 1900–1996.
Armonk, N.Y: M.E. Sharpe, 1999.
Minard, Philippe. La fortune du colbertisme. Paris: Fayard, 1998.
——. Les corporations en France au XVIIIe siècle: métiers et institutions. In La France,
malade du corporatisme? XVIIIe–XXe siècles ed. Steve Kaplan and Philippe Minard,
39–51. Paris: Belin, 2004.
Ming Chushi yiming daoren (pseud. Zhou Tingying?). Laijiang jishi benmo (A Complete
Narrative of the Lai River Events, pref. dated 1645). Qingshi ziliao 1 (1980): 137–157.
Ming shilu (Veritable Records of the Ming). Taipei: Institute of History and Philology of the
Academia Sinica, 1961–1966.
Mingshi (Standard History of the Ming, dated 1739). Beijing: Zhonghua shuju, 1974.
Ministerstvo gosudarstvennykh imushchest, Istoricheskoe piatidesiatiletnei deiatel’nosti
ministerstva gosudarstvennikh imushchestv, 1837–1887 (50 years activity of the ministry
of state dominions. Saint-Petersburg, 1888.
Ministerstvo Iustitsii, Otchet ministerstvo iustitsii za 1845 (Report of the ministry of justice).
Saint-Petersburg: 1846.
Ministerstvo vnutrennykh del (ministry of interior affairs), Zhurnal ministerstva vnutren-
nikh del (The journal of the ministry of the interior) 1859, 7: 1–46
Mironenko, Serguei V. Samoderzhavie i reformy: politicheskaia bor’ba v Rossii v nachale XIXe
v. (Autocracy and Reforms: Political Struggles in Russia in the Early Nineteenth Cen-
tury). Moscow: Nauka, 1989.
Mironov, Boris. Russkii gorod v 1740–1860 gody (The Russian Town, 1740–1860). Leningrad:
Nauka, 1990.
——. The Social History of Russian Empire. Boulder: Westview Press, 2 vv, 1999.
Mirzai, Behnaz A. “The 1848 abolitionist farmān: a step towards ending the slave trade in
Iran.” In Campbell (Ed). Abolition and its Aftermath: 94–102.
Mitchell, Richard and Howe, John. “The Evolution of the Contract of Employment in Aus-
tralia: a Discussion”. Australian Journal of Labour Law 12 (1999): 113–130.
Miyamoto Mataji. Kabu nakama no kenkyū [Research on trade associations]. Tokyo:
Yuhikaku, 1938.
Moertono, Soemarsaid. State and Statecraft in old Java. A Study of the Later Mataram
Period. Ithaca, New York: MA thesis Modern Indonesia Project, 1963.
Mohapatra, Prabhu P. “Assam and the West Indies, 1860–1920”. In Masters, Servants, and
Magistrates in Britain and the Empire 1562–1955, edited by Douglas Hay and Paul Craven,
Chapel Hill: North Carolina Press, 2004: 455–80.
——. “Regulated Informality: Legal Constructions of Labour Relations in Colonial India
1814–1926”, in Workers in the Informal Sector: Studies in Labour History 1800–2000 edited
by Sabyasachi Bhattacharya, and Jan Lucassen, New Delhi: Macmillan, 2005: 65–95.
Moissonnier, Maurice. La révolte des canuts. Paris: Editions sociales, 1975.
Mokyr, Joel (ed.). The Economics of the Industrial Revolution. Totowa: Rowman and Allan-
held, 1985.
Moll-Murata, Christine, “Chinese Guilds in the Qing Dinasty, 1644–1911”. Paper presented
at the Conference The Return of the Guilds, Utrecht, 2006. http://www.iisg.nl/hpw/
papers/guilds-mollmurata.pdf.
Montfalcon, J.-B. Histoire des insurrections de Lyon, en 1831 et en 1834 d’après des documents
authentiques; précédé d’un essai sur les ouvriers en soie et sur l’organisation de la fabrique.
Lyon: Louis Perrin et Delaunay, 1834.
310 references

Mookherrji, Sudhansu Bimal. The Indenture System in Mauritius, 1837–1915. Calcutta: K.L.
Mukhopadhyay, 1962.
Moon, David. The Abolition of Serfdom in Russia, 1762–1907. London: Pearson Education,
2001.
Morton, Fred. “Small Change. Children in the nineteenth-century East African Slave Trade.”
In Campbell, Miers and Miller (eds.). Children in Slavery Through the Ages: 55–70.
Mouat, F.J. Journal of the Statistical Society of London, 25 (1862): 175–218.
Moullier, Igor. Le ministère de l’Intérieur sous le Consulat et le Premier Empire (1799–1814).
Gouverner la France après le 18 brumaire. Thèse de l’université Lille III, 2004.
Mückenberger, Ulrich and Supiot, Alain. “Ordre public social et communauté. Deux cultu-
res du droit du travail.” In Zimmermann, Bénédicte, Didry, Claude and Wagner, Peter
(eds.). Le travail et la nation. Paris: Éditions de la Maison des Sciences de l’Homme,
2000: 81–105.
Müller, Hans-Heinrich. Märkische Landwirtschaft vor den Agrarreformen von 1807. Ent-
wicklungstendenzen des Ackerbaues in der zweiten Hälfte des 18. Jahrhunderts. Potsdam:
Bezirksheimatmuseum, 1967.
Murray, Stephen O. “Homosexuality among Slave Elites in Ottoman Turkey.” In Stephen O.
Murray and Will Roscoe. Eds. Islamic homosexualities: culture, history, and literature.
New York: New York University Press, 1997: 174–86.
Muszyńska, Jadwiga. Gospodarstwo chłopskie w starostwie sandomierskim 1510–1663. Kielce:
Wyższa Szkoła Pedagogiczna im. Jana Kochanowskiego, 1991.
Nagata, Mary Louise and Kiyoshi Hamano. “Marriage Market in Early Modern Kyoto.” His-
tory of the Family, An International Quarterly, 14 (2009): 36–51.
Nagata, Mary Louise. “Brotherhoods and Stock Societies: Guilds in Pre-modern Japan.”
In The Return of the Guilds edited by Jan Lucassen, Tine De Moor, and Jan Luiten van
Zanden. Special issue, International Review of Social History, 53, S16 (2008), Internation-
aal Instituut voor Sociale Geshiedenis: 121–142.
——. “L’adoption à Kyôto aux XVIIe et XIXe siècles.” Ebisu 31 (Automne-Hiver 2003): 59–83.
(Tôkyô: Maison Franco-Japonaise).
——. “Labor Migration, Family and Community in Early Modern Japan.” In Women, Gen-
der and Labor Migration edited by Pamela Sharpe, 60–84. London and New York: Rout-
ledge Press, 2001.
——. “Mistress or Wife? Fukui Sakuzaemon vs. Iwa, 1819–1833.” Continuity and Change 18,
no. 2 (2003): 1–23.
——. “Names and Name Changes in Early Modern Kyoto, Japan.” In Personal Names in
Asia: History, Culture and Identity edited by Yangwen Zheng and Charles MacDonald,
245–261. Singapore: National University of Singapore 2009.
——. Labor Contracts and Labor Relations in Early Modern Central Japan. London and New
York: Routledge Curzon, 2005.
——. “Leaving the Village for Labor Migration in Early Modern Japan.” In The Road to
Independence: Leaving Home in Eastern and Western Societies, 16th–20th Centuries edited
by Franz van Poppel, Michel Oris, and James Lee, 273–311. Bern-Bruxelles: Peter Lang,
2004.
Nair, Janaki. Miners and Millhands: Work, Culture and Politics in Princely Mysore. New
Delhi: Sage, 1998.
Nakai Nobuhiko. “Commercial change and urban growth in early modern Japan.” Trans-
lated by James L. McClain in Early Modern Japan, vol. 4 of The Cambridge History of
Japan, edited by John Whitney Hall, 519–95. Cambridge, UK: Cambridge University
Press, 1991.
Nakane, Chie. “Tokugawa Society.” In Tokugawa Japan. Edited by Nakane and Shinzaburō
Oishi. Translated by Conrad Totman, 13–231. Tokyo: Tokyo University Press, 1990.
Nakayama Shōtarō. “Shōyu jōzō gyō ni okeru koyō rōdō” [Hired labor in the soy sauce
brewing industry]. Kenkyū Kiyō 24 (Akashi, Hyōgo: Akashi Kōgyō Kōtō Senmon Gakkō
1982): 175–98.
 references 311

Nanjū sha kazoku tori shirabe sho shita gaki [Draft of an investigation of families in dis-
tress]. 1853, Hanaguruma-chō collection, Kyoto City Library for Historical Documents.
Neill, James. The Origins and Role of Same-Sex Relations in Human Societies. Jefferson,
N.C.: McFarland & Co., 2009.
Nelson, Thomas. “Slavery in Medieval Japan.” Monumenta Nipponica 59, no. 4 (Winter
2004): 463–92.
Nieboer, Herman. Slavery as an Industrial System. Cambridge: Cambridge University Press,
2010 (original: 1900).
Nieuwkerk, Karin van. ‘A Trade like Any Other.’ Female Singers and Dancers in Egypt.
Austen: University of Texas Press, 1995.
North, Michael. “Die Entstehung der Gutswirtschaft im südlichen Ostseeraum.” Zeitschrift
für historische Forschung 26, 1 (1999): 43–59.
——. “Lohnarbeit und Fronarbeit in der osteuropäischen Landwirtschaft vom 16. bis zum
18. Jahrhundert.” Zeitschrift für Agrargeschichte und Agrarsoziologie 36 (1988): 1–22.
——. Die Amtswirtschaften von Osterode und Soltau. Vergleichende Untersuchungen zur
Wirtschaft im frühmodernen Staat am Beispiel des Herzogtums Preußen in der zweiten
Hälfte des 16. und in der ersten Hälfte des 17. Jahrhunderts. Berlin: Duncker & Humblot,
1982.
North, Michael. From the North Sea to the Baltic: Essays in Commercial, Monetary and
Agrarian History, 1500–1800. Aldershot: Edward Elgar, 1996. 
Northrup, David. Indentured Labor in the Age of Imperialism, 1834–1922. Cambridge: Cam-
bridge University Press, 1995.
O’ Brien, Patrick. “Agriculture and the Industrial Revolution”. The Economic History Review,
30,1 (1977): 166–181.
——. “Path Dependency, or Why Britain Became an Industrialized and
——. Economic Growth in Britain and France, 1780–1914. Two Paths to the Twentieth Century.
London: Unwin, 1987.
Ogilvie, Sheilagh and Cerman, Markus (eds). European Proto-industrialization. Cambridge:
Cambridge University Press, 1996.
Ogilvie, Sheilagh. “Communities and the Second Serfdom in Early Modern Bohemia”. Past
and Present, 187 (2005): 69–119.
Oliver, Samuel Pasfield. “Sir Robert Townsend Farquhar and the Malagasy Slave Trade,”
Antananarivo Annual and Madagascar Magazine 15 (1891): 319–21.
Ontwerp eener Arbeidswetgeving waarvan de invoering door Deli Planters Vereeniging wordt
aanbevolen vóórdat tot afschaffing der poenale sanctie ter Oostkust Sumatra wordt over-
gegaan, met een Inleiding en Toelichting. s.l.: s.n., s.a.
Orloff, Ann Shola. “Gender and the Social Rights of Citizenship: The Comparative Analy-
sis of Gender Relations and Welfare States”. American Sociological Review, 58,3 (1993):
303–328.
Osakaya Sanzaemon and Suzuya Shōbei, [to Sugiyama Zenzaemon], “Uke tori shōmon
no koto.” Document of receipt, 11/1738, Sugiyama collection no. 62B, Kyoto University
Museum, Kyoto.
Pach, Zsigmond Pál. “Der Bauernaufstand vom Jahre 1514 und die ‘zweite Leibeigenschaft’.”
In Heckenast, Gustav (ed.). Aus der Geschichte der ostmitteleuropäischen Bauernbewe-
gungen im 16.-17. Jahrhundert. Budapest: Akademiai kiadó, 1977: 275–301.
——. “Labour control on the Hungarian landlords’ demesnes in the 16th and 17th cen-
turies.” Gunst, Péter, and Hoffmann, Tamás (eds.). Large estates and small holdings in
Europe in the Middle Ages and modern times. National reports. Budapest: Akademiai
kiado, 1982: 157–73.
——. “Zur Problematik der osteuropäischen Gutswirtschaft: Fronarbeit und Lohnarbeit
auf den ungarischen Herrengütern im 16.–17. Jahrhundert.” Studia historica in honorem
Hans Kruus. Talinn 1971: 154–75.
——. Die ungarische Agrarentwicklung im 16.–17. Jahrhundert. Abbiegung vom westeuropä-
ischen Entwicklungsgang. Budapest: Akadémiai kiadó, 1964.
312 references

Paillard, Yvan-Georges. “Les recherches démographiques sur Madagascar au début de

l’époque coloniale et les documents de ‘ l’AMI’.” Cahiers d’Études Africaines 27, 105–6
(1987): 17–42.
Palmer, Robert. English Law in the Age of the Black Death, 1348–1381 A Transformation of
Governance and Law. Chapel Hill, NC: University of North Carolina Press, 1993.
Pang Shangpeng. Pang shi jiaxun (The Pang Family Instructions). In vol. 976 of Congshu
jicheng chubian. Shanghai: Shangwu yinshuguan, 1939.
Pariset, Ernest. Histoire de la Fabrique lyonnaise. Étude sur le Régime social et économique
de l’Industrie de la soie à Lyon, depuis le XVIe siècle. Lyon: A. Rey imprimeur, 1901.
Parker, Geoffrey and Smith, Lesley M. (eds). The General Crisis of the Seventeenth Century.
London: Routledge and Paul Kegan, 1978.
Parsons, James B. The Peasant Rebellions of the Late Ming Dynasty. Tucson: The University
of Arizona Press, 1970.
Patnaik, Utsa and Dingwaney Manjari (eds). Chains of Servitude: Bondage and Slavery in
India. Madras: Sarngam books, 1985.
Patterson, Orlando. Slavery and Social Death: A Comparative Study. Cambridge: Cambridge
University press, 1982.
Peerthum, Satteeanund. “Le système d’apprentissage à L’île Maurice 1835–1839: plus esclave
mais pas encore libre.” In Maestri. Ed. Esclavage et abolitions: 285–94.
Pei Dazhong, and others. Wuxi Jinkui xianzhi (Gazetteer of Wuxi and Jinkui Districts, dated
1881). Reprinted in Zhongguo fangzhi congshu. Taipei: Chengwen chubanshe, 1970.
Perdue, Peter. “Military Mobilization in Seventeenth and Eighteenth Century China, Rus-
sia and Mongolia.” Modern Asian Studies, 30, 4 (1996): 757–793.
Perlin, Frank. “Proto-Industrialization and Pre-Colonial South Asia.” Past and Present,
No. 98, (Feb., 1983): 30–95.
Perrot, Michèle. Les femmes ou les silences de l’histoire. Paris: Flammarion, 1998.
Perrot, Michelle. Les ouvriers en grève: France, 1871–1890. Paris-La Haye: Mouton, 1973.
Peters, Jan. “Ostelbische Landarmut. Sozialökonomisches über landlose und landarme
Agrarproduzenten im Spätfeudalismus.” Jahrbuch für Wirtschaftsgeschichte 3 (1967):
255–302.
——. Märkische Lebenswelten. Gesellschaftsgeschichte der Herrschaft Plattenburg-Wilsnack,
Prignitz 1500–1800. Berlin: Berliner Wissenschafts-Verlag, 2007.
Petit, Heloïse and and Sauze, Damien.  “Une lecture historique de la relation salariale
comme structure de repartition des aléas. En partant du travail de Salais”. In Eymard-
Duvernay, François (ed.). L’economie des conventions: méthodes et resultats, Tome II,
Paris: La Découverte, 2006: 303–318.
Petráň, Josef. Zemědělská výroba v Čechách v druhé polovině 16. a počátkem 17. století. Praha:
Academia, 1963.
Pétré-Grenouilleau, Olivier. Les traites négrières. Paris: Gallimard, 2004.
Pineo, Huguette Ly-Tio-Fane. Lured Away: The Life History of Indian Cane Workers in Mau-
ritius. Moka, Mauritius: Mahatma Gandhi Institute, 1984.
Pintner, Walter M. Russian Economic Policy Under Nicholas I. Ithaca: Cornell University
Press, 1967.
Piore, Michael,and Sabel, Charles. The Second Industrial Divide: Possibilities for Prosperity.
New York: Basic books, 1984.
Pollard, Sidney. The Genesis of Modern Management: A Study of the Industrial Revolution in
Great Britain. Cambridge, MA: Harvard University Press, 1966.
Polnoe sobranie zakonov Rossiskoi Imperii (PSZ) (Complete collection of the laws of the
Russian Empire), three series: I: 1649–1825, 46 vv. Saint-Petersburg, 1830. II: 1825–1881,
55 vv., 1830–1884; III: 1881–1913; 33 vv. 1885–1916.
Pomeranz, Ken. The Great Divergence: China, Europe and the Making of the Modern World
Economy. Princeton, N.J.: Princeton University Press, 2001.
Ponting, Clive. A Green History of the World. London: Penguin, 1993.
 references 313

Poon, Pauline Pui-Ting. “The Well-Being of Purchased Female Domestic Servants (Mui
Tsai) in Hong Kong in the Early Twentieth Century.” In Campbell, Miers and Miller.
Eds. Children in Slavery Through the Ages: 152–65.
Popkin, Jeremy D. Press, Revolution, and Social Identities in France, 1830–1835. University
Park: Pennsylvania State University Press, 2002.
Porter, Andrew. Victorian Shipping, Business and Imperial Policy. Donald Currie, the Castle
Line and Southern Africa. Woodbridge: St. Martin’s Press, 1986.
Postel-Vinay, Gilles. “The Di-integration of Traditional Labour Markets in France. From
Agriculture and Industry to Agriculture or Industry.” In Grantham, MacKinnon, Labour
market: 64–83.
Pouchepadass, Jacques. Paysans de la plaine du Gange: croissance agricole et société dans le
district de Champaran (Bihar), 1860–1950. Paris: École française d’Extrême-Orient, 1989.
Prakash, Gyan. Bonded Histories: Genealogies of Labour Servitude in Colonial India. Cam-
bridge: Cambridge University Press, 1990.
Pratt, Edward E. Japan’s Protoindustrial Elite: The Economic Foundations of the Gōnō. Cam-
bridge, MA: Harvard University Press, 1999.
Pravilova, Ekaterina A. Zakonnost’ I prava lichnosti: administrativnaia iustitsiia v Rossii, vto-
raia polovina XIX v.-oktiabr’ 1917 (Legality and the rights of the person: administrative
justice in Russia, second half of the 19th century- to October 1917). Saint-Petersburg:
SZAGS, 2000.
Prothero, Iowerth. Artisans and Politics in Early Nineteenth Century London: John Gast and
His Times. Folkestone: Dawson, 1979.
Qi Biaojia. An Wu qinshen xigao (Draft Court Opinions from Cases Personally Tried as
Inspector of the Wu Region). In vol. 4 of Lidai panli pandu. Edited by Yang Yifan, and
others. Beijing: Zhongguo shehui kexue chubanshe, 2005.
Qinding da Qing huidian (Imperially Endorsed Collected Statutes of the Great Qing, dated
1899). Shanghai: Shangwu yinshuguan, 1908.
Qinding xuezheng quanshu (Imperially Endorsed Complete Guide for Education Commis-
sioners, dated 1814). Reprinted in vol. 16 of Qingdai ge buyuan zeli. Hong Kong: Fuchi
shuyuan, 2004.
Raeff, Marc. The Well-ordered Police State: Social and Institutional Change Through Law in
the German and Russia, 1600–1800. New Haven: Yale University Press, 1983.
Reddy, William, M. Money and Liberty in Modern Europe. Cambridge: Cambridge Univer-
sity Press, 1987.
Reid, Anthony (ed.). Slavery, Bondage and Dependency in South-East Asia. London: Pal-
grave MacMillan 1984.
Reid, Anthony. “The Decline of Slavery in Nineteenth-Century Indonesia.” In Klein.
Ed.Breaking the Chains: 64–82.
Renault, François. “The Structures of the Slave Trade in Central Africa in the 19th Century.”
In Clarence-Smith. Ed. Economics of the Indian Ocean Slave Trade: 146–65.
Reybaud, Louis. Etudes sur le régime des manufactures. Condition des ouvriers en soie. Paris:
Michel Lévy frères, 1859.
Richardson, Peter. “The Recruiting of Chinese Indentured Labour for the South African
Gold-Mines, 1903–1908.” Journal of African History 18,1 (1977): 85–108.
Richter, Karl. Geschichte des Niederlandes. Die Besiedlung. Sonthofen: Selbstverlag, 1960.
Ricks, Thomas M. “Slaves and Slave Traders in the Persian Gulf, 18th and 19th Centuries:
An Assessment.” In Clarence-Smith. Ed. Economics of the Indian Ocean Slave Trade:
60–70.
Robb, Peter. “Labour in India 1860–1920: Typologies, Change and Regulation.” Journal of the
Royal Asiatic Society, Third Series, 4,1 (1994): 37–66.
Roberts, Luke S. Mercantilism in a Japanese Domain: The Merchant Origins of Economic
Nationalism in 18th-Century Tosa. Cambridge, UK: Cambridge University Press, 1998.
Roberts, Richard and Zachuernuk Philip (eds). “On Slavery and Islam in African History:
A Tribute to Martin Klein.” Canadian Journal of African Studies, special issue: 34,3,
2000.
314 references

Rosanvallon, Pierre. Le modèle politique français: la société civile contre le jacobinisme de

1789 à nos jours. Points. Histoire. Paris: Le Seuil, 2006.
Ross, Robert. “The Dutch on the Swahili Coast, 1776–1779: Two Slaving Journals,” Interna-
tional Journal of African Historical Studies 19, 2 (1986): 304–60 and 19,3 (1986): 479–506.
——. “The Last Years of the Slave Trade to the Cape Colony.” In Clarence-Smith. Ed. Eco-
nomics of the Indian Ocean Slave Trade: 209–19.
Rowe, William T. Crimson Rain: Seven Centuries of Violence in a Chinese County. Stanford:
Stanford University Press, 2007.
Roy, Tirthankar. “Sardars, Jobbers, Kanganies: The Labour Contractor and Indian Eco-
nomic History.” Modern Asian Studies 42,5 (2008): 971–98.
——. The Economic History of India, 1857–1947. New York: Oxford University Press, 2001.
Royer, Jean-Pierre. Histoire de la justice en France. Paris: PUF, 2001.
Rude, Fernand. L’insurrection lyonnaise de novembre 1831. Le mouvement ouvrier à Lyon de
1827 à 1832. Paris: Anthropos, 1969.
——. “L’insurrection ouvrière de Lyon en 1831 et le rôle de Pierre Charnier.” La Révolution
de 1848 et les révolutions du XIXe siècle, 164 (1938): 18–49, 165 (1938): 65–117, 166 (1938):
140–179.
Rudert, Thomas. “Grenzüberschreitungen. Frühformen der Gutsherrschaft im mecklen-
burgisch-pommerschen Grenzgebiet im 16. Jahrhundert.” In. Peters, Jan (ed.). Gutsherr-
schaftsgesellschaften im europäischen Vergleich. Berlin: Akademie, 1997: 351–83.
——. Gutsherrschaft und Agrarstruktur. Der ländliche Bereich Mecklenburgs am Beginn des
18. Jahrhunderts. Frankfurt-on-Main: Lang, 1995.
Rudner, David West. Caste and Capitalism in Colonial India: The Natukottai Chettiars.
Berkeley, Los Angeles: University of California Press, 1994.
Rusiński, Władysław. “Some remarks on the differentiation of agrarian structure in East
Central Europe from the 16th to 18th century.” Studia historiae oeconomicae 13 (1978):
83–95.
Ryba, Jan. “Režijní hospodářství na toužimské panství ve druhé pol. 16. stol. a první
pol. 17. stol.” In Hojda, Zdeněk, Pešek, Jiří, and Žilynská, Blanka (ed.). Seminář a jeho
hosté. Praha: Univerzita Karlova, 1992: 217–25.
Ryndzinskii, Pavel. “Vymiralo li krepostnoe krest’ianstvo pered reformoi 1861 g.?” (Was
serfdom vanishing before the reform of 1861?). Voprosy istorii, 7 (1967): 54–70.
Sabel, Charles and Zeitlin, Jonhatan (eds). Worlds of Possibilities. Flexibility and Mass Pro-
duction in Western Industrialization. Cambridge, Paris: Maison des Sciences de l’Homme,
Cambridge University press, 1997.
Sabel, Charles, et Jonathan Zeitlin. “Historical Alternatives to Mass Production: Politics,
Markets and Technology in Nineteenth-Century Industrialization.” Past and Present, 108
(1985): 133–176.
Saito, Osamu. “The Labour Market in Tokugawa Japan: Wage Differentials and the Real
Wage Level, 1727, 1830”. Explorations in Economic History, 15, 1 (1978): 84–100.
——. “Wages, Inequality, and Pre-Industrial Growth in Japan, 1727–1894.” In Living Stan-
dards in the Past: New Perspectives on Well-Being in Asia and Europe. Edited by Robert C.
Allen, Tommy Bentsson and Martin Dribe, 77–98. Oxford, UK: Oxford University Press,
2005.
Salais, Robert, et Michael Storper. Les mondes de production: enquête sur l’identité écono-
mique de la France. Civilisations et sociétés. Paris: Éd. De l’École des hautes études en
sciences sociales, 1993.
Salman, Michael. “The meaning of slavery: The genealogy of ‘an Insult to the American
Government and to the Filipino People’.” In Campbell. Ed. Abolition and its Aftermath:
180–97.
Schackmann, Wil. De proefkolonie. Amsterdam: Mouria, 2006.
Scheltema, A.M.P.A. The Food Consumption of the Native Inhabitants of Java and Madura.
National Council for the Netherlands and the Netherlands Indies of the Institute of
Pacific Relations. Batavia: Ruygrok, s.a. [1936].
 references 315

Schilling, Renate. Schwedisch-Pommern um 1700. Studien zur Agrarstruktur eines Territori-

ums extremer Gutsherrschaft. Weimar: Böhlau, 1989.
Schleinert, Dirk. Die Gutswirtschaft im Herzogtum Pommern-Wolgast im 16. und frühen 17.
Jahrhundert. Cologne: Böhlau, 2001.
Schmidt, Christoph. Leibeigenschaft im Ostseeraum. Versuch einer Typologie. Cologne:
Böhlau, 1997.
——. “Über die Bezeichnung des stände (sostojanie-soslovie) in Russland seit dem. 18 Jahr-
hundert.” Jahrbücher für Geschichte Osteuropas 38 (1990): 199–21.
Schmiechen, James. Sweated Industries and Sweated Labour: The London Clothing Trades,
1860–1914. Urbana: University of Illinois Press, 1982.
Schottenhammer, Angela. “Slaves and Forms of Slavery in Late Imperial China (17th to
early 20th Century).” In Campbell. Ed. Structure of Slavery: 143–54.
Schrauwers, Albert, ‘The “Benevolent” Colonies of Johannes van den Bosch: Continuities in
the Administration of Poverty in the Netherlands and Indonesia.’ Comparative Studies in
Society and History 43,2 (2001): 298–328.
Schuman, Peter. Tussen vlag en voorschip. Een eeuw wettelijke en maatschappelijke emanci-
patie van zeevarenden ter Nederlandse koopvaardij 1838–1940. Amsterdam: De Bataafsche
Leeuw, 1995.
Scott, Tom (ed.). The Peasantries of Europe. From the Fourteenth to the Eighteenth Centuries.
London: Longman, 1998.
Scott, Tom. Society and economy in Germany, 1300–1600. Basingstoke: Palgrave, 2002.
Sen, Samita. ‘ “Without His Consent?”: Marriage and Women’s Migration in Colonial India.’
International Labor and Working Class History 65 (2004): 77–104.
Sergeeva, Nikolai I. “Obiazannye kret’iane i reforma 19 fevralia 1861 goda.” (The “obliged”
peasants and the reform of 1861). In Problemy obshchestvennoi mysli i ekonomicheskaia
politika Rossii XIX–XX vekov (Problems of social thought and political economy in Rus-
sia, nineteenth-twentieth centuries) Leningrad: Nauka, 1972.
Sewell, William Hamilton. Gens de métier et révolutions: le langage du travail, de l’Ancien
régime à 1848. Collection historique. Paris: Aubier-Montaigne, 1983.
Sharma, Sanjay. Famine, Philanthropy and the Colonial State: North India in the Early Nine-
teenth Century. New Delhi: Oxford University Press, 2001.
Shen Jiaben, and others, comp. Da Qing xianxing xinglü (The Penal Law of the Great Qing
Currently in Use). Beijing: Xianzheng biancha guan, 1910.
Sheriff, Abdul. “Localisation and Social Composition of the east African Slave Trade, 1858–
1873.” In Clarence-Smith. Ed. Economics of the Indian Ocean Slave Trade: 131–45.
——. “The Slave trade and its Fallout in the Persian Gulf.” In Campbell. Ed. Abolition and
its Aftermath: 103–19.
——. Slaves, Spices & Ivory in Zanzibar. London: James Currey, 1987.
Shinbo Hiroshi and Hasegawa Akira. “Shōhin seisan, ryūtsū no dainamikkusu” [The
dynamics of commodity production and distribution]. In Keizai shakai no seiritsu: 17–18
seiki [The development of commercial society: 17–18th centuries]. Edited by Hayami
Akira and Miyamoto Matao, 217–70. Tokyo: Iwanami Shoten, 1992.
Shirodkar, P.P. “India and Mozambique: Centuries Old interaction.” Purabhilekh-Puratatva
6, 1 (1988): 35–62.
Shlomowitz, Ralph. “Markets for Indentured and Time-Expired Melanesian Labour in
Queensland, 1863–1906: An Economic Analysis.” Journal of Pacific History 16, 2 (1981):
70–91.
——. “Mortality and Indentured Labour in Papua (1885–1941) and New Guinea (1920–
1941).” Journal of Pacific History 23, 1 (April 1988): 70–9.
Simitis, Spiros. “The case of the Employment Relationship: Elements of a comparison’, in
Steinmetz (ed.) Private Law and Social Inequality: 181–202.
Simon, Daphne. “Master and Servant”. In Saville, John (ed.). Democracy and the Labour
Movement: Essays in Honour of Dona Tor. London: Lawrence and Wishart, 1954: 160–200.
316 references

Sims, Vanessa Good Faith in Contract Law: A Comparative Analysis of English and German
Law Ph.D. Thesis, University of Cambridge, 2002: 85–86.
Singha, Radhika. A Despotism of Law: Crime and Justice in Early Colonial India. New Delhi:
Oxford University Press, 1998.
SIRIO (Sbornik Imperatorskogo Russkogo Istoricheskogo Obshchestva) (Collection of the
Imperial Russian Historical Society). Saint-Petersburg, 1872–1896. vol. 8, 93, 197 et 134.
Smissaert, H. De arbeidsovereenkomst der dienstboden. Haarlem: Tjeenk Willink, 1908.
Smith, R.E.F. Peasant Farming in Muscovy. Cambridge: Cambridge University Press: 1977.
Smith, Thomas C. Native Sources of Japanese Industrialization, 1750–1920. Berkeley and Los
Angeles: University of California Press, 1988.
Sollewijn Gelpke, J.H.F. Naar aanleiding van Staatsblad, 1878 No. 110. Batavia: Landsdrukkerij,
1901.
Solomon, Peter (ed.). Reforming Justice in Russia, 1864–1994, Power, Culture and the Limits
of Legal Order. Armonk: Sharpe, 1997.
Sommer, Matthew H. “The Penetrated Male in Late Imperial China: Judicial Construction
and Social Stigma.” Modern China 23, 2 (1997): 140–180.
Soom, Arnold. “Gutswirtschaft in Livland am Ausgang des 16. Jahrhunderts.” Zeitschrift für
Ostforschung 5 (1956): 34–57.
Souryi, Pierre François. The World Turned Upside Down: Medieval Japanese Society. Trans-
lated by Käthe Roth. New York, NY: Columbia University Press, 2001.
Spence, Jonathan D., and John E. Wills, eds. From Ming to Ch’ing: Conquest, Region, and
Continuity in Seventeenth Century China. New Haven, London: Yale University Press,
1979.
Špiesz, Anton. “Vývoj agrárnych poměrov v střednej a východnej Európě v novověku a
problém existencie druhého nevol’níctva na našom území.” Historický časopis 15 (1967):
512–58.
Sri Margana. “Hybridity, Colonial Capitalism and Indigenous Resistance: The Case of the
Paku Alam in Central Java.” In Sugarlandia Revisited. Sugar and Colonialism in Asia
and the Americas, 1800–1940’, edited by Ulbe Bosma, Juan Giusti-Cordero and G. Roger
Knight, (New York, Oxford: Berghahn 2007) 95–112.
Stanziani, Alessandro. “Free Labor-Forced Labor: an Uncertain Boundary? The Circula-
tion of Economic Ideas Between Russia and Europe From the Eighteenth to the Mid-
nineteenth Century”. Kritika: Explorations in Russian and Eurasian History, 9,1 (2008):
27–52.
——. L’économie en révolution, Le cas russe 1861–1930. Paris: Albin Michel, 1998.
Steinfeld, Robert J. Coercion, Contract and Free Labor in the Nineteenth Century. Cambridge:
Cambridge University Press, 2001.
——. The Invention of Free Labor. The Employment Relation in English and American Law
and Culture, 1350–1870. Chapel Hill & London: The University of North Carolina Press,
1991.
Steinmetz, Willibald (ed.). Private Law and Social Inequality in the Industrial Age. Com-
paring Legal Cultures in Britain, France, Germany and the United States. Oxford: Oxford
University Press, 2000.
Stevens, Carol Belkin. Soldiers of the Steppe. Dekalb: Northern Illinois University Press,
1995.
Stewart-McDougall, Mary Lynn. The Artisan Republic. Revolution, Reaction, and Résistance
in Lyon 1848–1851. Kingston et Montreal: McGill-Queen’s University Press and Alan
Sutton, 1984.
Stokvis, J.E. Koloniale Schande. Uitgave van het partijbestuur der S.D.A.P. Amsterdam,
1926.
Stoler, Ann Laura. Capitalism and confrontation in Sumatra’s plantation belt, 1870–1979.
New Haven [etc.]: Yale University Press, 1985.
Struve, Lynn A. Voices from the Ming-Qing Cataclysm: China in Tigers’ Jaws. New Haven,
London: Yale University Press, 1998.
 references 317

Stutzer, Dietmar. “Die Ertrags- und Lohnverhältnisse in den landwirtschaftlichen Betrie-

ben der Familie von Eichendorff in Oberschlesien um 1800.” Zeitschrift für Ostforschung
28, 1 (1979): 1–27.
Suermondt, W. and N.P. van den Berg. Nota over de suikerindustrie op Java, en de maatrege-
len die ter harer instandhoudinig vereischt worden. Batavia, 30 November 1886.
Sugihara, Kaouro. “Labour-Intensive Industrialisation in Global History”. Australian Eco-
nomic History, 47,2 (2007): 121–154.
Supiot, Alain. Critique du droit du travail. Paris: Presses Universitaires de France, 1994.
Sutherland, Heather Amanda. The making of a bureaucratic elite. The colonial transforma-
tion of the Javanese Priyayi. Singapore: Heinemann for the Asian Studies Association of
Australia, 1979.
Tanaka, Masatoshi. “Popular Uprisings, Rent Resistance, and Bondservant Rebellions in
the Late Ming.” In State and Society in China: Japanese Perspectives on Ming-Qing Social
and Economic History, edited by Grove, Linda and Daniels, Christian, 165–214. Tokyo:
Tokyo University Press, 1984.
Tawney, Robert. The Agrarian Problem in the Sixteenth Century, edited with an introduc-
tion by L. Stone. New York: Harper and Row, 1967.
Taylor, Romeyn. “Chinese Hierarchy in Comparative Perspective.” The Journal of Asian
Studies, 48,3 (1989): 490–511.
Terwiel, B.J. “Bondage and Slavery in Early Nineteenth Century Siam.” in Reid. Ed. Slavery,
Bondage and Dependency: 118–37.
Testart, Alain. L’esclave, la dette et le pouvoir. Paris: Éditions Errances, 2001.
The Muscovite Law Code (Ulozhenie) of 1649. Hellie, Richard, ed. and tr. (1988). Irvine, CA:
Charles Schlacks.
Thévenot, Laurent. “Les investissements de forme” Cahiers du Centre d’Etude de l’Emploi,
21–71. Paris: Presses Universitaires de France, 1986.
Thompson, Edward P. “Time, Work-Discipline and Industrial Capitalism.” Past and Present,
38 (1967): 56–97.
Thompson, F.M.L. “The Second Agricultural Revolution, 1815–1880”. Economic History
Review, 21 (1968): 62–77.
Thornton, John. Africa and Africans in the Making of the Atlantic World. Cambridge: Cam-
bridge University Press 1992.
Tideman, J. “De koelieordonnantie en hare toepassing.” Koloniale Studiën 2,2 (1918): 31–74.
Tinker,  Hugh. A New System of Slavery: The Export of Indian Labour Overseas, 1830–1920.
Oxford: Oxford University Press, 1974.
Tiray, Jan. “Jak se žilo na panství Teleckém za posledních pánů z Hradce (1589–1604).”
Časopis moravského muzea zemského 15 (1915): 49–66, 208–38.
Toby, Ronald. State and Diplomacy in Early Modern Japan. Princeton, NJ: Princeton Uni-
versity Press, 1984.
Toledano, Ehud R. State and Society in Mid-Nineteenth-Century Egypt. Cambridge: Cam-
bridge University Press, 1990.
Tomlinson, Brian R. The Economy of Modern India, 1860–1970. Cambridge: Cambridge Uni-
versity Press, 1996.
Totman, Conrad. The Collapse of the Tokugawa Bakufu, 1862–1868. Honolulu, HI: University
of Hawaii Press, 1980.
Toynbee, Arnold. Lectures on the Industrial Revolution in England, ed. (with an introduc-
tion by T.S. Ashton. Newton Abbott: A.M. Kelley, 1969) (originally published 1884).
Tracy, James. The Rise of Merchant Empires. Cambridge: Cambridge University Press, 1990.
Trzyna, Edward. Położenie ludności wiejskiej w królewszczyznach województwa krakowskiego
w XVII wieku. Wrocław: Zakład Narodowy im. Ossolińskich, 1963.
Turner, Mary. “The British Caribbean, 1823–1838: The Transition from Slave to Free Legal
Status”. In Hay and Craven (eds.). Masters: 303–322.
Turton, Andrew “Thai institutions of slavery.” In Georges Condominas. Ed. Formes extrê-
mes de dependence. Contributions à l’étude de l’esclavage en Asie du Sud-Est. Paris: École
des Hautes Études en Sciences Sociales, 1998: 411–57.
318 references

——. “Violent Capture of People for Exchange on Karen-Kai Borders in the 1830s.” In
Campbell. Ed. Structure of Slavery: 69–82.
Urbanized Economy Long Before France.” Economic History Review, 49,2 (1996): 213–248.
Urfus, Valentin. “Právní postavení námezdních pracovníků u nás v 16.-18. století.”
Pravněhistorické studie 2 (1956): 198–249.
Useful Arts and Manufactures vol. 3. London: Bernard Quaritch, 1885.
Vacek, František. “K dějinám selské roboty v Čechách.” Časopis pro dějiny venkova 23
(1936): 1–10, 57–69.
Válka, Josef. “Druhé nevolnictví a původní akumulace v 16. století.” Československý časopis
historický 6 (1958): 316–24.
——. “K současnému stavu otázky ‘druhého nevolnictví‘ ve středoevropských zemích.”
Časopis Matice moravské 73 (1954): 179–220.
——. Hospodářská politika feudálního velkostatku na předbělohorské Moravě. Brno: Státní
pedagogické nakladatelství, 1962.
Van den Bosch, J. “Nota van den gewezen kommissaris-generaal van de West-Indische
bezittingen, betrekkelijk een stelsel van kolonisatie voor Java [. . .].” In Het koloniaal
monopoliestelsel getoetst aan geschiedenis en staatshuishoudkunde, D.C. Steijn Parvé,
275. Zalt-Bommel: Joh. Noman en Zoon, 1851.
——. Nederlandsche bezittingen in Asia, Amerika en Afrika in derzelven toestand en aan-
gelegenheden van dit Rijk 2 vols.’s-Gravenhage en Amsterdam: Gebroeders van Cleef,
1818.
Van der Linden, Marcel. “Forced Labour and Non-Capitalist Industrialization: The Case of
Stalinism (c. 1929–c. 1956)”. In Free and Unfree Labour: The Debate Continues edited by
Tom Brass and Marcel van der Linden, 501–23, Berne: Peter Lang, 1997.
Van Geuns, M. De suikeronderneming Djatiroto en de regeling van het werkcontract op Java.
Overdruk uit: Soerabajasch Handelsblad van 6, 8, 11, 13, 15 en 18 aug. 1908. Soerabaja: sn, 1908.
Van Houten, S. De regtstoestand der werklieden in Nederland. Ontwerp van wet met Memorie
van Toelichting. ’s-Gravenhage: Martinus Nijhoff, 1870.
Veneziani, Bruno. “The Evolution of the Contract of Employment.” in Hepple Bob (ed.)
The Making: 31–72.
Vernus, Pierre. “Pour une histoire du syndicalisme patronal dans la soierie lyonnaise avant
1914: l’Association de la Fabrique lyonnaise. ” in Les organisations patronales. Une appro-
che locale (XIXe–Xxe siècles) directed by Pierre Vernus, 71–103. Cahiers du Centre Pierre
Léon 1. Lyon 2002.
Verslag van de arbeidscommissie betreffende de wettelijke vaststelling van minimumloonen
voor werknemers op Java en Madoera. Batavia: Kollf, s.a.
Verslag van de suiker-enquête commissie (1921). Soerabaja: Fuhri & Co., [1921].
Villermé, Louis-René. Tableau de l’état physique et moral des ouvriers employés dans les
manufactures de coton, de laine et de soie. Jules Renouard et Cie, 1840.
Vink, Markus. ‘ “The World’s Oldest Trade”: Dutch Slavery and Slave Trade in the Indian
Ocean in the Seventeenth Century.’ Journal of World History 14,2 ( June 2003): 131–77.
Vitalis, L. Opmerking omtrent den loop der suiker-industrie in den Nederlandsch O.I. archipel.
’s-Gravenhage: H.C. Susan, C.Hz, 1862.
Vogler, Günter. “Die Entwicklung der feudalen Arbeitsrente in Brandenburg vom 15. bis 18.
Jahrhundert. Eine Analyse für das kurmärkische Domänenamt Badingen.” Jahrbuch für
Wirtschaftsgeschichte 1 (1966): 142–74.
Voth, Hans-Joachim. “Time and Work in Eighteenth Century London”. The Journal of Eco-
nomic History, 58,1 (1998): 29–58.
Voznesenskii, Serguei V. Materialy dlia bibliografii po istorii narodov SSSR XVI–XVIIvv.
(Materials for bibliography of the history of soviet people, sixteenth-seventeenth cen-
tury). Leningrad: Nauka, 1933.
Wachowiak, Bogdan. Gospodarcze położenie chłopów w domenach Księstwa Szczecińskiego
w XVI i w pierwszej połowie XVII w. Szczecin: Państwowe wydawnictwo naukowe, 1967.
 references 319

——. Gospodarka folwarczna w domenach Księstwa Pomorskiego w XVI i na początku XVII

wieku. Warszawa: Neriton, 2005.
Wächter, Hans-Helmut. Ostpreußische Domänenvorwerke im 16. und 17. Jahrhundert. Würz-
burg: Holzner, 1958.
Wagner, William. Marriage, Property and Law in late Imperial Russia. Oxford: Oxford Uni-
versity Press, 1994.
Wajs, Hubert. Powinnośći feudalne chłopów na Mazowzu od XIV do początku XVI wieku
(w dobrach monarszych i kościelnych). Wrocław: Ossolineum, 1986.
Wakeman, Frederic Jr. The Great Enterprise: The Manchu Reconstruction of Imperial Order
in Seventeenth-Century China. Berkeley: University of California Press, 1986.
Wakita Osamu. “The Social and Economic Consequences of Unification.” Translated by
James L. McClain. In Early Modern Japan, vol. 4 of The Cambridge History of Japan,
edited by John Whitney Hall, 96–127. Cambridge, UK: Cambridge University Press,
1991.
Wallace, Daniel. “Grigorii Teplov and the Conception of Order: The Commission on Com-
merce and the Role of Merchants in Russia.” Canadian-American Slavic studies, 16 (1982):
410–432.
——. “The Merchantry and the Problem of Social Order in the Russian State: Catherine II’s
Commission on Commerce.” Slavonic and East European Review, 55 (1977): 185–203.
——. “The Merchants’ View of the Social Order in Russia As Revealed in the Town
Nakazy from Moskovskaia Guberniia to Catherine’s Legislative Commission.” Canadian-
american Slavic studies, 11 (1977): 503–522.
Waltner, Ann B. Getting an Heir: Adoption and the Construction of Kinship in Late Imperial
China. Honolulu: University of Hawaii Press, 1990.
Wang Mengji. Wang Mengji jiaxun (Family Admonitions of Wang Mengji). In Chen Hong-
mou, Wuzhong yigui (Five Sets of Bequeathed Rules, pref. dated 1742). In vol. 267 of Sibu
beiyao. Shanghai: Zhonghua shuju, 1936.
Warren, James Francis. “Chinese Prostitution in Singapore: Recruitment and Brothel
Organisation.” In Jaschok and Miers. Eds. Women and Chinese Patriarchy: 77–105.
——. “The Structure of Slavery in the Sulu Zone in the Late Eighteenth and Nineteenth
Centuries.” In Campbell. Ed. Structure of Slavery: 111–28.
Watson, James L. “Transactions in People: The Chinese Market in Slaves, Servants, and
Heirs.” in idem. Ed. Asian and African Systems of Slavery: 223–50.
——. Ed. Asian and African Systems of Slavery. Berkeley. University of California Press,
1980.
Weber, Matthias. Die schlesischen Polizei- und Landesordnungen der Frühen Neuzeit. Köln:
Böhlau, 1996.
Wertheim, W.F. “Planters tegen koelies: nieuwe studies over Sumatra’s Oostkust”, Bijdra-
gen tot de Taal-, Land,- en Volkenkunde 143, 2/3 (1987): 357–362.
Westra, Piet and Armstrong, James. Eds. Slave Trade with Madagascar. The Journals of the
Cape slaver Leijdsman, 1715. Cape Town: Africana Publishers, 2006.
Wiens, Mi Chu. “Masters and Bondservants: Peasant Rage in the Seventeenth Century.”
Ming Studies 8 (1979): 57–64.
Will, Pierre-Etienne. “L’histoire n’a pas de fin.” In La Chine et la démocratie, edited by Del-
mas-Marty, Mireille and Pierre-Etienne Will, 7–39. Paris: Fayard, 2007.
——. Official Handbooks and Anthologies of Imperial China: A Descriptive and Critical Bib-
liography. Forthcoming (As of 3 June 2010).
Williams, Edward T. “The Abolition of Slavery in China.” The American Journal of Interna-
tional Law 4,4 (1910): 794–805.
——, trans. “Report to the Throne of the Imperial Chinese Commission on Constitu-
tional Government Recommending the Abolition of Slavery, together with the Imperial
Rescript Approving the Report and Ten Regulations for its Enforcement.” The American
Journal of International Law 4, 4 (1910): 359–373.
320 references

Williams, Eric. Capitalism and Slavery. Chapel Hill: University Of North Carolina Press,
1944.
Williamson, Jeoffrey. “Why Was British Growth So Slow During the Industrial Revolu-
tion?”, The Journal of Economic History, 44,3 (1984): 687–712.
Wirtschafter, Elise Kimerling, Structures of Society. Imperial Russia’s “People of Various
Ranks”. Dekalb: Northern Illinois university press, 1994.
——. “Legal Identity and the Possession of Serfs in Imperial Russia.” Journal of Modern
History, 70,3 (1998): 561–587.
——. Social Identity in Imperial Russia. Dekalb: Northern Illinois University press, 1997.
Wong, Bin. China Transformed: Historical Change and the Limits of European Experience.
Ithaca et New York: Cornell University Press, 1997.
Woods, Donna. “The Operation of the Master and Servant Acts and Servant Acts in the
Black Country, 1858–1875”. Midland History, 7 (1982): 109–123.
Worden, Nigel. “Indian Ocean slavery and its demise in the Cape Colony.” In Campbell. Ed.
Abolition and its Aftermath: 29–49.
Worger, W.H. South Africa’s City of Diamonds: Mine Workers and Monopoly Capitalism in
Kimberley, 1867–1895. New Haven and London: Yale, 1987.
Woronoff, Denis. Histoire de l’industrie en France du XVIe siècle à nos jours. Paris: Le Seuil,
1994.
Wortman Richard. The Development of a Russian Legal Consciousness. Chicago: Chicago
University Press, 1976.
Wu Han. Dengxia ji (Under Lamp-Light). Beijing: Sanlian shudian, 1961.
Wu Yanhong. “Dui Mingdai beiyou falü diwei de kaocha” (An analysis of the legal sta-
tus of Inferiors and Juniors in the Ming). In Wan Ming shehui bianqian wenti yu yanjiu
(Researches and Questions regarding Social transformations in the Late Ming), edited
by Wan Ming, 364–402. Beijing: Shangwu yinshuguan, 2005.
Wu Yuancui. Linju manlu (Casual Records from Retirement in the Forest, dated 1607).
Reprinted in vol. 1172 of Xuxiu siku quanshu. Shanghai: Shanghai guji chubanshe, 1996.
Wu Zhenhan. “Mingdai de zhupu guanxi” (On the Relationship between Master and Bond-
servant). Shihuo yuekan 12 (1982).
Wyczański, Andrzej. Studia nad folwarkiem szlacheckim w Polsce w latach 1500–1580. War-
szawa: Państwowe wydawnictwo naukowe, 1960.
——. Studia nad gospodarką starostwa korczyńskiego 1500–1660. Warszawa: Państwowe
wydawnictwo naukowe, 1964.
Xiao Yong. Chishan huiyue (Agreement of the Chishan Association). In vol. 733 of Congshu
jicheng chubian. Shanghai: Shangwu yinshuguan, 1939.
Xie Guozhen. “Mingji Nubian kao” (An Examination of Late Ming Bondservant Rebellions).
In Xie Guozhen, Ming Qing zhiji dangshe yundong kao, 192–215. Shanghai: Shanghai
shudian, 2006 (1925).
Yagi Tetsuo. “Kinsei no shō kō gyōsha to toshi” [Early modern commercial crafts and cities].
In Shakai shi II [Social history] edited by Nakamura Yoshinao, 169–172. Tokyo: Yamakawa
Shuppansha, 1982.
Yang Dianzi, and others. Guangshan Xianzhi (Gazetteer of Guangshan District). Guang-
shan: Guangshan xianshu, 1889.
Yang Shi, comp. Er Cheng cuiyan (The Essential Words of the two Cheng Masters). In
vol. 621 of Congshu jicheng chubian. Shanghai: Shangwu yinshuguan, 1936.
Yang Yifan and Liu Ducai, comp. Zhongguo gudai difang falü wenxian (Ancient Local Legal
Materials from China). Beijing: Shijie tushu chuban gongsi, 2005.
Yilmaz, Gulay. “Becoming a Devşirme. The Training of Conscripted Children in the Otto-
man Empire” in Campbell, Miers and Miller (eds.). Children in Slavery through the Ages:
119–34.
Yu Chenglong. Yu Chenglong pandu jinghua (An Anthology of Judgments by Yu Cheng-
long). Reprinted in Qingdai mingli pandu qizhong huibian 3–80. Taipei: Laogu wenhua,
2000.
 references 321

Yunoki Manabu. “Kinsei Itami shuzō gyō no tenkai to Konishi ke” [Developments in the
sake brewing industry of early modern Itami and the Konishi family]. Edited by Itami
Shiritsu Hakubutsukan. Chiiki Kenkyū Itami 18 (1989): 1–23.
——. Sake Zukuri no Rekishi [The history of making sake]. Tokyo: Yūzankaku Shuppan,
1987.
Zaionchkovskii, Peter. Otmena krepostnogo prava v Rossii (The abolition of serfdom in
Russia). Third edit. Moscow: Nauka 1968.
Zeng Yuwang, Yiyou biji (Chronicle of 1645). Reprinted in Qingdai riji huichao, 3–36. Shang-
hai: Shanghai renmin chubanshe, 1982.
Zhang Chuanxi, comp. Zhongguo lidai qiyue huibian kaoshi (An Historical Anthology of
Chinese Contracts with Annotations and Explanations). Beijing: Beijing daxue chuban-
she, 1995.
Zhang Lüxiang, Yangyuan xiansheng quanji (Complete Works by Zhang Lüxiang, dated
1704). Beijing: Zhonghua shuju, 2002.
Zhang Mingbi. “Xiaobi ban ji” (Notes on the “Sliced Noses” society). In Zhang Mingbi, Ying-
zhi quanji, juan 4 (pref. 1898).
Zhuang Yuanchen, Zhuang Zhongfu zazhu (Various Writings by Zhuang Zhongfu). Rare
manuscript held at the National Library of Beijing.
Zimányi, Vera. Economy and Society in Sixteenth and Seventeenth Century Hungary (1526–
1650). Budapest: Akadémiai kiadó, 1987.
Zurndorfer, Harriet. “Servage domestique, conflits sociaux et la transition Ming-Qing dans
la préfecture de Hui-Chou, 1644–1646.” In Alessandro Stanziani (ed), Le travail contraint
en Asie et en Europe, Paris, FMSH, 2010: 241–264.
Żytkowicz, Leonid. “The Peasant’s Farm and the Landlord’s Farm in Poland from the 16th
to the Middle of the 18th Century.” Journal of European Economic History 1,1 (1972):
135–54.
——. “Trends of Agrarian Economy in Poland, Bohemia and Hungary from the Middle of
the Fifteenth to the Middle of the Seventeenth Century.” Mączak, Antoni, Samsonowicz,
Henryk, and Burke, Peter (eds.). East-Central Europe in Transition: From the Fourteenth
to the Seventeenth Century. Cambridge, New York, Paris: Cambridge University Press;
Editions de la Maison des sciences de l’Homme, 1985: 59–83.
——. Studia nad gospodarstwem wiejskim w dobrach kościelnych XVI w. Warszawa:
Państwowe wydawnictwo naukowe, 1962.
 INDEX

Abolition 5, 18, 21, 34, 36, 64, 67 n. 12, 68, Bengal 90, 102, 280–281, 283 nn. 40–41
78, 80, 81, 82, 85, 91, 91 n. 25, 92, 92 Betsileo 110
nn. 27–28, 93 n. 38, 99 n. 82, 104 n. 117, BGB 59, 60 n. 117
104, 105, 108, 112 n. 178, 183, 212, 215, 216, Black 35, 95, 311, 320
217 n. 19, 223, 237 n. 1, 282 n. 36, 292, Bonded 4, 7 n. 21, 19, 21, 24, 25, 68 n. 15,
309, 319 70 n. 20, 89, 93 n. 34, 95, 104, 109, 112, 313
Abolitionist Movement 89, 90, 93, 106 Bondservant 237, 240 n. 10, 242–243, 317,
Adoption 115, 118, 125, 128, 129, 142, 253 320
n. 59, 270, 310 319 Borneo 96
Aetas 103 Boys 54, 88, 96, 98, 99, 100, 101, 108, 118,
Afghanistan 98 126, 250, 261, 271 n. 99
Africa 2, 4, 7, 15, 17, 56–57, 58 n. 106, 70 Burma (Myanmar) 92, 97, 102, 104, 106,
n. 20, 87, 90 n. 15, 92, 93 n. 34, 94–99, 109, 185
101–104, 107 n. 141, 175, 177, 279 n. 27,
283, 293–295, 304–306, 308, 312, 317, 320 Cape Town 94–95
African 5, 15–16, 56, 68, 90 n. 15, 92 Caste 7, 21, 25, 101, 103, 104, 179, 182, 183,
nn. 26, 31, 93, 95–96, 97 nn. 64, 67, 98 185, 274 n. 7, 278 n. 21, 284, 284 n. 44,
n. 69, 105–106, 179, 279 n. 27, 296, 313, 314
319 Cattle 90 n. 15
Agriculture 1, 9, 9 nn. 27–28, 10, 10 n. 33, Ceylon (Sri Lanka) 106
11, 12, 14, 31, 35, 100, 122, 124, 141, 203, 204, Charity 229, 255, 258, 300
206, 227, 228, 235, 236, 295, 298, 300, Child 2, 2 n. 2, 14–15, 87, 88, 92 n. 30,
308, 311–312 97–110, 103 n. 112, 110, 141, 296, 303
Ambohibeloma 110 Child Labour 2 n. 2, 87–89, 296, 303
Ambositra 110 Children 2 n. 3, 12, 15, 20, 87–88, 92–94,
Anti-slavery 69, 92, 109, 290 96–97, 99–101, 99 n. 86, 103–105, 103
apprentice 15, 31, 35, 41, 104, 105, 106, 123, n. 113, 136, 293
126–127, 134, 138, 141, 147, 164, 165, 166, China 1, 4–6, 11–12, 17, 19–22, 25, 87–88,
170 n. 74, 171, 224, 238 n. 4, 242, 295 90, 94, 99, 101, 107, 129, 176–177, 237,
Apprenticeship 21, 36, 47, 71, 104, 124, 239–240, 244 n. 21, 245, 247, 250 n. 46,
126, 138, 141, 224 254, 255 n. 67, 260 n. 79, 265 n. 87, 269,
Arbeitskraft 58 270, 271, 292–293, 300–302, 304–305, 312,
Arbeitsverhältnis 60 315–317, 319–320
Armenian 99 Cholera 90
Artisans 31, 36 n. 15, 40, 41, 42, 48, 65, 66, Christian Filipinos 96 n. 56, 103
89, 118, 123, 126, 162 n. 48, 165, 176, 177, Circassian 99
216, 313 Code civil 58 n. 108, 59 n. 110
Asia 1–2, 9, 11, 16–18, 22, 24–25, 56, 82, Coercion 2, 3 n. 4, 11, 20, 65, 38 n. 29, 39
87, 92 n. 32, 93 nn. 34–35, 94, 96, 97 n. 33, 41 n. 40, 42 n. 42, 67, 68 n. 14, 76,
nn. 59–63, 65, 100–101, 102 nn. 105, 108, 83, 186, 200, 224, 235, 251, 273, 287 n. 54,
103 nn. 110–111, 104 n. 118, 107, 109 n. 164, 294, 316
111 n. 176, 177, 291–293, 295, 304–305, 310, collective bargaining 5, 31, 44, 59, 60–61
312, 314, 316, 318 Colonial 4, 6, 7, 24, 25, 31, 32, 56, 57, 63,
Australia 20, 106, 309, 317 64, 65, 67-70, 70 n. 20, 72, 74, 75, 76,
78 n. 55, 79, 80, 81, 82, 84, 85, 89, 90,
Behenjy 110 90 n. 28, 91, 92, 92 nn. 29, 94 n. 42, 101
Beijing 240, 255 n. 63, 257 n. 71, 299–302, n. 102, 103, 104, 104 nn. 118–119, 106
306–307, 309, 313, 315, 320–321 n. 132, 107 n. 141, 108 n. 155, 110, 112, 175,
324 index

176, 183, 184, 187, 274 n. 5, 275, 277, 283 Employers 3, 12, 15, 16, 21, 31, 31–62, 74,
n. 41, 284, 285, 287, 289–291, 294, 299, 77, 78, 80, 81, 82, 106, 108, 120, 123–126,
302, 304, 309, 312, 314–316 128, 134, 141, 142, 148, 171, 178–188, 273
Colonialism 2, 187, 306–307, 316 n. 3, 282, 286, 287
Colonies 1, 3, 4, 5, 8, 15, 17, 24, 56, 57, 66, Employment at will 43
67, 68, 69 n. 17, 72, 75, 77, 83, 84, 91, 103, Enlightenment 5, 7, 67, 103
108, 112, 175, 183, 184, 315 Ethiopia 97 n. 63, 100, 108
Compensation 31, 33, 33 n. 7, 38 n. 29,
44–45, 46, 47, 49, 52 n. 77, 57, 163, 200, Factory 40–41, 46, 52, 75, 76, 77, 78, 123,
223, 264, 264 n. 86 126, 151, 219, 224, 273, 287, 304
Confucian 95, 243, 254, 256, 258, 259 Famine 25, 90, 97, 103, 131, 255 n. 63, 273
n. 77, 271, 303, 305 n. 4, 275, 275 n. 10, 284, 285, 298, 301, 315
Contract 3, 4, 7, 20, 29, 31–65, 66, 65 n. 4, Fiji Islands 106
68, 68 n. 14, 70, 76, 77, 78, 80, 81, 83, Forced labour 2–3, 20, 22, 63–65, 68, 71,
84, 85, 89, 103, 104, 77 n. 48, 105, 106, 106 72, 108–110, 191–198, 200–203, 206–214,
n. 130, 120, 123, 124, 125, 131, 145–148, 150, 226, 277
151, 155, 171, 172, 175, 178, 179, 180–188, 211, Freedom of contract 34, 56
220, 224, 227, 229, 240, 240 n. 11, 242
nn. 16–17, 253, 253 n. 59, 255–257, Georgian 99
261–264, 280–281, 282, 285, 287, 291, 294, Gold 51 n. 75, 110 nn. 166, 168, 110 n. 170,
296, 305, 308–309, 315–316, 318 296
Contract of employment 29–32, 34–35, Gondar 100
40, 43–45, 54–56, 59–61 Gujarat 176
Corvée 5, 66, 69, 71, 72, 73, 75, 76, 108,
109 n. 158, 191, 192, 193, 195, 200, 202, 208, Hierarchy 16, 142, 154, 156, 240, 241, 243
209, 214 n. 20, 245 n. 25, 246, 255, 256, 257, 268,
Cotton 13, 72, 99, 118, 122, 130, 301 269, 277 n. 16, 317
Creole 106 Hindu 95, 176, 182, 183, 185, 187
Crime 36, 37, 38, 101, 216 n. 10, 230, 231, Historiography 5, 7, 8, 13, 106, 112, 200,
243, 244, 246 n. 29, 249, 251, 253, 257, 221, 233, 235, 238
270, 271, 299, 315 Homosexuality 98 n. 73, 100 n. 91, 250
Criminal 3, 18, 21, 23, 25, 38, 44, 57, 93, n. 46, 290
216, 223, 248, 249, 250, 252, 258, 278 Hong Kong 111, 178, 313
n. 20, 283 n. 41
Indentured 1, 3 nn. 5, 8, 4, 6, 20 nn. 78,
Debt 2, 4, 19, 21, 76, 93, 93 n. 36, 96 n. 55, 56, 81, 83, 84, 85, 88 n. 2, 91 n. 26, 101
97, 100, 101–105, 101 n. 101, 112, 118, 120, n. 96, 105, 106 nn. 134–136, 140, 107
125, 148, 149, 155, 186, 187, 217, 220, 230, nn. 144–147, 108 n. 150, 111 n. 174, 175,
242, 280, 285 175 n. 1, 185 n. 13, 311
Demesne, 191–214 Industrial Revolution 9, 11 n. 39, 12–5,
Dyaks 96 17, 35, 89, 293, 296, 300–301, 306, 309,
311–312, 317, 320
East Indies 66, 69, 80, 84, 91, 106, 178 n. 6 industrialization 10 n. 35, 13 n. 50, 145,
Education 69, 182, 228, 247, 249, 259, 215 n. 6, 273, 308, 311
299, 313
Egypt 88, 89 n. 13, 109, 111, 112 n. 177, Japan 1, 5–6, 8, 11–12, 14–15, 17–18, 21–24,
176–179, 180 n. 9, 181 n. 10, 183, 186, 188 89, 115, 118–120, 122–126, 128–129, 132–133,
Elite 24, 25, 65, 73, 75, 84, 90, 91, 97, 98, 140–142, 177, 295–296, 298–301, 303, 310,
112, 124, 216, 217, 218, 220, 227, 230, 236, 313–314, 317, 319
241 n. 11, 275, 313, 123 n.18 Java 63–64, 66, 69–72, 74–79, 82, 84–85,
Emancipation 8, 79, 91, 92 nn. 32–33, 119, 102, 109, 292–299, 302, 304, 307, 309, 314,
213, 220, 223, 226, 227, 230, 231, 233, 234, 316, 318
235, 237, 237 n. 1, 249, 249 n. 43, 264,
266, 282 n. 36, 298, 300–301, 303–304 Kinship 97, 244, 253, 254, 319
 index 325

Labour rents 191–192, 194–196, 200, New Guinea 106

206–207, 209, 212–213 Noble 4, 196, 199, 206, 207, 207 n. 64, 208,
Labourers 4, 5, 12, 21, 24, 25, 31, 33, 35, 36, 212, 216–233
37, 37 n. 19, 38, 39 n. 32, 42, 46, 49, 52, North America 5, 29, 61, 89, 177
64, 70, 72, 74, 77, 78, 80, 83, 106, 107,
111, 185, 194, 198, 199 n. 37, 201, 203, 203 Ottoman Empire 4, 98–99, 108, 178
n. 49, 207, 208, 274, 281, 283, 285, 286
Labour-intensive 1, 8–11 Pakistan 98, 176
Leninist 238 Pawnship 15, 15 n. 63, 69, 93, 102, 104, 306
Pemba 94
Mahajanga 110 Penal 57, 63–65 , 74, 77, 78, 78 n. 52, 79,
Majapahit 102 80, 81–85, 93, 108, 156, 164, 166, 168, 186,
Malay 106 223, 237, 243, 260, 263, 273 n. 3, 280, 283,
Malaya 101, 106, 175, 185 283 n. 41, 285, 285 n. 46, 287, 315
Managerial prerogative 44, 56, 58 Persia 98
Manchu 240, 241 n. 11, 255, 319 Persian Gulf 96, 99–100
Marxist 9, 74, 191, 238, 240 n. 11 Philippines 92, 102, 184 n. 12
Mascarenes 88, 94–95 Plantations 16, 56, 57, 63–65, 67 n. 12, 76,
Master 3, 4, 5, 15, 25, 31–34, 36–38, 39 80–85, 95, 99, 104, 106, 106 nn. 130–131,
n. 30, 40, 42–43, 44, 45, 48 n. 56, 49, 50, 107, 175, 186, 279, 279 n. 26, 285, 291
51, 55–57, 77, 93, 97, 98, 100, 102, 103, 104, Poor law 31, 35, 41, 51, 55, 69
106, 118, 124, 147–148, 155, 157, 161, 172, Population 9, 12 n. 47, 14, 20–21, 25, 29,
179–181, 187, 216, 217, 224, 225, 230, 237, 45, 56, 63, 66, 69, 72, 84, 85, 89, 90, 98,
240, 240 nn. 11–12, 242, 242 n. 16, 243 101, 102, 109, 110, 115–118, 119 n. 6, 122,
n. 20, 244, 246, 253, 254–270 296, 315, 125 n. 24, 127, 129, 130, 131, 133, 134, 135,
320 137–140, 145, 176, 182, 183, 191, 193, 195,
Master XE “Master” and Servant Acts  196, 209–211, 215 n. 3, 218, 224, 225, 232,
36, 37, 40, 42–43, 46–49, 50–57, 180, 187, 245, 270, 295, 298, 302, 305, 308
320 Port Louis 105
Mauritius 91, 94, 99, 104–105, 107–108, Poverty 67–68, 122, 160, 242, 252, 262
175, 185 n. 13, 186, 309 Property 23, 76, 105, 109 n. 158, 193, 203,
Mechanisation 9 216, 217, 224, 235, 241 n. 13, 248, 253, 261,
Merchant 35, 37, 41, 66, 97, 101, 118, 262, 266–269, 279, 307, 319
145–149, 150–164, 166–173, 176 n. 2, 177, Proto-industry 1, 10, 11, 18, 19, 22, 23, 24,
179, 182–184, 186, 188, 218–220, 222–227, 37 n. 31, 122, 124, 128, 141, 212, 308
233, 236, 245, 307, 313, 317 Punishment 21, 23, 39, 81, 83, 101, 216, 217,
Merina 108, 109, 111 223, 242, 249, 251, 252, 260, 270, 283, 283
Middle East 87, 92, 94, 100–101, 176 n. 41, 284, 284 n. 44, 299
Military 48, 70, 110, 111, 217, 217 n. 15,
229–231, 277, 278 n. 21, 280–82, 283 nn. Qing 19, 237, 238 n. 3, 240, 245 n. 27,
39–40, 284, 301–312 246 n. 28, 249–250, 265, 271, 292–293,
Ming 19, 237–241, 242 n. 15, 243–245, 246 295–296, 304, 309, 313, 315–317, 320
n. 28, 247–252, 253 nn. 58–59, 254–255, Queensland 106–107
257–262, 265–266, 268 n. 89, 269,
270–272, 293, 295–296, 298–300, Reciprocity 36, 40, 253–258, 260, 263,
303–304, 306, 308–309, 312, 316–317, 269, 272, 285
319–320 Réunion 90 n. 14, 94, 107 n. 143
Mobility 1, 4, 16, 21, 23, 104, 184, 185, 186,
188, 194, 199, 217, 221, 233, 234, 250 Seasonal 11, 12, 51–53, 122, 152, 154, 208
nn. 43–44, 280, 280 n. 28, 282, 286, 287, Seigniorial 193, 196, 197, 199, 200, 202, 212,
289, 302 214, 228, 233, 236
Mortality rates 90, 94, 97, 108, 278, 280, Self-employment 31–33, 54, 54 n. 94, 56
280 n. 30, 287 Servant 1, 3, 4, 5, 6, 7, 16, 21, 25, 31–59, 64,
Muslim 92–93, 96 n. 56, 98, 176, 187 66, 68, 69, 73, 77, 77 n. 49, 78, 79, 80, 84,
326 index

96–98, 100, 111, 118, 119, 120, 123–128, 133, Thailand 88, 101–102, 104, 108, 185
135–137, 139, 142, 178, 179–183, 185, 186 Traders 7, 14, 16, 66, 94, 99 n. 82, 101, 145,
187, 194, 198–200, 202–210, 219, 224, 227, 148, 154, 175, 177, 179, 307
229, 230, 237–260, 296 Trading 5, 24 n. 93, 25, 72, 74, 88 n. 5, 91,
Servitude 3 n. 5, 7, 19, 63, 67, 83, 87, 92 99 n. 86, 175–177, 183, 295
n. 32, 93, 97 n. 65, 101 n. 103, 102, 104 Turkestan 98
nn. 115, 118, 107 n. 141, 112 n. 178, 113,
187, 189, 237, 238, 238 n. 4, 240, 241, 242, Unemployment 34, 45, 48, 53–54, 67,
248, 251, 253, 258, 259, 260, 269–272, 155, 186
299, 312 United States 3, 10, 20, 35, 65, 68 n. 14,
Sexual 98–100, 111–112, 241, 250–251 89, 95, 184, 233, 295, 316
Sindworkies 177–178, 183–187
Slave, slave trade 23, 64, 66, 72, 88 n. 5, Vietnam 102
89, 91 n. 23, 92 nn. 31, 33, 93 nn. 34, 39, Village 22, 23, 25, 26, 65, 70, 73, 74 n. 35,
94 nn. 42–43, 45, 95 nn. 46–48, 51–52, 75, 76, 84, 107, 111, 115–117, 121 n. 14,
92–96, 96 n. 53, 97 nn. 61–63, 65, 98, 99 122–125, 125 n. 23, 127–129, 129 n. 32,
nn. 82, 86, 100 nn. 87–88, 101–103, 107 140–141, 143, 191–193, 193 n. 3, 194–195,
n. 141, 108–112, 119, 120, 124, 222, 237, 195 n. 8, 196–197, 197 n. 21, 198 n. 24, 201,
247, 248, 270, 282 n. 36, 292, 297, 202, 208, 210, 211 n. 76, 212 n. 84, 219, 231,
317–318 235, 236, 275, 276, 298, 300, 302, 304, 310
Slavery 2, 2 n. 1, 4 n. 11, 3–5, 5 nn. 12,
6–8, 14–19, 20 n. 74, 21, 22, 24, 56, 64, Wage 3, 10, 31, 33–35, 36, 39, 39 n. 32, 42,
66, 67, 67 n. 12, 68, 69, 70 n. 20, 81, 82, 45, 46, 48 n. 56, 50, 58, 74, 78, 79, 83, 120,
88, 89 n. 9, 91, 92, 93 nn. 34, 36–39, 94 121, 125, 147, 153, 158, 159, 161, 163, 173,
nn. 41–42, 96 n. 55, 97 nn. 63–65, 67, 180, 181, 186–188, 193, 194, 198–200, 203,
98 nn. 69, 71, 99 n. 80, 99 nn. 82–83, 211, 273–275, 277, 280, 282, n. 37, 286,
100 nn. 88, 90, 94, 101 n. 102, 102 n. 109, 293, 302, 314
104 nn. 117–119, 105 n. 127, 107 n. 143, 108 Wage labour 4–6, 8, 20, 59, 65–66, 68,
n. 155, 109 nn. 158, 160, 162, 119 nn. 6–7, 74–75, 80, 84–85, 89, 91, 106, 192, 194,
237, 238, 239, 246, 272, 273, 282, 292–293, 197–203, 207–208, 210–214
297–299, 304–306, 308, 310, 312–313, 315, Warfare 298
317–319 Women 16, 48, 65, 73, 85, 93, 97, 98, 107,
Social insurance 31, 33–34, 45, 52, 108, 111 n. 174, 112, 122, 126, 127, 134, 138,
54–56 140, 147, 156, 224 n. 56, 244, 252 n. 51,
Social security 44, 55, 61 259, 279 n. 27, 284 n. 45, 290, 299, 304,
Somalia 92, 94 307, 310, 315
Status 2, 3, 4, 5, 15, 16, 18, 19, 20, 21, 25, Work 3, 6, 8, 14 n. 57, 16, 19, 21, 23, 24, 25,
31–34, 42–45, 49–53, 54–57, 60, 69, 72, 29, 31–33, 35, 36, 37, 39, 40, 41–43, 49, 55,
80, 82, 88, 97, 99, 101, 103, 104, 109, 117, 57, 59–61, 63, 67, 69–71, 76–78, 81, 83–85,
118, 119, 148, 161, 172, 178, 184, 215, 216, 85 n. 75, 104, 105, 108, 110, 147–149, 151,
219, 222–236, 237–262, 265–272, 298, 153–156, 158, 160, 166, 168, 169, 171, 173,
317 180, 181, 188, 192, 196, 197, 199, 202–211,
Sugar 63, 71, 74–80, 84, 85, 91, 99, 106, 212 n. 85, 213, 221, 226, 239, 257, 273
107, 175, 186, 282 n. 36, 297, 304, 316 n. 1, 274, 276, 283, 290–291, 296, 302,
Sulu 94 n. 42, 96 n. 56, 98 306, 310, 317–318
Sumatra 64–65, 80–85, 96, 106, 292, 311,
316, 319 Zanzibar 94–95, 108