bradley, keith. the cambridge world history of slavery

8/9/2019 Bradley, Keith. the Cambridge World History of Slavery

1/753


2/753

TH E CAMB R IDGE WOR LD H ISTOR Y OF SLAVERY

Volume : a d – a d

Most societies in the past have had slaves, and almost all peoples have at sometime in their pasts been both slaves and owners of slaves. Recent decades have seena significant increase in our understanding of the historical role played by slavery and wide interest across a range of academic disciplines in the evolution of theinstitution. Exciting and innovative research methodologies have been developed,and numerous fruitful debates generated. Further, the study of slavery has cometo provide strong connections between academic research and the wider publicinterest at a time when such links have in general been weak. The Cambridge World History of Slavery responds to these trends by providing for the first time, in fourvolumes, a comprehensive global history of this widespread phenomenon fromthe ancient world to the present day.

Volume of The Cambridge World History of Slavery is a collection of essaysexploring the various manifestations of coerced labor in Africa, Asia, and the

Americas between the opening up of the Atlantic world and the formal creation of the new nation of Haiti. The authors, well-known authorities in their respectivefields, place slavery in the foreground of the collection but also examine othertypes of coerced labor. Essays are organized both nationally and thematically andcover the major empires, coerced migration, slave resistance, gender, demography,

law, and the economic significance of coerced labor. Nonscholars will also findthis volume accessible.

David Eltis is Robert W. Woodruff Professor of History at Emory University and research associate of the W. E. B. Du Bois Institute, Harvard University. Hehas also held visiting appointments at Harvard, Yale, and Oxford universities.Eltis received his PhD from the University of Rochester in . He is author of The Rise of African Slavery in the Americas , co-author (with David Richardson)of Atlas of the Transatlantic Slave Trade , and co-compiler of Slave Voyages at

www.slavevoyages.org. He co-edited and contributed to Extending the Frontiers:Essays on the New Transatlantic Slave Trade Database (with David Richardson)and Slavery in the Development of the Americas (with Frank D. Lewis and KennethL. Sokoloff) and edited Coerced and Free Migrations: Global Perspectives .

Stanley L. Engerman is John H. Munro Professor of Economics and Professor of History at the University of Rochester. He has also previously taught at Harvard,

Yale, Oxford, and Cambridge universities. Engerman received his PhD in eco-nomics from Johns Hopkins University in . He is the author of Slavery,Emancipation, and Freedom: Comparative Perspectives and the co-author of Time on the Cross: The Economics of American Negro Slavery (with Robert Fogel) andNaval Blockades in Peace and War: An Economic History Since 1750 (with Lance E.Davis). He is also co-editor of A Historical Guide to World Slavery (with SeymourDrescher); Finance, Intermediaries, and Economic Development (with Philip T.Hoffman, Jean-Laurent Rosenthal, and Kenneth L. Sokoloff); and The Cambridge Economic History of the United States (with Robert E. Gallman).

Cambridge Histories Online © Cambridge University Press, 2011


3/753

THE CAMBR ID GE WOR LD HISTOR Y OF SLAV ERY

General editors

David Eltis, Emory University

Stanley L. Engerman, University of Rochester

Volume I: The Ancient Mediterranean World

Edited by Keith Bradley and Paul Cartledge

Volume II : ad 500–ad 1420

Edited by David Eltis and Stanley L. Engerman

Volume III : ad 1420–ad 1804Edited by David Eltis and Stanley L. Engerman

Volume IV : ad 1804–ad 2000

Edited by Seymour Drescher, David Eltis, and Stanley L. Engerman



4/753

T H E C A M B R I D G E

WORL D H IS TO RY OFSLAVERY

VOLUME

ad 1420–ad 1804

Edited by

D A V I D E L T I S

Emory University

S T A N L E Y L . E N G E R M A N

University of Rochester



5/753

c a m b r i d g e u n i v e r s i t y p r e s s

Cambridge, New York, Melbourne, Madrid, Cape Town,Singapore, São Paulo, Delhi, Tokyo, Mexico City

Cambridge University Press Avenue of the Americas, New York, n y -, u s a

www.cambridge.org Information on this title: www.cambridge.org/

C Cambridge University Press

This publication is in copyright. Subject to statutory exceptionand to the provisions of relevant collective licensing agreements,no reproduction of any part may take place without the written

permission of Cambridge University Press.

First published

Printed in the United States of America

A catalog record for this publication is available from the British Library.

Library of Congress Cataloging in Publication data

Revised for volume The Cambridge world history of slavery / edited by David Eltis and Stanley L. Engerman.

p. cm.Includes bibliographical references and index.

isbn ---- (hardback). Slavery – History. I. Eltis, DavidII. Engerman, Stanley L. III. Title.

ht861.c

.–dc

isbn ---- Hardback

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party Internet Web sites referred to in this publication and does not

guarantee that any content on such Web sites is, or will remain, accurate or appropriate.



6/753

CONTENTS

List of Maps, Figures, and Tables page ix Contributors xiSeries Editors’ Introduction xiii

Dependence, Servility, and Coerced Labor in Time and Space d a v i d e l t i s a n d s t a n le y l . e n g e r m a n

p a r t i : s l a v er y i n a f r i c a a n d a s i a m i n or

Enslavement in the Ottoman Empire in the Early ModernPeriod e h u d r . t o l e d an o

Slavery in Islamic Africa, – r u d o l p h t . w a r e i i i

Slavery in Non-Islamic West Africa, – g . u g o n w o k ej i

Slaving and Resistance to Slaving in West Central Africa r o q u i n a l d o f e r r e i r a

White Servitude w i l l i a m g . c l a r e n c e - s m i t h a n d d a v i d e l t i s

p a r t i i : s l a v e r y i n a s i a

Slavery in Southeast Asia, – k e r r y w a r d

Slavery in Early Modern China p a m e l a k y l e c r o s s l e y

v



7/753

vi contents

p a r t i i i : s l a v er y a m o n g t h e i n d i ge n o u s a m e r i c a n s

Slavery in Indigenous North America

l e l a n d d o n a l d

Indigenous Slavery in South America, – n e i l l . w h i t e h e a d

p a r t i v : s l a v er y a n d s e r f d o m i n e a s te r n e u r o p e

Russian Slavery and Serfdom, – r i c h a r d h e l l i e

Manorialism and Rural Subjection in East Central Europe,– e d g a r m e l t o n

p a r t v : s l a v e r y i n t h e a m e r i c a s

Slavery in the Atlantic Islands and the Early Modern Spanish Atlantic World

w i l l i a m d . p h i l l i p s , j r .

Slavery and Politics in Colonial Portuguese America: TheSixteenth to the Eighteenth Centuries

j o ˜ a o f r a g o s o a n d a n a r i o s

Slavery in the British Caribbean p h i l i p d . m o r g a n

Slavery in the North American Mainland Colonies l o r e na s . w a l s h

Slavery in the French Caribbean, – l a u r e n t d u b o i s

Slavery and the Slave Trade of the Minor Atlantic Powers p i e t e r e m m e r



8/753

contents vii

p a r t v i : c u l t u ra l a n d d e m o g r ap h i c p a t t e r ns i nt h e a m e r i c a s

Demography and Family Structures

b . w . h i g m a n

The Concept of Creolization r i c h a r d p r i c e

Black Women in the Early Americas b e t t y w o o d

p a r t v i i : l e g a l s t r u ct u r e s , e c o no m i c s , a n d t h e

m o v e me n t o f c o e r ce d p e o p l e s i n t h e a t l a n t i c w o r l d

Involuntary Migration in the Early Modern World,– d a v i d r i c h a r d s o n

Slavery, Freedom, and the Law in the Atlantic World,–

s u e p e a b o d y

European Forced Labor in the Early Modern Era t i m o t h y c o a t e s

Transatlantic Slavery and Economic Development in the Atlantic World: West Africa, – j o s e p h e . i n i k o r i

p a r t v i i i : s l a v er y a n d r e s i s t a nc e

Slave Worker Rebellions and Revolution in the Americasto m a r y t u r n e r

Runaways and Quilombolas in the Americas m a n o lo fl o r e nt i n o a n d m ´ a r c i a a ma n t i n o

Index



9/753



10/753

LIST OF MAPS, FIGURES, AND TABLES

maps

. Culture Areas of Indigenous North America page

. East Central Europe, ca. . Brazil, Eighteenth Century . Portuguese Empire in America, Eighteenth Century

figures

. Sambabaia Quilombo . River of Perdition Quilombo . Quilombo on a Tributary of the Perdition River

. Ambrozio Quilombo . Sam Gonçalo Quilombo

tables

. Debts to be collected by the postmaster of Cuyuni . Distribution of registered slave baptisms: São Gonçalo,

– . Slave populations of the British Caribbean, –

. Annual percentage decline (and increase) in the slavepopulations of the British Caribbean, –

. Imports of slaves in Dutch Brazil, –, by Africanregion of origin

. Suriname’s trade balance/balance of payments, –,average per year

. The Dutch slave trade, – . Distribution of slave departures from Africa on Danish vessels . The African origins of Suriname slaves

. Involuntary migration in the Old World, –,estimates and projections

ix



11/753

x list of maps, figures, and tables

. Africans and whites taken to the Americas, –, by subperiods

. National participation in transatlantic slave trade, – . Numbers of slaves shipped by African region of departure, all

carriers, – . Merchandise carried to the African Coast by the Mary in . Cowries carried to the Gold Coast from Britain, –

(three-year averages in tons) . Distribution of commodities carried to the Bight of Benin

from Britain, select years, – . Distribution of commodities carried to the Bight of Biafra

from Britain, select years, – . Distribution of commodities carried to the Bights of Benin

and Biafra from Britain, select years, – . Demographic profile of slaves in Taubaté (–) and

Rio de Janeiro (–) . Demographic profiles of escaped slaves advertised in

newspapers in the Caribbean and the southern United States(–)

. Population estimates of some Minas Gerais quilombos (–)



12/753

CONTRIBUTORS

Márcia Amantino, Department of History, Universidade Federal do Riode Janeiro, Brazil

William G. Clarence-Smith, Department of History, School of Orientaland African Studies, UK

Timothy Coates, Department of History, College of Charleston, USA

Pamela Kyle Crossley, Department of History, Dartmouth College, USA

Leland Donald, University of Victoria, British Columbia, Canada

Laurent Dubois, Department of History, Duke University, USA

David Eltis, Department of History, Emory University, USA

Pieter Emmer, Department of History, Leiden University, Netherlands

Stanley L. Engerman, Departments of Economics and History, University of Rochester, USA

Roquinaldo Ferreira, Department of History, University of Virginia, USA

Manolo Florentino, Department of History, Universidade Federal do Riode Janeiro, Brazil

João Fragoso, Department of History, Universidade Federal do Rio de Janeiro, Brazil

Richard Hellie, University of Chicago, USA (deceased)

B. W. Higman, Department of History, Australian National University, Australia

Joseph E. Inikori, Department of History, University of Rochester, USA

Edgar Melton, Department of History, Wright State University, USA

Philip D. Morgan, Department of History, Johns Hopkins University, USA

xi



13/753

xii contributors

G. Ugo Nwokeji, Department of African American Studies, University of California, Berkeley, USA

Sue Peabody, Department of History, Washington State University, USA

William D. Phillips, Jr., Department of History, University of Minnesota,USA

Richard Price, Department of Anthropology, College of William and Mary,USA

David Richardson, Wilberforce Institute for the Study of Slavery andEmancipation, University of Hull, UK

Ana Rios, Department of History, Universidade Federal do Rio de Janeiro,

BrazilEhud R. Toledano, Department of History, Tel Aviv University, Israel

Mary Turner, Institute of Commonwealth Studies, University of London,UK

Lorena S. Walsh, Colonial Williamsburg Foundation, USA (retired)

Kerry Ward, Department of History, University of Michigan, USA

Rudolph T. Ware III, Department of History, University of Michigan,

USA

Neil L. Whitehead, Department of Anthropology, University of Wisconsin,USA

Betty Wood, Department of History, Cambridge University, UK



14/753

SERIES EDITORS’ INTRODUCTION

This is the third volume of The Cambridge World History of Slavery , explor-ing the various manifestations of coerced labor in Africa, Asia, and the

Americas between the opening up of the Atlantic world and the formal

creation of the new nation of Haiti. Slavery has been among the mostubiquitous of all human institutions, across time and place, from earliesthistory until, some would argue, the present day. Yet its durability andubiquity are not widely recognised and, where they are, they seem poorly understood by the general public and scholars alike. A central aim of thesevolumes, which cover many different times and places, is to help to placethe existence and nature of slavery against the backdrop of the broaderhuman social condition.

Slavery has appeared in many different forms and is not always easy to

separate from other forms of coerced labor. Nevertheless, there are basicsimilarities that emerge from the contributions that follow. Most criticalof these is the ownership of one human by another, and the ability tobuy and sell the human chattel such ownership creates. A second commoncharacteristic is the fact that chattel status is a heritable condition passeddown through the mother. Such characteristics are not to be found in themore general category of ‘coerced labor’, as normally practiced. The lattertypically involves a general loss of citizenship rights, but not necessarily ownership of one person by another and inherited status. Some scholars

regard slavery as part of a spectrum of coerced labor and dependency, butthe institution has maintained a distinctive legal existence in almost allsocieties.

xiii



15/753



16/753

CHAPTER 1

DEPENDENCE, SERVILITY, AND COERCED

LABOR IN TIME AND SPACE

david eltis and stanley l. engerman

Slavery is generally regarded as the most extreme form of dependency and exploitation. This project attempts to cover types of dependency inaddition to slavery, although it is clear from both the overall title and

the program for the project’s third volume that slavery gets considerably more attention than do other types of dependency. This reflects in part themodern preoccupation with individual freedom and equality before the law accorded by citizenship now acknowledged, at least as an ideal, just abouteverywhere in the modern world. Slavery may not be completely eradicatedtoday, but it had lost irrevocably the ideological struggle perhaps as early asthe first half of the nineteenth century, with only minor rearguard actions(in ideological terms, that is) in the antebellum South and less certainly inHitler’s Germany and the Soviet gulags. Such a circumstance – amazing

in its rapidity and completeness from a worldwide historical perspectiveof human behavior and beliefs – is taken for granted today. The morecomplete the victory of the view that slavery should not exist nor shouldhave ever existed, the more remote slavery itself appears, but at the sametime the greater the modern fascination with the institution becomes.

And the more remote it appears, the easier it is to treat slavery simply as an evil practiced by evil men, and the harder it is to understand it inhuman terms. At the very least, modern preoccupations with freedom andindividual rights drive the fascination with slavery. This phenomenon, an

outcome of the Enlightenment, shapes the form of the modern assault onslavery.

General explanations of the rise and fall of slavery have not fared wellin recent years, as the great resources thrown into the study of slavery from primary sources have revealed the richness and complexity of theinstitution. As this suggests, such explanations tend to date from an era predating our present age of extensive empirical research, and for the mostpart focus on slavery – or rather separate slavery from other forms of dependency – counter to what we wish to do. Such explanations are quite

good at describing how slavery functions but are weakest at accounting forfirst, its rise, second, its fall, and third, why at times nonslave dependency (for instance, serfdom) emerges as more important than chattel slavery.



17/753

the cambridge world history of slavery

Most important of all, perhaps, they fail to explain the eligibility issue – inother words, why certain peoples are seen as qualifying for slavery (whereasothers are not), and why this changes over time. This last issue has becomeof much greater interest in the last decade or so, as the realization spreads

that all peoples in the world have been at some time in their history both slaves and owners of slaves, often at one and the same time. Having dismissed general theories, we will nevertheless mention three of them hereas sometimes helpful. There is the general Marxist position, implicit in the

work of those who followed Marx, if not Marx himself, who had little tosay on the subject, which in broad terms takes the position that any ruling class would wish to impose slavelike conditions on the rest of society andis prevented from doing so only by resistance on the part of the potentialslaves. This position is tempered by an argument – quite incorrect, in our

view – that chattel slavery is not compatible with industrialization because,in crude terms, advanced capitalism needs consumers and skilled workers

who respond to incentives. Thus, it is argued, slavery exists when conditionshobble the ability of people to resist enslavement and tends to disappear

with the onset of industrialization. A second general position is that of Jack Goody, who accepts the overwhelming power element of the previousargument but interprets it in terms of states rather than classes. This hasthe advantage of recognizing that most peoples in history have not enslavedfull members of their own society and have sought slaves from elsewhere.

It also projects to the level of the state the explanation Adam Smith offeredfor slavery at the personal level, which was “man’s love to domineer.” Suchan impulse would probably hold for both states and individuals even if using free rather than slave labor might lead to more profits. Based mainly on his study of African societies, Goody offers the general propositionthat any time a state was significantly more powerful than its neighbors,one could expect the powerful state to use the weaker as a source of slaves. A third general explanation is the now well-known Nieboer-Domarhypothesis that focuses on the environment. It is a land-labor argument

that elegantly lays out the social consequences of land abundance. In short,it holds that slavery will tend to emerge in such an environment becauseone cannot have free land (in other words, a frontier open for settlement),free workers willing to work for wages, and a nonworking land-owning class at the same time. Only two of these three elements can exist at once.Hence serfdom emerged in early modern Eastern Europe, and slavery emerged in the Americas. We find this persuasive, but there is nothing toaccount for why serfdom emerged and not slavery (and vice versa), why slavery never appeared in many land-abundant environments – especially

hunter-gatherer societies – and why slavery disappeared in the Americas atleast several generations before the closing of the land frontier on the twocontinents.



18/753

dependence, servility, and coerced labor

Instead of dwelling further on these general theories, we would like,at least at this stage of the project, to note the different forms of forceddependency that have existed, as well as some common patterns in theinstitution of slavery and how these have changed over time. If we are to

gain any insight into slavery, however, it must be assessed as part of a con-tinuum of dependency typically seen as occupying the opposite pole fromfree labor and separated from it by such institutions as indentured servi-tude, convict labor, debt peonage, and serfdom, to mention just a few of the intervening categories. Institutionalized dependency and servitudehad been accepted without question in Western and non-Western culturesalike, from the dawn of recorded history until the modern historical era,and they have formed one of the basic institutions that have appeared inalmost every culture. Earlier discussions of dependency, and more specifi-

cally slavery, where they occurred, were couched in terms of how individualslaves should be treated, who should be a slave, and how one could fallinto or lose slave status, but not whether the institution itself should exist.Moreover, however firmly the modern mind sees free labor as the antithe-sis to slavery, free labor arguably did not exist at all until the nineteenthcentury in the sense of the master-servant contract being enshrined in civilrather than criminal law. For example, free labor emerged first in the UnitedStates. As late as in England, a worker who refused to comply withthe terms of his contract was viewed as stealing from the employer. Indeed,

when the post-emancipation British West Indies colonial authorities intro-duced what the Colonial Office in London regarded as a harsh labor code,it was pointed out that the new code was basically adapted from the BritishMaster and Servant Act. More recently, Kevin Bales has estimated that million slaves lived in the late-twentieth-century world. It is possible toquestion the definition he uses – it appears to cover a range of dependency relations rather than chattel slavery per se – but even accepting it for themoment, million constitutes far less than percent of today’s globalpopulation. Two and a half centuries ago (as Arthur Young, among others,

pointed out), a definition of “unfree status” similar to that employed by Bales would have encompassed a majority share of the mid-eighteenth-century’s working population, whereas a definition of free labor in themodern sense would have covered few, if any, waged workers in or inany preceding era. Broadly, then, institutionalized coercive relationships,

whether for profit or for some more overtly social purpose, were normalbefore the nineteenth century and have diminished rather dramatically since.

Perhaps the first step is to recognize changes in the way societies have

defined the various forms of dependency. Thus, as already hinted, even thenature of free labor has changed substantially within the confines of theperiod to which volume three of the present project is devoted – waged



19/753


labor in seventeenth-century England (and even in mid-nineteenth-century America) being taken as a sign that an individual could not possibly bea full citizen. Among the more overt forms of dependency and coercedlabor, convict labor (in the sense of those guilty of offenses being required

to labor) by the state has increased dramatically since the early mod-ern period. Prior to this, and in many non-Western environments long afterward, those guilty of crimes against the community might be physi-cally chastised or expelled. Punishment had few implications for labor. In

Western societies, physical chastisement came to be supplemented by, or insome instances replaced with, incarceration, and expulsion became system-atized into transportation. In both cases, however, convicts were frequently expected to labor as well. The Siberian case is well known. Exile was stipu-lated as early as , but the forced labor of exiles is an eighteenth-century

phenomenon, with, in the British case, a rapid switch from colonial North America to the antipodes as the place of exile. The most striking exampleis perhaps Australia, where shortly before the ending of transportation inthe s, convicts brought halfway around the world formed a similarproportion of the total population as had slaves in South Carolina lessthan a century earlier, and a far greater proportion than was ever the casein Siberia. They were also responsible for much of the infrastructure thataccelerated the economic development of Australia. Despite this, the exac-tion of labor was never the major reason for the creation of convicts in the

first place, or even, after conviction, for the existence of schemes that usedthe labor of those convicted, such as workhouses, prison gangs, galleys,soviet gulags, and transportation to distant colonies. Indeed, the history of coerced labor in the context of the history of the community’s or state’sneed to punish transgressors seems a story of lost economic opportunity.One possible reason for this is that few schemes to harness the labor of convicts appeared to have warranted the expenditures they incurred – atleast within the norms that most societies regarded as acceptable for thetreatment of convicts. If convicts had been treated like African slaves, then

there might have been different economic consequences.In classical times, prisoners of war were probably the major source of

slaves, especially in the early expansionary days of the Roman Empire, as was also the case more recently in Africa and the indigenous Americas.Historically, capture in war has always been a justification of slavery. If a victor has the power to end a person’s life, then presumably the victor alsohas the power to inflict social death, or slavery, as opposed to biologicaldeath. A typical pattern at the conclusion of a battle was to inflict thelatter on adult males and the former (slavery) on women and children.

Such behavior is observed in the struggles between core states in WesternEurope and the peoples that spearheaded the great migration prior to thefall of the Roman Empire and on down to the early Middle Ages. It was



20/753


also prevalent in struggles between most premodern polities everywhere inthe world. The first effect of the emergence of large states and empires –

whether in China, Mesoamerica, or the aforementioned case of Rome, where state structures allowed the control of men as well as of women

and children – was that men, too, became slaves. Yet in the European world, treatment of prisoners of war changed rather decisively around thetwelfth century, as relative equality of power between European states (andalso between Islamic and Christian powers) and the attendant fear thatthe defeated power might be the victor in the next conflict meant thatgradually more and more prisoners of war came to be exchanged or ran-somed. Yet when Western European nations ended enslavement of oneanother, they still carried on extensive warfare resulting in large-scale deaths,rape, and pillaging. Whatever the reason, there is almost no evidence of

prisoners of war being enslaved in the European Atlantic world during the era of American slavery, and indeed, no indication of servitude of any length being exacted by the victors in the many intra-European wars of the era (except, perhaps, for Dutch prisoners being put to work draining the English fens in the seventeenth century for the duration of hostilities).The major exception was prisoners of civil wars and those on the Celticfringe that resisted the expansionary impulses of the core states of WesternEurope, they were sent in large numbers to American plantations, at leastin the seventeenth century, but always as servants with fixed terms rather

than as chattel slaves, and with offspring who were free.Debt bondage was a form of servitude based upon an initial agreement

to borrow funds and continued until the time, if ever, the debt was repaid.The debt was payable by the family of the borrower if the latter wasunable to repay while alive. Lenders were accused of extending too muchcredit or charging an excessively high interest rate so that repayment wasnever possible. The borrower would therefore become bound for very long periods, perhaps for life. Debt bondage was a system of coercion sometimesassociated with the post-chattel-slavery era, as manifested in nineteenth-

century India, but it was practiced widely and in some cases earlier inother parts of Southeast Asia, as well as in Latin America, Africa, andChina.

Serfdom has a history going back to at least ancient Greece and formedthe basis of agricultural production and rural social structure alike in

Western European medieval countries. The classic explanations of its rise,in what might be called its first resurgence in the aftermath of the fall of the Roman Empire, allow for some peasant agency. The feudal contractprovided some protection from marauding invaders for those working the

land in return for feudal obligations to the lord, who provided the security.From the late fifteenth to seventeenth centuries, serfdom went througha second renaissance in Eastern Europe and, on a much smaller scale, in



21/753


Scotland after seemingly heading toward insignificance in the West. Boththe scale and the intensity (that is, the restrictions applied to the peasant)increased in the east through to the eighteenth century, as the Russianand Prussian states extended the area under their control eastward. By

late in that century, there were probably more serfs in Europe, including Russia, than ever before. Expansion also meant that the term “serf” cameto cover a much wider range of servile relationships than earlier. Serfdommay have disappeared in Scandinavia, England, and the Netherlands, butin most parts of Western Europe, including Germany and France, peasantsstill owed residual obligations to landholders. Indeed, in Germany, suchobligations acted as a major restraint on German migration to both eastand west, as German peasants had to compensate their lords before they could legally migrate. Peasant support for the early stages of the French

Revolution is testimony enough to the significance of similar obligations west of the Rhine.

The new “full” serfdom that developed in Eastern Europe from the six-teenth century varied somewhat from its Western predecessor. Althoughprimarily a means of ensuring that landholders would have a supply of labor, and the state a pool of potential soldiers, a new form of serfdom alsoshowed up, stripped of its military aspects, in mines in Scotland, Germany,and even in the lead mines of Elizabethan England. In the Scottish case,valuations of the collieries reflected the number, age, and sex of the serf

workforce in a way familiar to those who have studied probate records ordeeds in plantation regions in the Americas. In addition, the second serf-dom showed much less evidence of the contractual (implicit or otherwise)basis for serf status that historians have seen in its Dark Ages predecessor.The new lands acquired by an expanding Russian state were taken fromindigenous, mainly Turkic, peoples and remained highly insecure. Hun-dreds of thousands of Russians and other Slavic peoples fell victim to slaveraids and died in servitude in Islamic and Christian Middle Eastern regions,as indeed the origin of the term “slave” suggests. Nevertheless, there is little

sense of a contractual relationship between the peasant on the one hand andthe state, or the local pomeschiki class in Russian history, on the other. Theexpansion of serfdom occurred overwhelmingly at the initiative of anexpanding militaristic state. Equally important, some Eastern serfs cameto have fewer ties with the land in law, in the sense that both state andseigneurial peasants in Russia could be forcibly moved to new lands in a

way that would not have been imaginable in medieval Western Europe,and which was redolent of chattel-slave status. Under such circumstances –given the heritability of serf status – drawing a legal or behavioral line

between serf and slave status becomes difficult.If the resurgence of serfdom in the east changed the nature of serfdom,

completely new forms of coercive relationships appeared in northwest-ern Europe. The aforementioned master-servant contract, as it evolved in



22/753


the aftermath of the Great Plague, recognized the right of the master tophysically chastise the servant and charge the servant with theft in theevent that the latter did not meet the terms of the contract. From thebroad global perspective, what is extraordinary about such a relationship

is the voluntary nature of the initial contract and the fact that it could berenewed at least once a year. Rural fairs in northwestern Europe becamenot just markets for surplus produce but, late in each year, nascent marketsfor labor as well. In the global history of dependency and coercive labor,this was a watershed in the evolution of agency on the part of those

without property or without kin. The evolution of the master-servant rela-tionship has received very little attention, at least from the comparativeperspective. Equally unique in global terms was the system it spawnedfor facilitating large-scale transoceanic travel. As it evolved in England,

the master-servant contract provided the initial basis for the repeopling of the Americas, and much later, the first large-scale movement of Asianpeoples to the semitropical Americas. In its first manifestation, it came tobe called indentured servitude; in its second, contract labor. In both cases,there was a largely voluntary contract in which individual workers gave upseveral years of their working lives in return for the cost of passage. During the period of the contract, there were clear analogies with slavery in thatthe contract could be sold and severe restrictions placed on the rights of the

worker to move or to avoid the obligations incurred. Once more, the full

weight of the criminal law was applied against the servant for noncompli-ance, but not against the master. The length of the term of labor requiredappears to have varied closely with key variables such as the age and skilllevel of the laborer and the distance (and thus the cost) of the migrant’spassage.

Major change occurred within the slavery category over the centuriespreceding its abolition. There are, arguably, three aspects of slave societiesthat at a preliminary view are to be found across cultures, although theincidence and distribution of these forms do seem to vary in a systematic

fashion. As with attempts at definition, these may seem vague and indef-inite, but they help provide some analytical grounding for an importantissue. First, and perhaps most common from a transglobal perspective, wasslavery as a system of augmenting and sustaining the survival of the groupas a social entity, whether based on some conception of kinship or set of religious beliefs. Such slavery is more likely to be “open,” that is, to providefor eventual entry into full membership of society through a process of “a gradual reduction in marginality” of either the slave or, more likely, thedescendants of the slave (though the stigma of slave origins could survive

for many generations). Slavery of this type could be associated with largestate structures, as in many Islamic polities, or in smaller societies on eitherside of the shift to settled agriculture, as in the indigenous Americas andpre-nineteenth-century Africa.



23/753


A second type of slavery was, as a system, directly organized by the stateto achieve communal goals – perhaps the maintenance of public works,as in irrigation systems, fortifications, or the clearing of salt deposits topermit agriculture, or to provide soldiers for offensive or defensive purposes.

Examples could be found in most phases of Chinese history (referred tosometimes as “Oriental despotism”), in fourteenth- to sixteenth-century Korea, and in Ancient Egypt. Both the Janissaries of the Ottoman Empireand the genizaros of Spanish New Mexico would also qualify. A third typeis as a system for extracting high levels of output from labor for profitof private individuals. Although the state was not directly involved as anowner (though in the early modern period, Atlantic European navies didship some slaves across the Atlantic, and European armies bought Africanslaves for military purposes – galley oarsmen as well as the regular army),

the state normally had to provide the legal structure for the enforcementof ownership rights of slaveholders and, ultimately, the armed force thatsustained the private use of slaves. There are probably no occupationsthat have been performed by nonslaves that have not also been performedby slaves, yet historically, some activities have clearly had a larger slavecomponent than others. Concentration of slaves in particular tasks may be attributed broadly to the ability of nonslaves to avoid activities that

were particularly unpleasant. For two centuries after the mid-seventeenthcentury, field labor on plantations in the Americas was evidently one such

activity. In some societies in the classical era, the focus on production didnot preclude the eventual entry of some slaves into mainstream society. Wecan probably all think of cases that fit none of these three categories – thetribute slaves that came into the Aztec Empire from the north, many of

whom ended up as sacrificial victims, to provide one example. Yet somebroad categorization is useful to get an analytical grasp on an institutionas ubiquitous as slavery – few peoples on the globe have not at some pointin their history been slaves and owners of slaves, often at the same time.

Given these changing conceptions of dependency, it is somewhat tricky

to evaluate the relative importance of the different forms of dependency and coercion over time. Even without such a consideration, the differenttypes do on occasion occur together. Thus, the bulk of European convictssent overseas before were in fact sold in the same manner as indenturedservants to private owners, with only a longer term of service separating them from their nonconvict counterparts. But as social observers from

James F. Brooks, Captives and Cousins: Slavery, Kinship, and Community in the Southwest Border-lands (Chapel Hill, NC, ), pp. –.

The historiography on slavery in the Aztec Empire is extremely thin, but see Robert D. Shadow and Maria J. Rodriguez, “Historical Panorama of Anthropological Perspectives on Aztec Slavery,” inBarbro Dahlgren and Ma De Los Dolores Soto de Arechavaleta (eds.), Arqueologia del Nort y del Occidente de Mexico: Homenaje al Doctor J. Charles Kelley (Mexico City, ), pp. –.



24/753


Aristotle to Marx and Foucault have noted, there can be no doubt thatin addition to changes within a given form, major shifts have taken placein the relative importance of different forms. As already suggested, recentinterpretations stress that free labor as we understand it today did not

exist prior to the nineteenth century. But even understood in seventeenth-century terms, it had neither a long history nor a very wide currency outsiderelatively small enclaves in Western Europe. For convicts and perhapsprisoners of war, significant numbers could not be expected before thecreation of a state system and bureaucracy to maintain them and administertheir activities. Galleys in the Mediterranean drew on this form of labor(as well as on nonconvict slaves) from antiquity to the eighteenth century,but it is unlikely that convicts ever formed more than a tiny share of either the labor force or, more broadly, the unfree, even in societies with

sophisticated state structures. The same is true of indentured servitude andcontract labor, which did not appear at all until the seventeenth century and thereafter never accounted for anything approaching majority statusin any society. Serfdom, by contrast, was usually widespread if it existedat all, especially if we define it in the broadest possible way to include allrelationships where individuals gained access to land to produce their owncommodities in exchange for varying circumscriptions of personal actionsand the acknowledgment of obligations to others.

The chronology of the initial appearance of the three systems discussed

in this chapter broadly follows the order in which they were described.Slavery dedicated to augmenting the numbers and sustaining the identity of societies or religions is usually associated with Islam, sub-Saharan Africa,or the indigenous Americas, but it now seems to have application formany parts of the premodern world. As that world is also largely pre-orthographic, historical evidence of it tends to come from oral tradition orfrom those post-orthographic societies with which the premodern society interacted. This means essentially that evidence of such slavery is scarce inthe years before Chinese and European expansion, but there seems little

reason to doubt that it existed and, indeed, may well have been universalin post-neolithic societies. More broadly, an argument might be made thatthe basic social structure in such environments was not class but kinship,and that slavery was a normal component of kinship structures. This isnot to suggest that slavery then was widespread. Too many slaves wouldbe likely to overwhelm the absorptive function of the institution andthreaten collective identities – as indeed happened in several indigenous

American societies in the aftermath of the demographic calamity triggeredby Old World contact. A slave in the two later types of slave systems

described earlier was usually without any rights in law and passed on hisor her status to any offspring. In kin-based societies, by contrast, slavesor their descendants might gradually receive back certain rights as they



25/753


demonstrated acceptance of kinship identity through their behavior. Asthere is no clear dividing line between slave and nonslave, assessments of the extent of such slavery must necessarily be fragile. Nevertheless, in theabsence of severe demographic stress, people without rights at any given

point in time must have formed a very small proportion of the populationsof kin-based societies. From another perspective, however, one that countsas servile all those who were not full members of the kin group and weretherefore in part dependents of those who were full members, then wemight say that the servile would often, perhaps normally, account for themajority of the population.

Systems of slavery dedicated to the extraction of labor, whether for publicprojects or for the production of export crops organized for the benefit of private individuals, are normally associated with stratified societies that

have moved some distance beyond the agricultural revolution. When theseappear, it is possible to think in terms of “slave societies” instead of “societies

with slaves,” to use Moses Finley’s well-known designations. It is alsoprobable that slavery of this type was what the major social science modelersof slavery, both Marx and Engels, Nieboer, and Domar, had in mind.Indeed, this form of slavery is what most people have in mind when they think of the subject at all, especially those who have used the term “slavery”to draw attention to abusive or exploitative labor situations from early timesto the present day. Many Caribbean islands had more than three quarters

of their populations as chattel slaves with no prospect of change of statusprior to the abolitionist era. Brazil probably approached a point where half of its population was enslaved at several points prior to the early nineteenthcentury. Yet because of the absolute nature of the definition of slavery inthese societies, and the rarity of any intermediary category between slavery and freedom, the proportion of the population that had full rights wasactually quite high from the global historical perspective adopted here,and high, too, compared to the share of free people – using here moderndefinitions of freedom – that existed in the countries of Western Europe

that owned these islands. Though the share of slaves in Rome, Greece,and the slave Americas was much higher than was ever the case in kin-based societies that used slavery as a way of augmenting their numbers andsustaining their identities, there have been relatively few “slave societies”in history. They appeared relatively late in human social evolution, andthough they have had a very high profile in recorded history – being associated usually with imperial systems and “human progress” to borrow David Brion Davis’s ironic association – they probably never accounted foranything like the majority of slaves on the globe at any point in history.

Thus, most slaves in history have experienced their servitude in what aretoday termed premodern social environments. It also seems highly probablethat the number of slaves in the Americas has always lagged behind thenumber of serfs in the Old World.



26/753


The advent of large-scale slave societies did not mean that the originalkin-based form of slavery disappeared. The two, sometimes all three, formsof slavery existed at the same time. In the Atlantic world, some scholarsargue that the kin-based system of slavery at the periphery of capitalist

development in both Africa and the indigenous Americas was transformedby a burgeoning Atlantic-based market system into something more akinto slavery in the plantation Americas. Thus, by the nineteenth century,the Cherokee in the United States owned cotton plantations worked by

African slaves, and slaves owned by Africans in different parts of sub-Saharan Africa grew peanuts and cloves for sale into the Atlantic economy.

Yet the total value of such activities is so small when compared to thevalue of any major crop in the white-dominated plantation Americas thatsuch a slippage into a new form of slavery cannot have been extensive.

A much stronger consequence of contact between different systems wasthat plantation societies drew on their kin-based counterparts for slaves,first in the Americas, then on the African coast, and finally in Dutch Asia.Slaves traded between the two systems were individuals without any rights

whatsoever in either sphere, but the trade ensured that they shifted from anenvironment where a reduction of their social marginality was possible toone in which the gradual reclaiming of rights was an unlikely eventuality.

Returning to the overview of dependency and coerced labor over the very long run, we can observe three major patterns. First, though slavery was

ubiquitous, the share of slaves in kin-based slave systems was not likely tohave been very great. However, if we define freedom as emanating from fullmembership of a given society so that, first, one has the right to participatein the decision making of the kin or community in which one lives, andsecond, one is in possession of most of the bundle of rights that makeup possessive individualism, then the share of free individuals in kin-basedsocieties was also small. Thus, the vast majority of people in most societiesin history have been neither slave nor free, once we consider the limitedrights to political participation that existed, and not just freedom from

labor coercion. A second pattern is the polarization process that appearsto have been associated with the rise of more complex economies andimperial systems. The share of both slave and free in such societies appearsto have risen sharply, and the intervening categories of dependence havealmost disappeared. This observation is another way of approaching theparadox that has drawn the attention of Orlando Patterson, who has arguedthat our understanding, indeed awareness, of freedom was dependent onslavery. The lines between slave and free (defined in terms of citizenship)

were clearly delineated in Greek, Roman, and, with a religious orientation,

Islamic societies, too. The slave-free dichotomy was perhaps at its starkestin the Americas.

Orlando Patterson, Freedom in the Making of Western Culture (New York, ).



27/753


A third major pattern has been the rise and fall of the incidence of coercive systems in the last five centuries, in a world in which kin-basedsystems of slavery continued to thrive. From the fifteenth to the early twen-tieth centuries, systems of serfdom, slavery, convict labor, and indentured

servitude expanded dramatically and in close unison. Four out of fivetransatlantic migrants prior to arrived in the Americas owing serviceto another, most of them having been physically coerced into leaving theircountry of origin. Yet in little more than a century, coercive migrant sys-tems had disappeared. The last slave ship crossed the Atlantic in ,the last transoceanic contract labor vessel (with terms of service enforced

with penal-code sanctions) arrived in , and the last convicts returnedfrom Devil’s Island to France in . A related and even more importantdevelopment was the virtual disappearance of all ideological justifications

of inequality and dependence. In the twentieth century, there have beenintense debates on the meaning of freedom, but none at all on its desir-ability. The net result is that from the perspective of the early twenty-firstcentury, while inequality is clearly rife in the modern world, there is noattempt to justify it in the terms employed in the earlier debates. Theideological shift has swept away not only the American slave plantation,but, almost as comprehensively, the kin-based systems of slavery in theindigenous Americas, in Africa, and in Asia. At no point in history hasthe share of the global population who see themselves as full members of

society been as great as it is now. Although slavery today is seen as the epitome of evil, its stigma is not

entirely a function of modern conceptions of freedom. However muchslavery has historically formed part of a range of dependent relations, ithas tended to be regarded across cultures at best as a particularly hard andunfortunate fate, and at worst as the ultimate degradation for any humanbeing. In many social environments, it has been viewed not as an alterna-tive to death, but as a fate worse than death, although most societies thathad some form of human sacrifice also had slavery. Individuals who sold

themselves into slavery did so only as a last resort, thus suggesting thatavoidance of slavery was of paramount importance to them. The stigma of a slave-ancestor in most non-Western societies was (and in many, stillis) widespread. Long before the abolition process was complete, Frederick Douglass made it clear to supporters of other social reforms that antislavery should have priority because there was nothing at all to compare with itsmalevolent impact. Scholars of the social history of the colonial Americashave equated the conditions of indentured servants, convict and contract

David Roediger, “Race, Labor, and Gender in the Languages of Antebellum Social Protest,” inStanley L. Engerman (ed.), Terms of Labor: Slavery, Serfdom and Free Labor (Stanford, CA, ),pp. –.



28/753


laborers, and even free wage-workers with those endured by slaves. Nev-ertheless, there could have been few slaves at any point in the history of slavery in the Americas who would have spurned an opportunity to switchtheir status with that of any one of these nonslave groups, just as few in

history have opted to enslave themselves. The distinctiveness of slavery inhistorical (as opposed to modern) terms seems to lie in the close to absoluteone-sidedness of power in the master-slave relationship, at least in formallegal terms. Even where slavery might offer freedom from starvation and onoccasion greater life expectancy, the disutility of the institution in the formof being in the power of another was overwhelming. Nonslaves always hadmore protection against the power of a social superior or an employer thandid slaves. In the end, social norms offered far more protection for serfs,convicts, servants, prisoners of war, contract workers, debt peons, appren-

tices, and the myriad other forms of dependency (including children and wives) than they did for slaves. Put another way, these groups were lessmarginal to society than were slaves – a conclusion that appears to holdfor all societies. Even in societies where the exaction of labor was not thecentral function of slavery, they were less likely than slaves to be sacrificed,sold off in times of social stress, or denied rights over offspring and spouses.

What follows from the uniquely degrading nature of slavery observedhere is a central set of questions for the present volume. What is it that deter-mines who is to be a slave, and how does this shift over time and between

societies? Given that the potential for abolition has always existed in thesense that, in every culture, there were large numbers of people – usually the vast majority – who were considered exempt from slavery. Is abolition,then, nothing more than the extension of this exemption to everyone ina given society and, eventually, the attribution of all the characteristics of full personhood to all aliens as well? If so, then just as important as thetype and function of coercion is the question of which groups are viewedas eligible for coercion, and why direct coercion has come to play a very much smaller role in the way societies function than has hitherto been the

case. It is striking that few of the major models of slavery have made mucheffort to address the issue of eligibility for enslavement. Whether land-laborratios (Nieboer-Domar), or power imbalances between societies (Goody),or simply the love to domineer (Adam Smith), general explanations havefocused very much on the conditions under which slavery might appear orintensify, and on the prerequisites of its abolition. For most of the history of slavery, such a focus was entirely appropriate. Major centers of slavery have often drawn slaves from one particular region so that the name forslave became synonymous with the name of the dominant peoples in the

See, for example, Hilary McD Beckles, White Servitude and Black Slavery in Barbados, 1627–1715 (Knoxville, TN, ).



29/753


region of provenance. In French Canada, “panis,” an ethnic designation, was the standard name for slaves whatever the ethnicity of the slave. Incolonial Brazil a similar merging came about under the term “tupi,” andthe origin of the term “slave,” as is well known, comes down to us from a

time when the great bulk of slaves entering the Mediterranean area weredrawn from slavic regions. Yet prior to the fifteenth century, it was rare tohave eligibility for enslavement defined in terms of physical characteristicsor even racial constructions. Cranial deformation, or its absence, among Northwest Pacific Coast peoples comes closest, but it was never an abso-lute marker for slavery. For the great modelers of slavery, it was enoughto acknowledge that slavery was associated with extreme degradation, andthen move on to the social, psychological, or environmental factors thatshaped how extensive the institution of slavery would be, and what form

it would take. And most of the historiography on slavery has followed suitby keying on rather narrow cost-benefit considerations and power relation-ships between groups when addressing historical shifts in the compositionof people making up slave populations – as opposed to explaining why slavery per se has existed.

It is impossible to address the question of eligibility without taking intoaccount how any group responsible for enslavement perceived and defineditself in relation to others. In recent decades, this has come to be knownas the question of identity. Societies have tended to reserve enslavement

for those whom they have defined as not belonging, but this has notalways meant that all aliens were enslaved, or that all slaves were aliens.There have been many instances in history of societies generating slavesfrom within their own ranks, but this has usually occurred only afterthe potential slave has violated, or is thought to have violated, the mostprofoundly held norms of society. In addition, exposure of infants (parentsabandoning a child), typically practiced by all social ranks, was a source of internally generated slaves in many societies, including ancient Rome andChina, which suggests that some acculturation or nurturing process was a

prerequisite of “belonging,” or insider status.In early Rome, citizens could be reduced to slaves, and twins in many

Igbo communities were sold into the Atlantic slave trade directly fromIgboland. It was easier to become a slave from within some societiesthan from within others, just as the ease of reduction of marginality (and

Leland Donald, Aboriginal Slavery on the Northwest Coast of North America (Berkeley, CA, ),pp. –.

H. Hoetink’s work on “somatic norms,” not often cited recently, is an exception to this comment. Almost all the twins in a sample of , Africans taken out of slave ships by British cruisers and

landed in Sierra Leone between and were on vessels that left Bonny, New Calabar, and OldCalabar [“Nutritional Trends in Africa and the Americas: Heights of Africans, –,” Journal of Interdisciplinary History , (): –], as were the vast majority of the small number of recaptivesin the Liberated African Registers with disabilities.



30/753


thus protection against the worst consequences of enslavement) varied.In parts of Europe in the Middle Ages, the problem of “excess children”

was solved by infanticide and abandonment, whereas in some societies in Africa and Asia, the same issue was resolved by involuntary enslavement

of such children. In the slave Americas, manumission – a clear example of reduction of marginality – was more possible in Iberian than in English-speaking areas, but on the other hand, the Iberian Americas were the very last to abolish slavery on the two continents – an unexpected negativecorrelation. The most well-known survey of slavery and kinship in an

African context focuses almost entirely on the movement from outsider toinsider status. The reverse process – in effect how an insider becomes a slave – has received little attention for any society. Generally, however, there

was some formal process whereby the erstwhile insider was redefined as an

outsider, or else, as in the case of Russia, owners believed that their humanchattels were physically different from themselves when the reality pointedin quite the opposite direction. Nevertheless, the vast majority of slavesin history have originated from outside the group that was responsible fortheir enslavement.

The conception of not belonging appears to form the core elementof eligibility for slavery across cultures whether or not the institutionfunctioned primarily to extract labor or to add and integrate newcomers tothe slave owners’ social group. In addition, however, gender and age were

major considerations at different times. Where the main aim of slavery was to augment one’s social or religious group, then women and children would likely be preferred to adult males, who, as already suggested, mightbe put to death immediately or, as in Tupinamba societies in Brazil, held forsacrifice at a point in the future decided by the captor. The trade in slavesacross the Sahara Desert to the Islamic Mediterranean, which grew from a trickle of people in the early days of Islam to a stream ultimately rivaling innumbers the better-known transatlantic trade, was overwhelmingly female,and some of the few male slaves involved were destined to be eunuchs.

As the previous discussion suggests, societies seeking to augment theirnumbers, and ultimately their cultures and/or religions, were extremely eclectic in their selection of potential slaves. The whole point of acquiring such slaves was not just to inflict “social death,” but also to facilitate socialrebirth. The basic aim was to create a new social identity – to producemore people who in the end behaved and thought like the host group,and might fight alongside them. Children from any culture presumably have the potential for assuming new identities, and the chief purpose of preserving women after capture was to ensure a broader base for society

Suzanne Miers andIgorKopytoff (eds.), Slavery in Africa: Historicaland Anthropologival Perspectives (Madison, WI, ).



31/753


to reproduce itself in its own image. Slavery supported by major statestructures could more easily cope with adult male slaves. The ultimateaim, however, was nevertheless to refashion newcomers according to theneeds of the host society. Most countries in the modern world have had the

same attitude toward today’s voluntary immigrants. In the United Statesand Europe, there is a set of core values, loosely described as Western,that all newcomers are expected to believe in and accept. Assimilation

was intended to be the main outcome of the “melting pot.” The majordifference between this and enslavement as traditionally enforced is thatthe decision to migrate today is voluntary. Entry into the new society isno longer preceded by social death – or in the case of the Mintz and Priceformularization of creolization in the Americas, by the traumatic relocationinflicted by the Middle Passage.

But in some cases, women and children were simply not available. Fromthe point at which Christendom and Islamic societies reached military stalemate in early Middle Ages, slaves acquired by one from the othertended to be drawn from the mainly male world of ships and the military,and despite the fact there may have been more English slaves held in North

Africa than black slaves in the English Caribbean in the second half of theseventeenth century, neither were at that time very numerous. In recentyears, historians have begun to draw explicit comparisons between kinshipstructures and bondage in widely separated parts of the world, especially

sub-Saharan Africa and the temperate areas of the indigenous Americas. Whereas Europeans carried off . million Africans to the Americas in just more than three centuries, Africans absorbed few if any Europeansinto their own societies. In the celebrated case of Bullfinche Lambe, anEnglishman was held captive by the king of Dahomey for several years and

was eventually released. But the basic reason for the imbalance, apart fromthe fact that few Europeans could survive in sub-Saharan Africa, was thealmost total absence of European women and children on the African coast,and the essentially nonconfrontational nature of the relationship between

African polities and European slave traders. Not only were there few Euro-pean captives, but Europeans could usually pay what was necessary togain the release of captives before the “reduction of marginality” had pro-ceeded very far. In the temperate North Americas, by contrast, there waslarge-scale settlement by Europeans in the aftermath of the demographicdisaster that overtook the Indian population. These factors ensured thatsome French, English, and Spanish (and many more of Euro-aboriginal

David Eltis and David Richardson, “A New Assessment of the Transatlantic Slave Trade,” in idem

(eds.), Extending the Frontiers: Essays on the New Transatlantic Slave Trade Database (New Haven, CT,), pp. –.

See, for example, the process described by Suzanne Schwarz, Slave Captain: The Career of James Irvine in the Liverpool Slave Trade (Wrexham, ).



32/753


parentage) women and children, and even some men, became absorbedinto indigenous societies via the enslavement mechanism, though obvi-ously the numbers never approached those of Indians enslaved by Euro-peans. Though initially Europeans sought labor, they later strove to bring

about the complete assimilation of whole Indian populations in what wasone of the most blatant of European attempts to recast social identity.Even African slaves were expected to take a role in white society – a partialassimilation presaged in the Iberian areas with conversion to Christianity,learning a new language, and conforming to European behavioral norms.From this perspective, the differences between slavery designed to extractlabor and slavery designed to augment and perpetuate the host social groupfades somewhat.

If societies seeking slaves for any purpose happened to draw dispro-

portionately on a particular region or ethnicity, this was rarely becausethe people of the target group were seen as having the best potential forenslavement. “Slavish personalities” attributed to slaves and serfs, or thepeoples from whom slaves were drawn, have formed part of the ideology of slavery since written records first appeared. Nevertheless, any examinationof the composition of slave-labor forces and conquered peoples in themajor empires from ancient China down to the very early modern Atlanticempires indicates that the enslaved comprised a bewildering mix of peoples.The same was true of systems such as those in the indigenous Americas,

where the Iroquois, for example, brought in so many newcomers from somany ethnic backgrounds that social identity could never become uni-form in spite of the fact that enslavement and absorption were designedto achieve the opposite effect. Broadly, for all systems, shifting powerrelationships between societies, as well as environmental and demographicconsiderations, usually ensured a heterogeneous mix of slaves in the long run.

In the late Middle Ages, this situation began to change. Partly becauseof the aforementioned stalemate between Islam and Christendom, and

more particularly because of rising slavic, especially Russian, power inthe east, the flow of slaves into Islamic areas from the north and westgradually declined. As a consequence, the relative importance of southernregions – specifically sub-Saharan Africa – as a source of slaves for Islamincreased. Growing distinctions between white and black slaves are quiteapparent in Islam from the ninth century if not even earlier. Different termsevolved – “abd” for all slaves, but “mamluk” came to be used for whiteslaves – and higher valuations for whites appear in the record, presumably reflecting the greater likelihood of whites being ransomed. The harder

Richard White, The Middle Ground: Indians, Empires and Republics in the Great Lakes Region,1650–1815 (Cambridge, ).



33/753


tasks were assigned to blacks. Thus, the linking of race and slavery inthe context of the Americas had little precedent in the Old World. By thenineteenth century, the overwhelming share of slaves in the Arab world

was from sub-Sahara Africa, or was of sub-Saharan African descent, and

the association of black skin with slavery became ever stronger. Europeanexpansion into the Atlantic world from the early fifteenth century brought,first, cheap transoceanic transportation, second, a demographic calamity in the Americas, third, the prospect of exportable quantities of preciousmetals and high-value crops, and fourth, labor productivity that was muchhigher in the New World than in the Old. The resulting transatlantic slavetrade, after relatively modest beginnings in the sixteenth and seventeenthcenturies, built up to hitherto unimaginable levels of mass movement of peoples in the eighteenth and early nineteenth centuries. That the vast

majority of the people who made this move were unwilling Africans wasprecisely because the peoples of Europe had some choice over the decisionto migrate. When the options in the Americas narrowed to working on a sugar estate – the major reason for people crossing the Atlantic for nearly three centuries after the late s – then voluntary migrants avoidedthe plantation labor force option and did their moving within Europerather than between continents, but when alternative forms of agriculturedeveloped, free-labor migration was renewed on a much larger scale. Onceagain, the fact that conceptions of freedom in Europe had shifted to permit

most individuals control over the decision to migrate, plus, as describedlater, the invisible barrier that prevented Europeans from enslaving otherEuropeans, generated more coercion and more slavery for non-Europeans.

In both the Atlantic and, less certainly, the Islamic world, for the first timeeligibility for enslavement began to be defined not in terms of which groupto exclude, but rather which groups to include. Muslims debated the issueextensively, and in addition, had a formal proscription against enslaving other Muslims (as opposed to automatically manumitting those enslavedchattels who converted to Islam). The Spanish debated the enslavement

of indigenous Americans, but the striking feature of the establishment of African slavery in the Americas was the set of underlying assumptions about who could be enslaved. Indeed, the absence of a major debate – except forthe Spanish case – is probably responsible for the failure of historians toexplore the eligibility issue in the European context.

There was nothing in European history to suggest either a turning away from coercion or restrictions on who should be subject to enslavement.Europe was a conglomeration of competing polities with extensive writtenrecords of military conflicts, civil wars, and the repression of minority sys-

tems of thought, especially religious thought. Most states evolved unequal

Bernard Lewis, Race and Color in Islam (New York, ), pp. –.



34/753


and rigid social structures early on that fostered some slavery and a greatdeal of serfdom, as well as centralized judicial systems that meted outpunishments for wrongdoing and heresy that appear harsh compared toslavery itself. Slavery had been extensive in Roman times, and for nearly six

hundred years after the fall of Rome, a slave trade from the less developednorth, west, and east of Europe sent a stream of slaves drawn from vari-ous European peoples to the more prosperous areas of the south and theMediterranean – increasingly Islamic after the seventh century. Relationsbetween European polities and the fringe areas of Europe, especially themarauding leading edges of the Great Migrations, had a large enslavementcomponent on both sides, and the system of serfdom thought to havedeveloped as a response to these pressures was clearly related to slavery.In addition, in the late Middle Ages, plagues reduced western European

populations by one-third and created a large shortage of labor. But despiteall the precedents and pressures that appeared to point to more coercionand more slavery, between the eleventh and fifteenth centuries, the internalEuropean slave trade dried up completely, slavery itself diminished signif-icantly throughout Europe, and the institution disappeared altogether inthe north and west of the subcontinent.

More remarkably, when Europeans expanded into new lightly popu-lated, land-abundant territories, the overseas component of that expansionin the West (but curiously, not its eastern counterpart) demonstrated that

Europeans were prepared to enslave the peoples they found in those territo-ries, and to relocate millions of others – Indians, Africans, and Asians alike.Demographic collapse, and in the Spanish case some ideological reserva-tions, soon eliminated indigenous Americans as slaves, and though theDutch carried Asians to South Africa as slaves in the seventeenth and eigh-teenth centuries, and all Europeans used enslaved Asians in Asia in mainly domestic contexts, only distance from the Americas ensured that no Asians

worked in Caribbean cane fields prior to the nineteenth century. In early Brazil, the Portuguese drew extensively on Indian communities for slave

labor to produce sugar. The Spanish used variations of corv ́ee labor as wellas some Africans to exploit precious mineral deposits of New Spain andSouth America. In the Chesapeake, the English in the seventeenth century had some Indian slaves and many peoples of African descent who werenot only not slaves, but full members of society as well. French Canadians

were prepared to buy African slaves but could not afford them, and they

See William D. Phillips, Slavery from Roman Times to the Early Transatlantic Trade (Minneapolis,

) for the main slave trade routes across Europe down to the eleventh century. Presumably if the Spanish had developed a large export sugar industry in the sixteenth century,like the Portuguese, their reservations on the use of Indian slaves would have been more in tune withthose of the Portuguese.



35/753


ended up with a slave-labor force that was exclusively Indian. But fromthe mid-seventeenth to the early nineteenth centuries, the economic cen-ter of gravity of the Americas was in the Caribbean and subtropical South

America – a centering that was made possible by a labor force of slaves who

were exclusively African or of African origin. Slavery here came to havea racial exclusivity that was more pronounced than in the Islamic world.In both Islam and Europe, a core culture characterized a wide variety of ethnicities and languages as major elements of a common identity, andtherefore, insider status. Both Islam and Europe traveled a road that began

with the exemption of its own members from slavery, and by the end of theperiod, both cultures were conferring eligibility for enslavement on only one of the many groups of “others,” or outsiders – Africans.

But was the next step in this process of redefining eligibility necessarily

the elimination of the final category such that slavery was seen as inap-propriate for any human being? The intensity and depth of abolitionismin parts of the West suggests that the final step was more likely to hap-pen (or to happen first) in the Atlantic rather than in the Islamic world,and within regions where slavery played a marginal role. As Adam Smithdescribed it in the case of the Pennsylvania Quakers, the demand curve foremancipation could be downward sloping, emancipation occurring first

where slavery was less important. But interpreting the economic argumentis itself often difficult, and the current literature has provided at least three

arguments – the first, that slave owners found the institution unprofitable,the second, that while still profitable, slavery was less so than sectors of the economy drawing on free labor, and third, that slavery came to play an ever-diminishing role in the major slave-owning nations and could beabolished without serious implications. Each of these arguments has dif-ferent implications for the nature of slavery. The debates on the causes of abolition have perhaps drawn too sharp a dichotomy between economicand other (moral, religious, cultural) factors in the process. The greaterthe costs of emancipation, whether because of the ongoing profitability of

the slave system, or the costs of compensating slaveholders, the more likely emancipation will be delayed.

Nevertheless, whatever the cost, unless there is a moral argument of somekind pointing to the need for slavery to end, the institution will continue.Even where the costs to ending slavery were low, this situation alone hasnever by itself led to abolition, nor apparently have high costs ever perma-nently prevented it from happening. Compensation, though not alwayspaid, was always an issue. First, should it be paid (and to whom, and in

what form)? Here the answers were generally clear – compensation to the

slave owners in cash, bonds, or labor time. Second, should emancipation

Robin Winks, Blacks in Canada: A History (New Haven, CT, ), chapter .



36/753


be immediate or gradual, and what should be the role of government in theprocess? An argument for gradual emancipation goes back to Jean Bodinand Condorcet – if slavery is as destructive to slave psychology and cul-ture as the antislavery argument claims, then immediate freedom without

some adaptive period or government control could only be disastrous. In nocase were slaves or serfs ever provided with compensation by their ownersor the state, a pattern that reflects the belief in the importance of property rights. There were a few discussions of compensation to freed people atthe end of the U.S. Civil War, but this was not a source of major debate.Claims for compensation to the descendants of the enslaved (reparations)developed as an issue only in the twentieth century.

Viewing abolition through the lens of social identity does offer someprospect of finessing these older debates, as well as coming to terms with

the continuance of slavery in those parts of Asia and Africa that viewedslavery as an integral part of societies organized around kin groupings. Suchan approach also reduces the distance between slave systems dedicated tothe exaction of labor and those whose aim is to augment the social group.The former always attempted some assimilation, and the latter always hadlabor needs, the most unpleasant of which were invariably performed by slaves or those who were most marginal to society.



37/753



38/753

PART I

SLAVERY IN AFRICA AND ASIA MINOR



39/753



40/753

CHAPTER 2

ENSLAVEMENT IN THE OTTOMAN EMPIRE

IN THE EARLY MODERN PERIOD

ehud r. toledano

introduction

From the middle of the fifteenth century until its demise after World War

I, the Ottoman Empire was arguably the most important Islamic power onthe face of the earth. At the height of its expansion, it ruled a vast territory from the western Mediterranean to the Persian Gulf, from southern Polandto southern Sudan. Many of the sultan’s subjects were not Muslim, did notspeak Ottoman Turkish, and were illiterate, poor, and lived in villages, notin cities. Yet they were all governed by a Muslim, Turkish-speaking, urban,affluent, and predominantly male elite of officeholders. Perhaps the only phenomenon that cut across all these social barriers was enslavement, fordespite the at times enormous differences in lifestyle, enslaved persons came

from all walks of life: They were male and female, rich and poor, powerfuland powerless, rural and urban, Muslim and non-Muslim, and speakersof all the dialects in the empire, with origins as far-flung as central Africa and the eastern Caucasus. What united them was a shared legal statusof bondage, with the variety of social impediments it entailed in eachpredicament.

Perhaps more than anything else, it was this mélange of types that madeOttoman enslavement unique, complex to study and explain, and highly intriguing as a social phenomenon. For its significance lay mostly in its

social and cultural aspects rather than its role in the Ottoman economy. Whereas practically all historically known societies – including Islamicones – enslaved individuals either from within or from outside their bound-aries, few had evolved such a stratified and highly diversified unfree pop-ulation. If until the early seventeenth century most of the enslaved wereprisoners of war, from that point on – but mainly in the eighteenth andnineteenth centuries – people were reduced to slavery through capture andtrade. The main reason for this shift in recruitment was rather simple – ter-ritorial expansion as a result of military conquest ceased almost completely,

For the detailed arguments underlying this essay, see the following books by Ehud R. Toledano: As If Silent and Absent: Bonds of Enslavement in the Islamic Middle East (New Haven, CT, ); andSlavery and Abolition in the Ottoman Middle East (Seattle, ).


bradley, keith. the cambridge world history of slavery

Documents