Вы находитесь на странице: 1из 551


The Ottoman empire as a political entity comprised most of the present Middle East
(with the principal exception of Iran), North Africa and South-Eastern Europe. For
over years, until its disintegration during the First World War, it encompassed a
diverse range of ethnic, religious and linguistic communities with varying political and
cultural backgrounds.
Yet was there such a thing as an Ottoman world beyond the principle of sultanic rule from Istanbul? Ottoman authority might have been established largely by
military conquest, but how was it maintained for so long, over such distances and so
many disparate societies? How did provincial regions relate to the imperial centre, and
what role was played in this by local elites? What did it mean in practice, for ordinary
people, to be part of an Ottoman world?
Arranged in five thematic sections, with contributions from thirty of the worlds
leading specialists, The Ottoman World addresses these questions, examining aspects
of the social and socio-ideological composition of this major pre-modern empire,
and offers a combination of broad synthesis and detailed investigation that is both
informative and intended to raise points for future debate. The Ottoman World
provides a unique coverage of the Ottoman empire, widening its scope beyond Istanbul to the edges of the empire, and offers key coverage for students and scholars alike.
Christine Woodhead is Teaching Fellow in History at the University of Durham.


Edited by Susan Doran and Norman Jones
Edited by Paul Stephenson
Edited by Stefan Brink in collaboration with Neil Price
Edited by Gwendolyn Leick
Edited by Andrew Rippin
Edited by Toby Wilkinson
Edited by Gordon Morris Bakken
Edited by Pedar W. Foss and John J. Dobbins
Edited by John Jeffries Martin
Edited by Philip F. Esler
Edited by Anton Powell
Edited by John Wacher
Edited by Sushil Mittal and Gene Thursby
Edited by Augustine Casiday
Edited by William OReilly
Edited by Martin Hewitt


Edited by

Christine Woodhead

First published
by Routledge
Park Square, Milton Park, Abingdon, Oxon OX RN
Simultaneously published in the USA and Canada
by Routledge
Third Avenue, New York, NY
Routledge is an imprint of the Taylor & Francis Group, an informa business
Christine Woodhead for selection and editorial matter; individual
contributions, the contributors.
The right of Christine Woodhead to be identified as the author of the
editorial material, and of the authors for their individual chapters,
has been asserted in accordance with sections and
of the Copyright, Designs and Patents Act .
All rights reserved. No part of this book may be
reprinted or reproduced or utilised in any form or by any electronic,
mechanical, or other means, now known or hereafter invented,
including photocopying and recording, or in any information
storage or retrieval system, without permission in
writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging in Publication Data
A catalog record for this book has been requested.
ISBN: (hbk)
ISBN: (ebk)
Typeset in Adobe Garamond Pro
by Swales & Willis Ltd, Exeter, Devon


List of illustrations


List of maps


List of tables


List of contributors



Note on Turkish and technicalities


Christine Woodhead

Nomads and tribes in the Ottoman empire
Reat Kasaba

The Ottoman economy in the early imperial age

Rhoads Murphey

The law of the land

Colin Imber

A kadi court in the Balkans: Sofia in the seventeenth and early

eighteenth centuries
Rossitsa Gradeva

Amy Singer

Sufis in the age of state-building and confessionalization

Derin Terziolu



Royal and other households
Metin Kunt

On the tranquillity and repose of the sultan: the construction of a topos

Hakan T. Karateke

Of translation and empire: sixteenth-century Ottoman imperial

interpreters as Renaissance go-betweens
Tijana Krsti

Ottoman languages
Christine Woodhead

Ethnicity, race, religion and social class: Ottoman markers of difference

Baki Tezcan

The Kzlba of Syria and Ottoman Shiism

Stefan Winter

The reign of violence: the celalis c.

Oktay zel


Between universalistic claims and reality: Ottoman frontiers in the
early modern period
Dariusz Koodziejczyk
Defending and administering the frontier: the case of Ottoman Hungary
Gbor goston

The Ottoman frontier in Kurdistan in the sixteenth and

seventeenth centuries
Nelida Fuccaro

Conquest, urbanization and plague networks in the Ottoman

Nkhet Varlk

The peripheralization of the Ottoman Algerian elite

Tal Shuval
On the edges of an Ottoman world: non-Muslim Ottoman
merchants in Amsterdam
smail Hakk Kad


Masters, servants and slaves: household formation among the urban
notables of early Ottoman Aleppo
Charles L. Wilkins


Subject to the sultans approval: seventeenth- and eighteenth-century

artisans negotiating guild agreements in Istanbul
Suraiya Faroqhi
Literacy among artisans and tradesmen in Ottoman Cairo
Nelly Hanna

Guided by the Almighty: the journey of Stephan Schultz in the

Ottoman empire,
Jan Schmidt

The right to choice: Ottoman, ecclesiastical and communal justice in

Ottoman Greece
Eugenia Kermeli

Ottoman women as legal and marital subjects

Baak Tu

Forms and forums of expression: Istanbul and beyond,

Tlay Artan


The old regime and the Ottoman Middle East
Ariel Salzmann

The transformation of the Ottoman fiscal regime c.

Michael Ursinus

Provincial power-holders and the empire in the late Ottoman world:

conflict or partnership?
Ali Yaycolu

The Arabic-speaking world in the Ottoman period:

a socio-political analysis
Ehud R. Toledano







Old bedestan, Istanbul

Inscription over the entrance to the imaret of Mihriah Sultan
at Eyp, Istanbul
Vakf complex at Belen, near Aleppo, s
Sleyman in procession through Istanbul, c.
Gazi Hsrev Bey medrese, Sarajevo
The reign of the celalis: causes, symptoms, characteristics and
Thula, Yemen, mountain fortress
Shahara, Yemen: suspended bridge leading to a mountain
Ottoman and Hungarian/Habsburg garrison troops in border
forts in the sixteenth and seventeenth centuries
Perspective of Bitlis
A room in an upper-class house in Damascus, c.
Abd al-Rahman Katkhuda sabil-kuttab, Cairo, c.
An Ottoman legal scholar at work
Mellings engraving of a late eighteenth-century coee-house
in Tophane
A plunge pool, from the Ahmed I album
Women in a hamam, from Fazl Enderns Hbnnme ve Zennnme
Tavern scene from Fazl Enderns Hbnnme ve Zennnme
A brothel scene, s
Neighbours disturbing a house of sinners
Dervishes in a forest on the Asian side of the Bosphorus
Women looking up at a female figure in the sky, by Vanmour
Ladies picnicking leisurely on the Bosphorus
The grandsons of Ali Paa of Ioannina
Mehmed Ali mosque, Cairo, mid-nineteenth century



Ottoman towns and cities in Anatolia and the Balkans

The eastern and southern Ottoman world
Western Syria: eighteenth-century eyalet divisions
Ottoman north-eastern frontiers in Europe, c.
Ottoman Hungary
Plague networks, c.
Provincial power-holders in Anatolia and the Balkans,



Campaigns of Selim I
Campaigns of Sleyman
Campaigns after Mehmed III
Decrease in the number of Ottoman soldiers in frontier forts in the
vilayet of Budin,
Ottoman creditors of the Ragusan merchant Bucchia and sums he
owed to them at his death
Shares in rural-agrarian malikane tax farms in Aleppo,



Gbor goston is Associate Professor at the Department of History, Georgetown

University. His field of research includes Ottoman military and economic history and
comparative study of the Ottoman, Habsburg and Russian empires. He is the author
of Guns for the Sultan () and co-author of the Encyclopedia of the Ottoman empire
Tlay Artan teaches at Sabanc University, Istanbul. Her research focuses on Istanbul
and the Ottoman elite in the seventeenth and eighteenth centuries, and she is currently working on the Ottoman royal hunt. She is the author of the section Art and
architecture in the Cambridge history of Turkey, vol. ().
Suraiya Faroqhi is Professor of History at Istanbul Bilgi University. Her latest book
is Artisans of empire: crafts and craftspeople under the sultans (). She is currently
researching the ways in which Istanbuls inhabitants during the early modern age
made use of the natural resources at their disposal.
Nelida Fuccaro is Reader in the Modern History of the Middle East at the School
of Oriental and African Studies, University of London. She is a specialist on frontier
societies and the author of The other Kurds: Yazidis in colonial Iraq () and Histories
of city and state in the Persian Gulf: Manama since 1800 ().
Rossitsa Gradeva is Associate Professor of History at the American University in
Bulgaria and a member of the Institute of Balkan Studies, Sofia. She has published
widely on Ottoman legal, administrative and military institutions in the Balkans and
on Muslim and non-Muslim communities. She is the author of Rumeli under the
Ottomans 15th18th centuries: institutions and communities () and War and peace
in Rumeli: 15th to beginning of 19th century ().
Nelly Hanna is Professor and Chair of the Department of Arab and Islamic Civilizations
at the American University in Cairo. Her work focuses on social groups outside the
establishment, such as artisans, traders and merchants, with a special emphasis on the


economy and its impact on culture and on society. Her major publications include In
praise of books: a cultural history of Cairos middle class 16th18th centuries () and
Artisan entrepreneurs in Cairo (16001800) and early modern capitalism ().
Colin Imber retired as Reader in Turkish at the University of Manchester in . He
is the author of The Ottoman empire, 13001481 (); Ebus-suud: the Islamic legal
tradition (); The Ottoman empire: the structure of power (, nd edition, );
and The Crusade of Varna, 14431445 (); and is the editor of Norman Calder,
Islamic jurisprudence in the classical era ().
smail Hakk Kad is an Assistant Professor at Istanbul Medeniyet University. His
PhD dissertation on the role of Ottoman merchants in OttomanDutch trade is being
prepared for publication. He is currently researching relations between the Ottoman
empire and South-East Asian countries during the nineteenth and early twentieth
Hakan T. Karateke is Associate Professor of Ottoman and Turkish Culture, Language
and Literature at the Department of Near Eastern Languages and Civilizations,
University of Chicago. He is currently working on the social history of the Ottoman
Turkish language.
Reat Kasaba is Stanley D. Golub Chair in International Studies at the University of
Washington, Seattle. He is the editor of volume of the Cambridge history of Turkey:
Turkey in the modern world () and author of A moveable empire: Ottoman nomads,
migrants, and refugees ().
Eugenia Kermeli is an Associate Professor at the Department of History, Bilkent
University, Ankara. She is interested in Ottoman law and has published articles on
land issues and zimmiMuslim relations in the Ottoman empire. She is currently
working on a book on legal pluralism in the Ottoman empire.
Dariusz Koodziejczyk is Professor of History at the University of Warsaw. He is the
author of OttomanPolish diplomatic relations (), The Ottoman survey register of
Podolia (), and The Crimean Khanate and PolandLithuania: international diplomacy on the European periphery ().
Tijana Krsti is Associate Professor at the Department of Medieval Studies, Central
European University, Budapest. Her research focuses on cultural and religious history
of the early modern Ottoman and broader Mediterranean world. She is the author of
Contested conversions to Islam: narratives of religious change in the early modern Ottoman
empire ().
Metin Kunt is Professor of History at Sabanc University, Istanbul. Most recently he
co-edited, with Tlay Artan and Jeroen Duindam, Royal courts in dynastic states and
empires ().



Rhoads Murphey is Reader in Ottoman Studies at the University of Birmingham.

His current research interests include Ottoman law in theory and practice. He is the
author of Exploring Ottoman sovereignty: tradition, image and practice in the Ottoman
imperial household, 14001800 ().
Oktay zel is Assistant Professor at Bilkent University, Ankara. He studies social and
demographic changes in rural Anatolia during the sixteenth and seventeenth centuries,
and is currently engaged in research on Caucasian immigrants in the late Ottoman
empire. He is the author of Dn Sancs () and After the storm: the collapse of rural
order in Anatolia (Amasya 15761643) (forthcoming).
Ariel Salzmann is Associate Professor at Queens University, Kingston, Ontario. Her
exploration of the Mediterranean and West Asian past employs comparative analysis
and interdisciplinary methodologies. Her book Tocqueville in the Ottoman empire:
rival paths to the modern state () interrogates accepted theories of political development in light of empirical research on later Ottoman governance.
Jan Schmidt is Lecturer in Ottoman Studies at Leiden University. He is currently
working on an edition of the Istanbul correspondence of Rudolf Kraus ();
his latest publication is a catalogue of the Turkish manuscripts in the John Rylands
University Library in Manchester ().
Tal Shuval teaches in the Department of History, Philosophy and Judaic Sciences in
the Open University of Israel. He is currently researching the relations between Jews
and Muslims in eighteenth-century Algiers and is the author of La ville dAlger vers la
n du XVIIIe sicle () and several articles on Ottoman Algeria.
Amy Singer is Professor of Ottoman Studies at Tel Aviv University. Her current
research focuses on Ottoman public kitchens and the city of Edirne. Most recently,
she is the author of Charity in Islamic societies () and the editor of Starting with
food: culinary approaches to Ottoman history ().
Derin Terziolu is Assistant Professor in the Department of History at Boazii
University, Istanbul. She has published articles on aspects of Ottoman cultural, intellectual and religious history, and is currently working on a book on the transformation
of Islamic piety in the early modern Ottoman empire.
Baki Tezcan teaches history and religion at the University of California, Davis. Most
recently, he co-edited Beyond dominant paradigms in Ottoman and Middle Eastern/
North African studies: a tribute to Rifaat Abou-el-Haj () and authored The second
Ottoman empire: political and social transformation in the early modern world ().
Ehud R. Toledano holds the University Chair for Ottoman and Turkish Studies in
the Department of Middle East and African History at Tel Aviv University. His major
publications include The Ottoman slave trade and its suppression, 18401890 (),
State and society in mid-nineteenth-century Egypt (), Slavery and abolition in the



Ottoman Middle East () and As if silent and absent: bonds of enslavement in the
Islamic Middle East ().
Baak Tu is Assistant Professor of History at Istanbul Bilgi University. Her areas of
research are gender history and theory, Islamic law, and the social history of violence
and crime. Her PhD dissertation, Politics of honor: the institutional and social frontiers of illicit sex in mid-eighteenth-century Ottoman Anatolia (), is currently
being prepared for publication.
Michael Ursinus is Professor of Islamic and Ottoman Studies at the University of
Heidelberg. He is currently researching demographic trends in the western Balkans on
the basis of Ottoman scal documents and is the author of Grievance administration
(ikayet) in an Ottoman province: the kaymakam of Rumelias Record book of complaints
of 17811788 ().
Nkhet Varlk is Assistant Professor of History at Rutgers University. She is completing a book entitled Plague and empire in the early modern Mediterranean world: the
Ottoman experience, 13471600 and editing a collection of articles entitled Plague and
contagion in the Islamic Mediterranean.
Charles L. Wilkins is Assistant Professor of Middle Eastern History at Wake Forest
University, North Carolina. His recent book Forging urban solidarities: Ottoman
Aleppo, 16401700 () examines the effects of state-sponsored war-making on an
important Ottoman provincial city and administrative centre. He is currently working
on the long-term social and political adaptation of northern Syria to Ottoman rule,
Stefan Winter is Professor of History at the Universit du Qubec Montral
(UQM) and has been Directeur dtudes invit at the cole Pratique des Hautes
tudes (EPHE). His current research focuses on tribal society and provincial administration in northern Syria in the Ottoman period.
Christine Woodhead teaches Ottoman and Mediterranean history at the University
of Durham. Her research and publications are on sixteenth- and seventeenth-century
Ottoman literary history and historiography. She is currently working on letter collections of members of the Ottoman ulema.
Ali Yaycolu is Assistant Professor of History at Stanford University. His research is
on the transformation of the imperial system, provincial communities and regional
notables in the eighteenth- and early nineteenth-century Ottoman world.



It has been a pleasure to edit this book, with regard both to its content and to its
contributors. In terms of content, the volume is testimony to the enormous and very
fruitful expansion of Ottomanist historical study in recent decades. To study Ottoman
history as an undergraduate in the s was to enter a completely unfamiliar and
challenging world, exciting in its very difference. However, from todays perspective
it is clear that the books and articles available then were few in number, and that most
writing was on the relatively narrow military and political history of a dynastic state.
This was not by choice but by default, as at that time there seemed to be little else.
Broader social history topics were thought to be impossible to study due to lack of
sources; Ottoman historical studies were largely in a world of their own. By contrast,
however, the student of Ottoman history in the early twenty-first century enters a
dynamic, ever-expanding subject area which now presents a wealth of sources, topics
and publications unimaginable to a previous generation.
This book began with a series of questions about the nature of the Ottoman world,
its diverse societies and the ways in which these evolved over the period of the empires
existence, both in their own terms and in terms of being Ottoman. Within these
general parameters, contributors were invited to write about what they found most
significant in their own subject areas. No attempt was made to provide a comprehensive coverage of topics, either chronologically or throughout the empire. Although
several essays begin with a brief consideration of the relevant historiography, each
develops according to its own priorities. The result is a stimulating variety of styles and
approaches typical of current Ottomanist historical research.
As befits its subject matter, this book also represents a collaborative project of truly
international dimensions. I am most grateful to all my colleagues, not simply for their
willingness to contribute to this volume and for producing such informative and
thought-provoking essays, but particularly for their patience and good humour in the
face of the delays which are almost inevitable with such a large-scale project, and of the
demands of a pedantic editor.
Specific thanks are due to Richard Stoneman, former senior editor at Routledge,
who initiated this volume in as an essential part of the Routledge Worlds series
by inviting me to edit it. I agreed to do so only after consultation with an inforxiv


mal advisory group of Ottomanist colleagues. All were encouraging, and most have
contributed to the final volume. Thanks are due to all these advisors, but especially
to Suraiya Faroqhi for her enthusiastic initial response, for recommending several
contributors whose research was previously unknown to me, and for setting such a
positive and inspiring example in her own work and particularly in the collaborative
volumes which she herself has undertaken. I am also grateful for the helpful advice and
remarkable patience shown by Routledge sub-editor Amy Davis-Poynter, and to the
Routledge cartography department for redrawing some rather primitive maps. Philip
Williamson has offered advice and encouragement at critical moments. In particular,
without the invaluable editorial assistance of Clair Boyd, the final text would not have
been put together so efficiently or so carefully in the last, crucial stages.
While it is a truism to say that we learn by teaching, it is certainly the case that
my own interests in Ottoman history would not have developed as broadly as they
have without the stimulus of undergraduate teaching in the Department of History
at the University of Durham, facilitated in turn by the unblinking willingness of the
University Library to purchase everything I requested. Without the constant need to
rethink, explain and justify to a sceptical audience, I would never have been able even
to consider editing this book. I therefore hope that this volume will be as relevant and
stimulating to non-Ottomanists, from students to established scholars, as it will be to
Ottomanist historians, for both teaching and research purposes.
This book has no dedication, except perhaps to previous generations of
twentieth-century Ottomanist historians in various countries who persevered in their
work despite discouragements such as archive closures, cultural negativity and political
disincentives. Without their persistence and scholarship, current Ottomanist historians would not be where we are now.
Christine Woodhead




As far as possible, in the interests of simplicity, all Ottoman words and phrases have
been spelled using the modern Turkish alphabet. A few exceptions occur in essays
dealing with the Ottoman Arab provinces. In such cases, a simplified Ottoman Arabic
transcription is used.
The following is a general guide to the pronunciation of those consonants and
vowels used in Turkish which either do not appear in the English alphabet or which
differ markedly from their English pronunciation. Vowels are usually pronounced
short. The stress in Turkish is generally on the last syllable of the word.



j as in jam
ch as in church
has little sound of its own; usually lengthens the preceding vowel:
e.g., Osmanolu = Osman-oh-lu = the son(s) of Osman
s as in this (not as in these)
sh as in ship
(i) short a as in apple
(ii) long a as in father (in Arabic and Persian words)
e as in red
[undotted i] as i in cousin
i as in pin
o as in otter
eu as in French jeu
u as in put
u as in French tu

Names of major cities and regions are usually given in their Anglicized form where
these exist. Otherwise, place-name styles vary according to the region or era under
discussion, with Ottoman and modern variants given where appropriate.

Note on Turkish

All dates are given in ce (Common Era) reckoning. A double-year date such as
indicates often an original hicri date in the Muslim calendar which spanned these two

References appear as Harvard-style endnotes at the end of each essay. Full bibliographic details are given in the composite bibliography at the end of the book. The
alphabetical order followed in the bibliography is that usually preferred in Englishlanguage publications and which ignores the extra subtleties of Turkish letters. For
this cultural insensitivity, I apologise to Turkish-speaking readers.



R. Danu

Mediterranean Sea



R. Isker



R. Mari






R . D an u




Izmir (Smyrna)




R. Sakarya













Black Sea




R. Bug


















R. M

R. Tisza





Buda (Budin)







R. Dr


R. Eup



R. Tigris












Ottoman towns and cities in Anatolia and the Balkans


Black Sea










R. E








R. Tigri








R. N










R. N






The eastern and southern Ottoman world



Christine Woodhead

as there such a thing as an Ottoman world? Among the numerous and varied
answers there must be to this question, three stand out for the purposes of this
book. The first and most essential is the obvious territorial definition. The Ottoman
empire as a political entity comprised most of the present Middle East (with the principal exception of Iran), North Africa and South-Eastern Europe. For over years,
until its disintegration during the First World War, it encompassed a diverse range of
ethnic, religious and linguistic communities with varying political and cultural backgrounds. Further questions on the nature of the Ottoman world now follow. For
instance, Ottoman authority might have been established largely by military conquest,
but how was it maintained for so long over such distances and so many disparate societies? How did provincial regions relate to the imperial centre and what role was played
in this by local elites? How did the inhabitants of major cities such as Cairo or Damascus
adjust to Ottoman rule, or did it adjust to them? What produced the consensus which
supported the empire in the seventeenth and eighteenth centuries? To what degree did
subject peoples see themselves as part of a larger political and economic whole? What
did it mean in practice, for ordinary people, to be part of an Ottoman world?
Traditional study of the Ottoman empire based upon narrative sources inevitably focused on the dynasty in Istanbul, its military undertakings and its centralized
administration. This tended towards a relatively narrow, government-centred assessment of the empires rise and decline, easily correlated with more, and generally less,
competent sultans. Echoes of this easy schematic approach, which tied the health of
the empire as a whole to dynastic and governmental stability, still linger in general
history textbooks, to the despair of most Ottomanist historians. Given its strong focus
on Istanbul-centred government, traditional historiography did not raise such questions about how the wider state functioned or about the lives and outlook of Ottoman
A second answer to the initial question is therefore to consider the nature of the
societies which came under Ottoman rule, and to ask whether the latter had sufficient
influence on social, political, economic or cultural developments across the empire
so as to produce some definable degree or quality of Ottoman-ness which was more

Christine Woodhead

than simply acceptance of the sultans government. However, this approach has been
neither popular nor, until relatively recently, possible. Unlike the British, French or
Spanish empires, the Ottoman empire did not bequeath to its successor states a world
language or a widespread and lasting change of religious faith. Most of the constituent
parts into which the empire dissolved in the nineteenth and early twentieth centuries
could easily be portrayed in nationalist terms as having retained their integrity despite
Ottoman rule, whether the latter was considered as irrelevant or oppressive, or both.
The resulting historiographical distortions have been as difficult to overcome as has
the decline paradigm.1
In conjunction with this deeper analysis of the Ottoman world, a third answer
would take a still broader approach, expanding the term to include those areas and
political entities, neighbouring and distant from Venice, to England, to Diu in
India and Atjeh on Sumatra, from Poland to Central Asia and the Sudan with which
Ottomans and their state had diplomatic, commercial or cultural relations. Here, the
Ottoman world is the extended world in which Ottomans operated, with the emphasis on trade, communication and comparable lifestyles, rather than on taxation and
militarized organization. The old view that the Ottomans had little interest in the
non-Muslim world except as a site of potential conquest is no longer tenable. Suraiya
Faroqhis study of the interactions of Ottoman officials and ordinary individuals in
the pre-modern era with non-Ottoman merchants, captives, travellers and pilgrims
demonstrates a world or mind generally no more closed than that of their European
counterparts.2 Several Ottomanist historians have also countered the persistent belief
that the Ottomans were not interested in trade, a view which stemmed from overemphasis on the activities of western merchants in the eastern Mediterranean. Not
only was the Ottoman-Muslim world sufficiently large to constitute in itself a world
economy; its preferred routes for commercial expansion extended not westwards into
Europe and across the Atlantic, but eastwards through the Indian Ocean and, during
the sixteenth century, southwards down the east coast of Africa.3
This volume is concerned mainly with the second of these three understandings of
the Ottoman world, with added perspectives offered by the third. It does not present
this world through the eyes of the Istanbul-based Ottoman elite, as used to be the case
in older works which relied primarily on official chronicles and government documents. To study of these sources is now added extensive use of judicial court registers,
endowment deeds and personal records relating to all levels of Ottoman society, male
and female, rich and poor, central and provincial, which enables historians to present
a very different picture.
In this respect the historian who, as a catalyst, has probably had the greatest influence upon the development of Ottoman historical studies in the past half century is
Fernand Braudel. His belief that the Turkish Mediterranean lived and breathed with
the same rhythms as the Christian . . . with identical problems and general trends
challenged Turkish historians to delve further into the Ottoman archives, to present
their findings to an increasingly interested outside audience and, ultimately, to engage
in social-science-based comparative research.4 In recent decades such research has
expanded Ottoman studies enormously in many directions. It has, for instance, emphasized the essential pragmatism of the Ottoman administration and shown the degree
to which efficiency was dependent upon willing co-operation at local level. Islanders
on Limnos, Palestinian peasants and nomadic tribal leaders all had the potential subtly


to destabilize Ottoman rule. In contrast to the traditional view of tightly run central
institutions imposing their systems on subject peoples, we see that, as often as not, the
government had to negotiate constantly in order to succeed in its aims.
Braudels image of Philip II of Spain as the spider at the centre of a huge imperial
web can also apply in modified form to the Ottoman case, certainly in the fifteenth and
sixteenth centuries. If the spider is the central administration in Istanbul, the threads
of its web extend along communication routes and are anchored securely only where
they intersect, in the major towns and cities which were centres of Ottoman provincial
administration. The empty spaces between the threads represent the rural, desert, mountainous or otherwise sparsely populated areas where degrees of Ottoman authority varied
considerably. A greater distance between urban administrative centres the larger the
gaps in the web generally meant a less visible Ottoman presence. The spiders web analogy can be taken further: its adaptability, flexibility and relative ease of repair represent
one aspect, its essential light touch a second. Most of these elements can generally be
found in the immediate post-conquest stages of Ottoman administration in any given
area. The importance given by the Ottomans to settlement policies and to the development of new markets and towns, particularly in the Balkans during the fifteenth century,
also supports the appropriateness of the web analogy. To the old view that the Ottoman
empire was a military state organized for and living off the spoils of war, and to the more
recent characterization of it as an agrarian empire funded by agricultural revenues, we
must add the political and economic significance of urban centres and note the importance to the central government of establishing strong working relationships with the
leaders of urban society, whoever they might be.


Considering how best to divide the study of Ottoman history into meaningful periods
can be a useful way to reassess dominant preconceptions. With no dynastic change for
over years, there is no obvious point of disruption at which to stop and take stock.
Ottoman historiography, with its considerable, and natural, emphasis upon institutional continuity, contributes significantly to the difficulty of identifying significant
change. Traditional rise and decline views suggested two major periodizations. The
first, favoured by European historians, based on external affairs and emphasizing the
military aspect of the Ottoman empire, generally took the Treaty of Karlowitz in
and the Ottoman loss of Hungary as the turning point. The second, taking its cue
from early seventeenth-century Ottoman reform literature, gave priority to political
developments, placing the beginning of decline at some time in the later sixteenth
century, either with the death of Sleyman in or with the onset of corruption
in the reign of his grandson Murad III ().5
In contrast to these centrally focused interpretations, one broad, empire-wide, socioeconomic perspective suggests a three-stage Ottoman empire. The middle stage begins
around , with the establishment of the Ottoman province of Hungary at the end
of a thirty-year period of spectacular territorial expansion. Administrative reorganization, commercial prosperity and the maintenance of a largely self-sustaining Ottoman
world economy characterize this stable state second period. Only from around the
s, due to a change in the dynamics of trade with Europe and ominous defeat in
the war of against Russia, does the empire move into a final third stage.6

Christine Woodhead

Another possibility is to identify a turning point in Ottoman history in ,

with the incorporation into the empire of Syria, Egypt and the Hijaz. This point of
division not only acknowledges the obvious administrative and commercial developments associated with the sudden doubling in size of the empire, but also emphasizes
religious, legal and cultural aspects of Ottoman existence. Leaving aside the practicalities of rule over such extended territory, the question here is how the Ottomans
the johnnys-come-lately in their ex-Byzantine base, of initially rather suspect faith
could make themselves acceptable as rulers to long-established Muslim societies in
the former Mamluk empire. Ottoman adoption of the role of protector of the pilgrimage routes to Mecca and of the Holy Cities themselves also coincided with the challenge for control of Anatolia posed during the sixteenth century by the Shiite state of
Safavid Iran. An increasing Sunnitization of the state resulted.
Other long-view periodizations of Ottoman history will emphasize different perspectives, such as fiscal/financial, architectural, literary, and the history of communications or of regional/provincial developments. However, dominant preconceptions
are equally well undermined by the determined focus of several historians on a specific
problem, as shown in the great gazi debate of the s and s.7 Was early Ottoman expansion driven by the desire for holy war (gaza, fought by gazis, holy warriors),
by the restlessness of its nomadic, tribal following, by the desire for economic advantage through ever-extending control of trade routes, by the need to defend itself against
other, very similar Turkish groups, or by simple opportunism? So much appeared to
hinge on correct understanding of this early phase in Colin Imbers term, the black
hole in Ottoman history which was made all the more tantalizing because of the
shortage of reliable contemporary sources.
A debate of similar significance and intensity has centred on the nature of the celali
uprisings of the s and early s. Were these essentially peasant movements
caused by socio-economic push factors such as population pressure, poor harvests and
over-taxation, or were they the result more of political pull factors such as recruitment
of armed peasants as mercenary troops and the possibility of enhanced social status
and/or financial opportunity in employment off the land? What role was played by
dissatisfied Ottoman officials and former military men, and by the government itself?
Crucially, why was the celali phenomenon confined largely to Anatolia?
The celali phenomenon contributed to a period of crisis in Ottoman administration
and ideology which arose from an accumulation of several factors. In the early imperial era, the Ottomans had been able to develop their administration by adopting and
adapting procedures and personnel which they found in situ in newly conquered areas.
Much Ottoman administrative procedure in the Balkans, for instance, was adapted
from previous Byzantine fief-holding and taxation practices. Similarly, successful
administration of the former Mamluk regions of Syria and Egypt after relied on
taking over much of the previous system, together with its middle and lower-ranking Mamluk officials. However, by around , as Ottoman territorial expansion
largely ceased, challenges for government were no longer those of incorporating new
regions but of solving new problems which arose internally and for which there were
no precedents, models or experienced practitioners already in existence. Such new
problems included not only the socio-economic disruption caused by the celali uprisings, but also changes in military technology which required alteration in the balance
between cavalry and infantry, and hence different recruitment patterns and methods of


payment. They included unprecedented financial difficulties compounded by a per

cent devaluation in the Ottoman silver currency in the mid-s and by the need to
alter methods of tax assessment and collection. They also included increased competition for appointment in both administrative and judicial professions, and the increasing prominence of clientage networks and associated accusations of corruption. The
response to such problems could no longer be to adopt and adapt existing procedures:
new solutions had to be devised. Hence, while the seventeenth century is certainly a
period of crisis and change, it is also a period of trial and error, in working out these
new approaches. Some of these worked, and some did not. From this resulted the
apparent chaos and seeming lack of direction which for long made the seventeenth
century the most difficult and least popular era of Ottoman history to study. Now,
however, as is evident in many of the essays in this volume and encapsulated recently
in Baki Tezcans notion of a second Ottoman empire from around , Ottomanist
historians have largely jettisoned the notion of a post- decline and are identifying instead significant elements of continuity and development.
One of these elements, the role of provincial elites throughout the seventeenth and
eighteenth centuries, constitutes a third area of major debate and reassessment by
historians.8 As long as an efficient, centralized system of government was regarded as
the key to Ottoman success in the late fifteenth and early sixteenth centuries, then any
kind of later decentralization tended to be viewed negatively. Thus the ayan (local
notables, particularly those with political or military influence) were seen to pose
an increasing challenge to Ottoman authority. However, treating both the Ottoman
chronicle record and later nationalist histories with due care, historians now view provincial elites of various kinds as permanent and essential elements within Ottoman
society, without whose co-operation the government would have had difficulty functioning at virtually any period. If early Ottoman rule in the Balkans was facilitated by
the appointment of local Christian lords as timar holders, later Ottoman rule in most
provincial areas, both urban and rural, was also facilitated by the co-option of local
ayan as tax collectors, local officials and recruitment officers.


This book began with a question about the nature even the existence of an Ottoman world. It is not designed to be an introduction to, or a comprehensive survey of,
Ottoman history. For these the reader should turn to the Cambridge history of Turkey
and to Caroline Finkels analytical narrative Osmans dream.9 Nor is this a book about
sultans, Janissaries and grand vezirs as such, about military campaigns, or about foreign or domestic policy emanating from Istanbul. Rather, it is about the constants and
givens of Ottoman society, about the definition of Ottoman, and about the forces
which supported and challenged this. It is about the limits of Ottoman reach, about
significant factors influencing the lives of ordinary people in the empire, and about
how the nature of the Ottoman world appeared in the eighteenth century. It addresses
aspects of the social, and to some extent the socio-ideological, composition of a major
pre-modern empire, offering a combination of broad synthesis and detailed investigation intended both to be informative and to raise points for future debate.
The volume contains thirty essays arranged in five sections, each of which presents a
specific aspect of the larger question. The essays in Part I address fundamental aspects

Christine Woodhead

of Ottoman society which were either constantly present throughout the empires
history or which helped lay the foundations for the later empire. Sufism and pious
endowments were cornerstones of everyday life for Ottoman Muslims, with endowments such as soup kitchens also open to non-Muslims. The almost ubiquitous presence of nomadic groups may have made settled farmers uneasy and the lives of tax
collectors more difficult, but, without their expertise in sheep and camel-raising, meat
would have been in short supply for everyone and commercial goods brought by overland caravans would have been more difficult to obtain. Above all, the essays in this
section emphasize the importance always attached by the Ottomans to a sound economy and to a well-regulated and fair legal basis to society.
Part II addresses the crucial question of what or who was originally meant by
the term Ottoman. Derived from the simple descriptor Osmanl, the followers or
household of Osman, Ottoman was a dynastic term, comparable to Habsburg or
Romanov, used to denote a royal household, which became an imperial government.
On account of the practice of recruiting slave troops and administrators from nonMuslim peoples, and the attractions of Ottoman service to non-Turkish volunteers, at
no stage even in the fourteenth century was the Ottoman ruling group completely
Turkish ethnically.10 Despite the seeming interchangeability of Ottoman and Turk
in some types of historical writing, these terms were never synonymous. Mostly, they
denoted two completely different layers of society. Understanding how the Ottoman
amalgam of sultan and servitors developed, and how it was constantly evolving and
expanding, both politically and culturally, is one of the keys to understanding the
nature and longevity of the empire. This section of the book contains essays on the
creation and projection of an Ottoman identity in the fifteenth and sixteenth centuries, and how the us and them division was being constantly renegotiated.
The development of an Ottoman rhetoric of universal sovereignty was almost
inevitable in the sixteenth century, and was paralleled by the comparable claims in
western Europe of the Holy Roman Emperor Charles V. In practice, the pragmatism and accommodation of local interests which were hallmarks of Ottoman government everywhere were particularly so in frontier regions. Given that the Ottomans
did not inherit a ready-made, clearly defined state, it is worth noting that virtually
all areas under Ottoman rule would have qualified at some stage as frontier regions,
to be incorporated gradually into the empire.11 Three of the essays in Part III deal
with active, open frontiers almost equidistant from Istanbul to the west, north and
east, each with its own variation of Ottoman control. Further frontiers in Algiers and
Yemen differed in the degree of interest and authority of the central government.
In both regions these receded noticeably during the seventeenth century, when the
costs and difficulties of maintaining an Ottoman presence began to outweigh the
returns. Essays on the spread of plague and on eighteenth-century Ottoman merchant
communities in western Europe emphasize continuing Ottoman connections, both
indirect and direct, beyond these frontiers.
Part IV builds upon the social constants considered in Part I by considering in
greater detail specific aspects common to the daily life of large numbers of ordinary
people, both Muslim and non-Muslim. Domestic slavery and household formation,
guild organization, non-Muslim courts, literacy levels, womens rights to divorce, and
how inhabitants of Ottoman cities, men and women, might have spent their leisure
time are all topics which not long ago it would have been thought difficult, if not


impossible, to study seriously in an Ottoman context. There is now a much broader

perspective in which to set the observations of European travellers on Ottoman society.
Two major social groups, merchants and scholars, are under-represented in this volume, partly though not entirely because the primary focus of interest in these groups
lies outside its scope. As mentioned above, views on the extent, independence and
commercial significance of Ottoman trade have been revised considerably in recent
years. With regard to scholars, contemporary Ottoman biographical dictionaries have
always made possible prosopographical studies of the Istanbul-based ulema. Recent
studies especially of seventeenth- and eighteenth-century Syrian scholars will surely
prompt a significant re-evaluation of our understanding of the pre-modern scholarly
outlook.12 It could also be said that there is insufficient attention given here to nonMuslim Ottoman communities. While at one level this is true, at another it could be
argued that, as non-Muslims are discussed in several of the essays simply as part of
Ottoman society generally, this is an appropriate reflection of how, in the pre-modern
period, they were, on the whole, perceived.
Finally, the essays in Part V broaden the perspective once again to consider imperial
cohesion and the crucial question of to what extent provincial elites and societies in the
seventeenth and eighteenth centuries considered themselves Ottoman. In particular,
Ehud Toledanos concluding assessment of the localization of Ottoman elites and the
Ottomanization of local elites in the Ottoman Arab world emphasizes the continuing
flexibility of Ottoman societies in their relationship to the imperial centre.
The chronological period covered by this volume ends for the most part in the early
nineteenth century, thus omitting the final century of Ottoman history. In this sense
it perhaps follows the periodization which takes the reign of Selim III ()
as a major turning point in Ottoman history, in terms both of this sultans nizam-i
cedid (new order) reform programme and of Napoleons invasion of Egypt in .
However, the rationale for ending here is not based strictly on political or foreign
policy markers. The Ottoman nineteenth century has been studied intensively from
many different angles, both for its own sake and as a precursor to the emergence of
the modern Turkish republic.13 The aim of this volume is to draw attention to recent
scholarship on the early and middle periods of the empire, those eras which are more
distant from us in time and which appear more alien in nature, on which some of the
more tenacious scholarly views have been held, and where now the most exciting new
research is to be found.


Hathaway : , on the tendency among Arab nationalist historians to view Ottoman rule
as at best a demoralizing prelude to nineteenth-century western-inspired reforms.
Faroqhi .
Faroqhi : ; Eldem :
. ; Hanna ; Casale .
Braudel : I, preface; see also Inalck .
But see Tezcan for a much broader perspective on the second Ottoman empire.
Faroqhi : esp. .
Summarized and extended in Lowry a.
On historiographical approaches, see Khoury .
Three of the four projected volumes of the Cambridge history are currently available: Fleet
(vol. ); Faroqhi (vol. ); Kasaba (vol. ).

Christine Woodhead
See e.g., Lowry a.
For the classic account of Ottoman methods of incorporation in the early period, see Inalck
. For archaeological approaches to Ottoman frontier history, see Peacock .
On the Ottoman ulema, see Zilfi , ; on seventeenth-/eighteenth-century Arab
thought, see e.g., El-Rouayheb , ; Akkach , b.
E.g., the essays in Kasaba . See also Ehud Toledanos discussion of developments in Ottoman histiography, chapter 30, pp. 4556, 4579, in this volume.






Reat Kasaba

The territories of the Ottoman empire intersected with what geographers refer to
as the sub-Arctic nomadic zone, which extended from the Mediterranean littoral,
through the Anatolian peninsula and the Iranian plateau, on to the mountains of Central Asia. For millennia, tens of thousands of tribes moved constantly across this belt
of high mountains and dry steppes and deserts. Starting in the eleventh century, Turkic and Mongolian tribes arrived in Anatolia and eastern Mediterranean lands. They
became integrated into the indigenous patterns of circulation and altered forever the
social and political make-up and the history of these regions. As they passed through
these lands, these tribes interacted with local communities; some melded into local
relations and networks and abandoned their journey, while others continued to move.
Superimposition of the long-distance migrations onto local structures and movements
created a highly fluid social environment throughout this territory. Especially in Anatolia, between the eleventh and fourteenth centuries it became difficult to distinguish
between the arriving, staying, or departing tribes, let alone between sedentary and
nomadic communities. This was the context within which the Ottoman empire grew
to become a world empire after the thirteenth century.
The integration of the goat- and sheep-herding Trkmen communities of the Anatolian peninsula and the camel-raising Bedouin and Arab tribes of North Africa and
the Middle East created a fluid and heterogeneous society that defied simple characterization in ethnic, religious or administrative terms. At least initially, the Ottomans
had neither the means nor the intention to settle permanently or discipline nomadic
tribes. Instead, classifying tribes in their existing state appeared to be a more pragmatic
approach. As their empire grew quickly in western Anatolia and the Balkans, the Ottomans continued to balance their interest in strengthening the empires administrative
structure and its peasant base with the obvious need to define a clear place for nomads
within Ottoman rural society. To this end, they developed special laws to monitor the
activities of tribes and recorded tribal affairs separately in special documents.1 Maintaining control over nomadic tribal communities was by no means an easy task. For
one thing, the area across which the tribes moved could be quite large. One clan could

R e s a t K a s a b a

spend summers at the source of the Euphrates in the interior of eastern Anatolia and
then move south to the Syrian desert for the winter, a distance of over miles.2 Some
of the tribes were huge, with as many as , to , individuals and sometimes
several hundred thousand sheep and camels.3 Given that Ottoman law recognized
sheep as constituting a herd, and that the state used this as the unit of accounting in
assessing the liabilities of tribes, these were indeed wealthy and formidable units.
The Ottomans not only kept the existing patterns of tribal migration intact, but
they also encouraged mobility, making this an even larger part of the make-up of
Ottoman society. For example, sedentary and nomadic communities were forced by
the state to move across long distances, either as a method of punishment or as a way
of settling newly conquered areas. It should be noted, however, that forceful relocation
of nomadic tribes for punitive or strategic reasons did not automatically entitle them
to land, since they were not necessarily encouraged or expected to adopt sedentary
farming in their new places.4 Instead, in line with established practices in their places
of origin, these communities were allocated grazing lands, since it was assumed that
they would continue their pastoral nomadism in the new regions.

In administrative parlance, Ottoman officials referred to tribes as airet. In order to
facilitate governance and taxation, they grouped the tribes in eastern Anatolia, Iraq,
Syria, and further east in the Arab provinces as Trkmen, Kurds, Arabs or Bedouin.
Those who had moved west of the Kzlrmak River in Anatolia and into the Balkans,
and had increasingly engaged in settled agriculture and become semi-nomads (yargebe or konar-ger), were referred to as yrks.5 Of the main groupings, those in the
east were closer to being absolute nomads than their counterparts in the west.
The largest administrative units the Ottomans recognized among the Kurds and
Trkmens were il and ulus.6 The two largest of these were the Boz Ulus, consisting largely but not exclusively of Trkmens, and Kara Ulus, consisting largely, but
not exclusively, of Kurds.7 Ulus confederations were divided into smaller groups, in
descending order, as boy (sometimes taife), cemaat and kabile. Yrks, on the other
hand, were spun off from Trkmen kabiles and were not organized in the larger units
of boy, il or ulus. Instead, they were classified and registered as kabiles and cemaats,
mostly on the basis of their tax and other obligations or of the places where they circulated.8 For the most part, who was included in the yrk and Trkmen formations
ulus, il, boy, cemaat or kabile was determined endogenously, with little influence from
outside. These groups were given names and were recognized by the Ottoman administration only after they had already taken shape through their own internal dynamics.
The tribal units which the Ottomans recognized for the purposes of administering
could be very large yet territorially loose. In the sixteenth century, there were more
than separate tribes in various sizes registered as part of the Boz Ulus confederacy, whose population then is estimated to have been more than , tents with
million sheep. Their seasonal migrations covered an area extending from Mardin in
south-eastern Anatolia all the way to Iran and Georgia.9
In addition to recognizing the existing groupings and regrouping them under new
labels, the Ottoman government appointed high-ranking officers to administer the
affairs of tribal communities and to assess and collect their taxes. Being at least partially

Nomads and tribes in the Ottoman empire

mobile, tribes were not subject to the authority of the sancak beyi (district governor).
Typically, ulus units were governed by voyvodas, and cemaats by kethdas, whereas tribes
who were registered into the army were supervised by seraskers. Like other administrative units in the empire, each confederation of tribes was also assigned a kad (judge),
who served as the direct representative of the central government and also adjudicated
in intra- and intertribal matters. As a further indication of the governments willingness to accommodate these communities, these kads would sometimes accompany
the tribes as the latter went through their seasonal cycles of migration.10 Even though
the titles of the officials who were in charge of tribes and those who were responsible
for peasant households and villages were identical, there were important variations
in the way in which the two sets of administrators were appointed. Perhaps more
importantly, while the central organization of the Ottoman administration and its
application in sedentary rural areas were highly centralized and hierarchical, there was
a strong element of bottom-up initiative and indigenous identification which shaped
the nature of Ottoman administration in tribal areas.
Of the indigenous tribes which the Ottomans came to dominate, the Kurdish communities constitute a special category. They were one of the largest ethnically distinct
and predominantly Muslim communities whose presence in this part of the world
long predated the Ottoman empire. The Ottomans were aware of the local power of
this community and their policies contributed to the long-term survival of the Kurds
as a distinct people. They used a policy of accommodation sometimes referred to as
istimalet, which consisted of making generous concessions to win over the Kurds,
while helping consolidate the power of local chiefs.11 This was the same policy that
was used towards Christian communities in western Anatolia and the Balkans in the
fourteenth and fifteenth centuries.12
In general, Ottomans favoured Sunni Kurds over their Alevi counterparts. Most
of the favours and preferential treatment were directed at them. Sunni Kurds were
also encouraged to form a buffer against Iran and as an ally in lengthy struggles with
the local Shii communities (Kurdish and otherwise), some of whom supported the
Iranian rulers. Some Kurdish chiefs took advantage of these conditions and used their
ties with the Ottoman government to amass fortunes and extensive power in eastern
Anatolia. So powerful did some of these chieftains become that they were able to
influence Ottoman policies and affect the shape of military campaigns in their areas.13
Murad IV () issued a series of imperial orders in and recognizing
the power of the Kurdish chiefs, reinforcing the hereditary nature of Kurdish tribal
chiefdom, and prohibiting local military commanders and governors from harassing
the Kurdish tribes.
Generally, nomadic tribes were exempted from many of the taxes and dues that
were levied on most peasant households. Even with the special taxes that were imposed
on them, the tax burden on nomads ended up being lighter than the obligations of
sedentary farmers. The most commonly imposed tax on all pastoral nomads was adet-i
agnam (sheep tax), which was determined on the basis of the size and quality of
the herds owned by a particular tribe or confederation of tribes. Nomads were also
required to pay a series of fees for their grazing lands and pastures, as well as special
dues if they harmed or lost their own or other peoples animals or slaves. Along with
the rest of the peasants, nomads also had to pay marriage tax (resm-i arus or gerdek
akesi). As was the general practice, most of these taxes were assessed and the obliga

R e s a t K a s a b a

tion of nomads was determined by taking into consideration whether the payer was
well off and also whether he was single, married, living with his parents or living
alone.14 Naturally, nomads were aware of their special status, and they used it in raising complaints against officials who sought to impose additional dues on them. In
such complaints they insisted that they were not tax-paying subjects (reaya) but had
special status because of the specific services they were performing for the sultan.15
Here, the special status and power of the tribal leaders became particularly important,
since they played an important role in negotiating with the government and affecting
the outcome of such complaints.


While the Ottoman land system was premised on the existence and preservation of
peasant households (ift-hane), this same system also preserved and incorporated the
large number of nomadic tribes and regulated their interaction with sedentary farmers. In Braudels words, far from being a residual category that was simply contained,
transhumance was institutionalized in the Ottoman Empire and was protected by the
state with special safeguards, rules, and privileges.16 Numerous laws and regulations
were issued by Ottoman officials to regulate the migratory routes of nomads and to
guarantee their livelihood and safety.17 Even the laws protecting the rights of peasants
against the incursions of nomads and their animals did not leave the nomadic population without any safeguards. In many regions, there were specially designated paths
along which the tribes were expected to herd their animals, but the peasants were also
required to mark their lands clearly and to build fences around them if these lands
were on the migratory routes of a tribe.
Before the eighteenth century, the settlement activities promoted and enforced by
the Ottoman government were limited in focus and purpose. Most typically, they
were carried out in order to (re)populate newly conquered areas in the Balkans and to
enhance security in frontier zones. Even then, tribes were not necessarily required to
abandon nomadism entirely. Otherwise, the only other large-scale resettlement carried
out before the eighteenth century targeted not the nomads but the peasants who had
fled in the face of the celali uprisings in the seventeenth century.18
Official tolerance towards tribes was rooted in the fact that pastoral nomadism had
become integral to the organization of the Ottoman empire in a number of different
ways. For one thing, land was relatively plentiful. This meant that in most regions,
especially in parts of western Anatolia, there was a shortage of labour, which became
particularly pronounced during harvest time. Migrant labour belonging to various
nomadic tribes moved over long distances to participate in various harvest activities in
the west. Workers coming from northern and eastern Anatolia could be found in and
around Bursa and Izmir. They routinely took advantage of the higher wages, which
could go up by as much as three times their normal rates during harvest season. Their
participation in harvest and other economic activities meant that most of the time
these tribes were functioning as transhumant communities in the west. Only part of
a tribe would move between the lowlands and the mountain pastures, while the rest
became involved in sedentary activities, including farming. Even those who moved
were not necessarily cut off from farming completely as they would incorporate it into
their seasonal migration. In the course of their migration they would cultivate the

Nomads and tribes in the Ottoman empire

abandoned or unclaimed lands (which could be plenty) that they encountered or work
as sharecroppers or seasonal labourers on lands cultivated by peasants in their regions.
The existence of these extensive relationships shows that the underlying system of settled agriculture was strong and that there was division of labour between this sector
and the migrating groups and individuals.
The interchange between villages and migrants could take even more complex
forms. For example, the collective ownership of land, called musha, that prevailed in
some of the Arab provinces, in particular in western Syria, did not grant permanent
ownership or possession to any single household that was part of this system. Instead,
individual parcels were periodically redistributed among the households, including
seasonal migrants who might be residing in the village. In this way, musha provided
the migrants with another means of entry into local economic networks.19 This shows
that, despite appearances, neither migration in rural areas nor the interaction of
nomads with local economy was random. The central government was aware of the
systematic nature of these relations and made sure to tax the nomads separately for any
settled farming in which they might be participating.
In addition to farming, nomads took part in manufacturing activities, such as carpet
weaving, rug making and other textiles.20 For example, around Jerusalem, both the settled peasants and nomads were engaged in the production of soap in the seventeenth
century, and the two competed with each other to obtain raw materials such as alkaline ash, which was a key ingredient.21 Some of the key export items such as natural
dyes and timber were gathered from the interior of Anatolia by nomads and sold to
merchants on major trade routes.22 Often, nomadic tribes were the sole purveyors of
animals such as camels, donkeys, mules and horses, which were the main means of
transportation for civilian and military purposes. We should note, for example, that
nomadic tribes were the sole suppliers of the more than , camels and as many
mules that the Ottoman army required on its campaigns. Their control of the empires
large and small animal stock meant that nomads were indispensable for the operation
of regional and imperial networks of trade. Also, the fact that they specialized in raising sheep made transhumant nomads the main suppliers of meat in the empire.
In this context it is important to point out the mutual dependence that characterized the relations between nomads and the settled villagers and townspeople. Just as
strongly as the sedentary groups, nomads were dependent on the commercial nexus
that linked them to other groups on their paths. They had to have access to regional
markets in order to obtain a wide variety of necessities, including food items, construction materials, horseshoes, and even some of their weapons. Such exchanges were a
regular part of the economies of various regions of the Ottoman empire.23
Nomads also provided some very important services in the integration and organization of the empire on a macro-level. For example, the so-called Arab camel drivers
could be found in most provinces of Anatolia, sometimes moving with their animals
across a large swathe of territory extending from the Dardanelles to Adana in the
south.24 Nomads facilitated the flow of goods and resources; they made it possible for
the Ottoman troops to move quickly across long distances, and they herded, gathered,
grew or manufactured valuable goods of consumption and trade. In times of famine and other natural or man-made disasters, the Ottoman government relied on the
services of nomadic tribes to move large quantities of grains and other foodstuffs and
necessities to disaster regions.25 Nomads were also employed in state-owned mines,

R e s a t K a s a b a

and in the construction of roads, bridges, forts and castles, as well as in the guarding
and protection of such structures.26
Nomadic tribes made an equally significant contribution to the military organization and success of the Ottoman empire in the early part of its history. It was by relying
on intrinsically mobile groups such as nomads, and unattached single men, that the
Ottoman government could mobilize large numbers of troops without threatening
the integrity and viability of the peasant economy. In fact, so central were their military activities that, in the fifteenth and sixteenth centuries, the Ottoman government
registered many of the yrk communities as military companies (msellem) ready to
be called to the front in times of war or provide support services in war or peace.27 In
some instances the Ottomans did not even have to channel or harness the movements
of local tribes but simply followed them. This was the case with the , nomads
who in the fourteenth century spontaneously moved and settled in the Balkans, where
they found willing partners among the semi-nomadic communities of Vlachs and
Albanians.28 On a broader scale and for a longer period, tribal groups were assigned the
task of guarding fortresses in the frontiers, bridges and major highways, all of which
were crucial in military campaigns as well as in peace time. In the Arab provinces the
Bedouin were paid special fees by the government so that they would provide some of
these services and, in particular, to keep pilgrimage routes open and secure.29
Not surprisingly, the importance and variety of the activities in which nomads were
involved and the crucial role the chiefs played in mobilizing their followers made some
of these leaders extremely wealthy, powerful and highly autonomous. One historian
estimates that even the poorest herdsman could be placed in the same category as a
peasant cultivating a full ift of dnms or acres of land. In , two
confederations in eastern Anatolia had become so wealthy that they paid million ake
in taxes to the central government.30

Like all itinerant groups in history, nomadic tribes in the Ottoman empire had the
option of using their mobility as a weapon. When they lacked the means to organize
a rebellion, they simply abandoned their designated pastures and paths of circulation,
left the towns where they were settled, and hid in the mountains or the countryside
to avoid the authorities.31 Alternatively, they would join other unemployed youth or
soldiers who had deserted or were discharged from the army, and together they would
roam the countryside, raid and rob villagers and merchant caravans. The end of the
sixteenth century was one such period of upheaval in the Ottoman empire, which
overlaps with a Mediterranean-wide increase in banditry, described by Braudel as an
explosion of liberation from the Mediterranean mountains.32
The Ottoman response to such actions was typically punitive, involving the forced
displacement of the individuals and tribes involved. But even such confrontations did
not necessarily imply a categorical opposition to nomadism on the part of the central
government, but rather a more focused response to ideological, political and/or foreign pressures. In any event, tribes which had been forced to relocate as punishment
did not necessarily stay in the places to which they were sent, but moved back and

Nomads and tribes in the Ottoman empire

forth between their places of origin and their new areas of settlement. For example,
in western Anatolia, the region around Saruhan, which was a very important source
of emigration to the Balkans from an early period, never completely lost its nomads
but continued to provide one of the end points in the back and forth movement of
migrants and nomads.33 Hence, in most instances, in the early part of the empires
history, pressuring nomads had the effect of spreading nomadism into an even larger
part of the empire, a process characterized by one author as reseeding of nomadism
across the Ottoman territories.34

The eighteenth century
From the late seventeenth century, the Ottoman empire responded to the new world
of expanding trade networks and territorial states by initiating a series of measures
intended to improve the empires security. These measures can be interpreted as the
first steps towards creating a functionally differentiated state structure that constantly
sought to improve its capacity to rule over an unruly and highly mobile society.
A key component of this shift in priorities was the growing interest on the part of
Ottoman officials to count, register and, ultimately, settle the nomadic and other
itinerant groups within the borders of the empire. It was becoming clear that, in the
new modern world, the territorial and political uncertainty that came with nomads
and migrants was no longer an asset. In fact, such fluidity was fast becoming a
The turning point in the empires relations with nomadic tribes and other mobile
groups came in during the short grand vezirate of Kprl Fazil Mustafa Paa,
when the first set of comprehensive orders to settle nomadic tribes were issued. These
were followed in with another series of directives that aimed to register as many
of the nomadic tribes as possible.35 In addition to monitoring the movements of these
communities, the registration drive helped create specially designated companies who
were then drafted into the army.36
The urgent need to reinforce the army during the war of the Holy League in Hungary () was an important factor which informed these particular measures.
However, the policy of settling nomads did not end there. The following two centuries
of the empires history witnessed a steady stream of orders and decrees dealing with
various problems related to nomadic populations, migrants and refugees and their settlement. These orders and the underlying policy that shaped them became clearer and
better focused as time went on.
An important aspect of the new wave of centralization in the seventeenth and
eighteenth centuries was the qualitative and quantitative expansion in the kind of
information the Ottoman government sought. This implied a new approach to rule,
and expansion of the powers of the civilian officials.37 As part of this new development, almost as soon as the settlement decrees began to be issued, a new Office of
Settlements (iskan dairesi) was established in the palace, and in all such initiatives were placed under its authority.38 During the following two centuries, the offices
that focused on the settlement of tribes and refugees expanded, and their powers

R e s a t K a s a b a

multiplied. This subtle shift of power towards civilian authorities created a deep sense
of unease among the military and led to a series of revolts in the eighteenth century.
Cognizant of established practices and expectations, and being aware of the fact
that the very mobility of the tribes gave them an important means of protection and
the possibility of evasion by flight the central government approached sedentarization cautiously. Rather than relying only on force, officials tried to make the option
of settlement and joining the army attractive by offering a wide range of incentives
to those tribes who agreed to change their status. Such tribes were exempted from all
regular duties (nomadic or sedentary) for several years at a time. In addition, they were
granted free seeds and oxen, better access to sources of irrigation, and other encouragements and subventions that were tailored specially for different regions. In return for
these privileges the nomads were typically required to abandon their peripatetic lives
and assume sedentary farming. They would pay a one-time special tax and desist from
attacking or harming villages and the villagers in their new areas of settlement. In some
parts of the empire, nomads obtained even broader privileges if they agreed to plant
certain crops that were highly valued and/or needed, such as wheat and cotton in central and southern Anatolia. The Ottoman government also offered similar incentives
to fugitive peasants who had quit their lands and were roaming the countryside.
Even though such encouragements could be attractive, tribal members resisted registration and sedentarization, since these were usually followed by additional demands
for taxes and military service.39 It was common for them to ignore, deflect, subvert
or resist government orders. While it was usually the tribal chiefs who complained
and wrote the petitions, it was not uncommon for individual members of a tribe to
take the lead in such matters. The most common causes for complaint by the tribes
was the distance they would have to travel, the size of the plots given to them, and
the limited resources made available. Local officials reviewed such petitions and usually referred them to higher-ranking bureaucrats. Not surprisingly, the desert areas of
Rakka in Syria where a large number of nomads were forced to settle formed a particularly strong source of discontent for the tribes that were sent there. Lack of water
and limited vegetation were the main factors that the tribal leaders cited in demanding
a change in their orders of settlement. Occasionally, tribes mobilized a large number
of their members to resist the central governments order, not only by fighting Ottoman forces but also by making their places of settlement inhospitable by burning and
destroying the existing fields.40 Whether tribes would rebel and whether local notables
would support the rebellious tribes depended on local conditions. The particular position taken by a tribe did not necessarily stay the same either. In response to changing
circumstances, a tribe could switch from being rebellious to being supportive of the
government, or vice versa.41
In trying to settle the nomadic tribal communities and others, the Ottoman government also clashed with villagers whose interests were threatened by the introduction
of new groups in their midst.42 An important part of this conflict was rooted in the
ownership patterns in Ottoman agriculture, which had created several distinct groups
with equally strong and competing claims of rights, especially in the more fertile areas
of Anatolia and Mesopotamia. In addition to the peasants who had been working
the land, these groups included tax farmers, absentee landlords, and members of the
sultans household, all of whom had become de facto landlords with the introduction
of life-long leases in .43 In a pattern that would repeat itself many times in the

Nomads and tribes in the Ottoman empire

eighteenth and nineteenth centuries, absentee landlords could agree to government

policies of sedentarization and pacification for their own reasons, while the peasants
who were the actual cultivators in possession of the land strongly opposed the arrival
of refugees and the settlement of nomadic tribes. In fact, most of the complaints of
which records have survived, and which provide information about the processes and
problems of settling nomads, were filed by villagers. Especially in western Anatolia, the
common claim was that the land was already congested and that settling new groups
would make life very difficult for people already living there. It was not uncommon for
the Ottoman government to heed these complaints and reverse its position.44
Even though there were regional successes, overall, the sedentarization policies of the
seventeenth and eighteenth centuries did not produce a permanently settled society.
In fact, by the early nineteenth century the trend had once again turned in the opposite direction, with nomadism and nomadic tribes growing. While exact numbers are
impossible to come by, one contemporary account estimates that, on the eve of the
mid-nineteenth-century tanzimat, there were, just in Adana, as many as , nomads
compared with merely , settled peasants.45 Rather than shrinking, some of the
tribes were steadily expanding their power as well, some behaving as if they had all but
seceded from the empire. This development was caused, in part, by the impossibility of
imposing a uniform pattern of settlement across the empire from Europe to Asia.

The nineteenth century

In the early nineteenth century, external wars and internal disorder made it harder
for the central government to carry out its directives for creating a more sedentary
empire. In particular, the inability of the Ottomans to fill the vacuum left by the Russians after the war allowed local notables and Kurdish chiefs to expand their
power in the east. Some Kurdish tribes quickly gained control of a series of citadels in
the border areas of eastern Anatolia, creating a region that was effectively cut off from
the rest of the empire. They then used these as bases to attack travellers, traders and
peasants.46 One of these Kurdish chiefs, Badr Khan, eventually controlled the entire
region between Diyarbakir and Mosul, becoming so powerful that in the s and
the s he issued coins in his own name.47 Even though they were expanding their
power, building palaces, maintaining armed forces, and even establishing dynasties,
the Kurdish tribes never abandoned their pastoral nomadism completely. They continued to circulate in and around their regions, and fought with each other in order to
gain advantage and keep alive their tribal identities. In this way they could also keep
the centre at bay, being always prepared to revert to their nomadic lives if conditions
changed. This made it nearly impossible for the Ottoman empire ever completely to
subdue and control the Kurdish tribes in the east.
In the period between the abolition of the Janissaries in and the creation of
a new professional army, the Ottoman government carried out a series of campaigns
using former soldiers and local irregulars. The purpose of these campaigns was to
pacify the border areas that had become all but ungovernable. In particular, the government targeted the local notables of the Black Sea region, rebellious tribes in Libya,
Kurdish emirates in northern Iraq, Kurdish families in the east and the south, and the
Bedouin who had been pushed north to Syria during the eighteenth-century Wahhabi
uprisings and the later campaigns of the Egyptian armies.48

R e s a t K a s a b a

Among all the regions and groups with whom the Ottomans had to deal, they were
particularly challenged by the situation in the Taurus region in the south. To deal with
the large confederation of tribes which all but controlled of a large swathe of territory
here, the Ottoman government created a special fighting force called frka-i slahiyye
(the army of reform), consisting of former soldiers and Albanian, Zeybek, Georgian,
Circassian and Kurdish fighters who were recruited locally.49 Even though it was set
up as a military operation, the campaign by the frka-i slahiyye was carried out as a
comprehensive programme of pacification and sedentarization which also created new
towns and novel methods of administration; all were designed to strengthen the ties
between the centre and these regions.
Staging military campaigns was only one part of the Ottoman response to the growing power of tribes and local families. Over time, these were joined by new laws and
new institutions that gave the military and other steps a higher degree of consistency.
In a government publication summarizing all laws and regulations passed between
and , it is stated that all the nomadic tribes in Anatolia would henceforth
be settled in their winter pastures, and that they too would be included in the imperial
censuses. From then on, tribes would be expected to engage in agriculture and pay
their taxes accordingly.50 In , in a further attempt to undermine local centres of
power, the government deprived the tribal sheikhs from the official recognition that
had been accorded to them, and incorporated them into the administrative hierarchy
of provincial administration with titles such as kaymakam and mutasarrf.51
These regulations concerning nomads and tribes were accompanied by several laws
designed to strengthen the overall institutional framework of the empire. One such
initiative that aimed both to enhance the power of the central government and to protect settled peasant agriculture was the land code enacted in . While the overall
thrust of this law was to undermine the large landholding notables and to undo the
harmful effects of the tax-farming system, it was also a clearly and strongly anti-tribal
measure. Most significantly, in a direct blow to the foundations of tribal life, the
law replaced communal property and identity with the principle of individual ownership. Individual households could take advantage of this and acquire the lands which
they had already been cultivating, by registering with the central government.
The central government also requested reports about the behaviour of tribes following their (re)settlement. If tribes persisted in unruly behaviour, they would be moved
further away and punished in new ways. The Kurdish tribe of Hemvend was one community that was the subject of repeated government orders and attention, especially
in the last quarter of the nineteenth century. After causing numerous problems in
the region between Baghdad, Mosul and the Iranian border, and after some of their
leaders escaped from Yemen where they had been imprisoned, they were divided into
groups, and one part was forced to move in to the Ikodra region of Albania.52
Those who remained in the area but refused to settle near their places of circulation
around Mosul were pushed towards Adana, Ankara, Sivas and Konya.53 At the same
time, specific orders were given for the confiscation and selling of their animals, so
that they would have nothing to come back to.54 Another group of the Hemvends who
continued to organize raids between Mosul and Kerkk were ejected and settled in
Cyrenica.55 In the execution of this last order, the central government kept a particularly close watch on the Hemvend as they moved from Mosul to Albania and Libya, by
way of Izmir.56 In , because of the growing cost of moving the Hemvends around

Nomads and tribes in the Ottoman empire

the empire, the government ordered that some of their animals should be sold and the
money raised be used to meet some of these expenses.57
Other Kurdish confederations that have played prominent roles in the modern history of Iraq were also a source of concern for the Ottomans in the nineteenth century.
In , Ismail Paa, vice-director of the Ottoman Military Inspectorate, was given
special powers and the public prosecutor in Mosul was ordered to work with him, so
that the Berzenci and Talabanli tribes would be forced to cease attacking each other
and the banditry in the region could be brought under control.58 Several months later,
the central government was alerted once again to the ongoing conflict between these
two confederations.59 Two months after this, Ismail Paa reported that the trial of the
leaders of the two confederations was completed. Their weapons were confiscated and
they were ordered to abandon their nomadic and bandit ways.60 In executing these
orders, the government continued to use force. According to one estimate, between
and , , tribal members were killed as they were being pushed to settle
across the empire. This gave rise to a very rich folklore of pain and persecution at the
hands of the state, and remained a central motif in the culture of these tribes long after
they took up sedentary life.61
Available data suggest that the number of settlements in tribal areas, and villages
that were associated with specific tribes, steadily increased during the course of the
nineteenth century.62 It seems also that the central government took special care to
create as many mixed villages as possible, in order to break the cohesion of tribal solidarities as they took up settled life.63 As a corollary to this trend, data also show that
the nomadic population declined in the Ottoman empire during the same period.
In Anatolia, the proportion of this decline is calculated as per cent between
and .64 In the Arab provinces there was a similar decline. Around the province
of Basra, the proportion of nomads declined from to per cent between
and ; around Baghdad, the ratio was and per cent respectively.65 However,
as happened in earlier periods, this era of relative settlement would also unravel in
a period of internal strife and external warfare in the early decades of the twentieth
century. Under such circumstances, the tribes rallied their people and restored some
of their nomadic practices, re-creating a widely diverse and fluctuating picture for the
Ottoman empire.

The waves of sedentarization policies and their outcome in the eighteenth and nineteenth centuries show that there was no linear or definitive transition from nomadic
to settled order in the Ottoman empire. More typically, in many parts of the empire
we see a cyclical alternation between nomadism and sedentary life, sometimes involving the very same communities. For the most part, these cyclical patterns and transitions were maintained and reinforced by the continuing strength of tribal relations.
At least in part, the very policies of sedentarization conceived and implemented by
the Ottoman government ended up supporting this fluctuating pattern. For example,
the Ottoman state frequently resorted to the old tactic of using some tribes to punish
or settle others. A well-known example of such a policy is provided by the constitution and operation of the Hamidiye regiments. Formed in , these regiments
were staffed exclusively by Sunni Kurds, who were given a free hand and strong

R e s a t K a s a b a

government backing to pacify Alevi Kurds and Armenian and other non-Muslim
minorities in eastern Anatolia. The tribes recruited into the Hamidiye stayed within
their own tribes and were led by their chiefs. Their number increased from about
forty regiments in to sixty-three in .66 Some of these chiefs created local
networks of administration that looked like the replicas of the imperial institutions.
Ibrahim Paa of the Milan tribe, who was made a Hamidiye commander, became so
powerful that he was referred to as the uncrowned king of Kurdistan. During the late
nineteenth and early twentieth century, the Hamidiye became very well known for
their brutality, raiding the property of rival Kurdish tribes, especially Armenians, and
pushing them out of their land.
In most cases state officials had little choice but to rely on tribal leaders to carry out
their policies in various localities, and were forced to negotiate with the very elements
whose power they were trying to curb. For example, it almost always fell on the tribal
chiefs to guarantee that their tribes would stay at their designated places of settlement.67 If they did not, the chiefs would bear the brunt of punishment. In return, the
chiefs were given a higher quality of land and granted privileged access to higher echelons of bureaucracy; they themselves were rewarded with posts, salaries, medals and
lump sum payments. The old practice of allowing them to settle in familiar grounds
around their winter pastures and also of making provision for the continuation of
annual husbandry were some of the other concessions that the central government
made in its dealings with tribes. Sometimes the very act of settlement helped reinforce
tribal organization and identity, since the tribes now had to deal with new contingencies such as gaining access to water, which could only be achieved as a group and
even in cooperation with other tribes.68 In such instances, instead of being subsumed
under the centralizing power of the state, while serving the government, tribal chiefs
enhanced their position and became even more powerful and indispensable. The longterm importance of tribal and nomadic activities for the Ottoman empire and the particular policies the Ottoman state followed over the years guaranteed mobility a place
in the social organization of the empire. Furthermore, migratory activities survived
the transition from empire to nation. As a result, Turkey and other new states in the
Middle East during the twentieth century continued to govern large and occasionally
restless mobile groups.


Cook : ; Islamoglu-Inan : .
Orhonlu : ; Planhol : .
: ; Planhol : ; Orhonlu : , n. .
Inalck a: .
etintrk ; Inalck .
: .
Inalck : ; Van
. Bruinessen a: .
etintrk : ; Inalck : ; Smer : ; Emecen : .
Gndz : ; Planhol : .
etintrk : ; Orhonlu : .
Van Bruinessen a: .
Lowry a: .
Evliya elebi : , ; Sinclair .

Nomads and tribes in the Ottoman empire

Smer : ; Orhonlu : ; etintrk : ; Inalck b.
Ahmet Refik a: , , documents , .
: , .
Inalck : .
On the celali uprisings, see chapter , by Oktay zel, in this volume.
Inalck : ; Smer : .
: .
Inalck : ; Smer : .
: ; Singer : .
Inalck : .
Ahmet Refik b: .; Ger : ; Orhonlu : , n. .
Ahmet Refik a; Inalck ; etintrk .
Gkbilgin .
Barkan , Part II: , ; Adanr : .
Zeevi : .
Murphey : , .
Barkan : .
Braudel : I, .
Barkan , Part II:
. ; Gken .
Planhol : ; cf. Inalck : .
Two crucial sources for the early policies of sedentarization are Ahmet Refik a and Orhonlu
. Ahmet Refiks book is a compilation of orders issued by the central government in order
to administer and settle nomadic tribes. Although not a comprehensive list, it is nevertheless a
valuable source, containing orders covering the period between and .
Gkbilgin .
Abou-el-Haj .
Orhonlu : .
Ahmet Refik a: , , documents , .
Orhonlu : .
Khoury : ; Masters : .
Ahmet Refik a: .
On life-long leases, see chapter , by Ariel Salzmann, in this volume.
Orhonlu : .
Gould : .
Fraser : I, , , .
Von Mltke : .
Marufoglu: : .
Cevdet Pasa : .
In Kavanin ve nizamat mecmuas, cited in Marufoglu : .
Marufoglu : .
Ibid.: .
BOA-MV (Basbakanlk Osmanl Arsivi, Meclis-i Vkela Mazbatalar) /, / (, Za
BOA-MV / ( Za ).
BOA-MV / ( Re ).
BOA, Dahiliye Nezareti, Muhaberat- Umumiye dare Kalemi / ( C ); Marufoglu
: .
BOA-MV / (, Z. ).
BOA-MV / (, S. ).
BOA-MV / (, Ra ).
BOA-MV / (, Za ).
Gould : .

R e s a t K a s a b a

Ibid.: ; Lewis : , .
Gould : .
Ibid.: xv.
Nakash : .
Van Bruinessen a: ; Klein .
Orhonlu : , n. .
Nakash : .




Rhoads Murphey

n order to gain a sense of the base economic conditions prevailing in the western
Anatolian hinterlands of cities such as Bursa, which after became the nucleus
of the fledgling Ottoman state, it is instructive to start with the vivid and evocative
details provided in the narrative of Ramon (or Raymond) Muntaner detailing the
marauding activities of the Catalan companies in the Marmara region between
and . The overriding concern of the bands of military enterprisers who came to
dominate frontier society in the late thirteenth century was to maximize opportunities
for short-term profits within a given locality, and then to move on to new sites for
exploitation when local resources gave out. This point is made explicit on a number of
occasions in Muntaners account. In chapter , relating the aftermath of Frey Rogers first series of campaigns in Anatolia, Muntaner makes the following observation:
The Boca Daner [i.e., the anakkale Boaz/Dardanelles] is surrounded by good
and fertile places in all parts [i.e., the Biga Yarmadas/peninsula stretching eastward towards Bursa]. You will find that, on each side, there was a very fine town
and a very fine castle at the time we went there. All has been destroyed and ruined
by us.1

Following Frey Rogers assassination in , separate commanderies and captaincies

were established in Thrace. Each drew its livelihood from small-scale, largely uncoordinated raids aimed mostly against Byzantine towns and settlements along the north
shore of the Marmara. Muntaners boastful acknowledgement of the anarchic conditions that prevailed in the Byzantine borderlands in these years provides sometimes
startling reading. In chapter he notes: We sowed nothing nor ploughed . . . but
took each year as much . . . as we wanted . . . So we lived for five years on aftercrops
and the most wonderful raids were made man could ever imagine.2
He then relates the division of the spoils at the conclusion of a minor skirmish with
Byzantine forces in the vicinity of Gallipoli.
We took thirty-seven horsemen and . . . [a number of] foot. The next day in
Gallipoli we had an auction of the horses and prisoners . . . and of the booty

Rhoads Murphey

twenty-eight gold hyperpers for each armed horse and fourteen for each light
horse and seven for each foot soldier, so that every one had his share . . . This was
done not through our worth, but by the virtue and grace of God.3
In chapter Muntaner returns to the theme developed in chapter and once
again reassures his audience that, as a point of honour and regimental pride, for five
years we never sowed or planted or dug over the land.4 In chapter , he explains
why he was compelled to abandon his base at Gallipoli in favour of a new location
south of Salonica.
It is the truth that we had been in the peninsula of Gallipoli and in that district for
seven years since the death of the caesar [i.e., Frey Roger] and we had lived there
five years on the land and there was nothing left. And so, likewise, we had depopulated all that district for ten days journey in every direction; we had destroyed all
the people, so that nothing could be gathered there. Therefore we were obliged to
abandon that country.5
Muntaners final summing up of the activities of the company in the Aegean and Marmara basin is offered in chapter .
When I parted from the company . . . they went to a peninsula called Cassandra.
. . . At the entrance to the peninsula they pitched their tents and from there they
raided as far as the city of Salonica . . . And they consumed that district as they had
done those of Gallipoli, Constantinople and Adrianople.6
The general ethos described in Muntaners account of the Catalan company a
body run on roughly similar lines to modern business enterprises based on profitsharing principles was not of course confined to the particularities of one tradition, whether Muslim or Christian. Nor does it typify conditions prevailing in just
one geographical sector. What Muntaner describes applied broadly, perhaps even
universally, to the general lawlessness of frontier society in the interstices between
the late Mongol and Byzantine Ouikomene or, in Hodgsons terms, the Inhabited
Quarter.7 In earlier centuries the region had developed citied environments possessing well-established urban institutions and regulatory mechanisms sufficient to meet
Hodgsons definitional criteria. However, throughout the era of declining imperial
control between the mid-thirteenth and mid-fourteenth centuries, substantial parts
of Western Asia had lapsed into the conditions of near anarchy evoked in Muntaners
In their endeavours to build a flourishing economy from such a low base, the early
Ottoman sultans actively pursued a number of policies to encourage and protect
agricultural and commercial development and to foster the growth of urban economies. The following discussion takes for granted such well-known features of Ottoman revenue collection and distribution practices as the timar system, poll tax and
customs levies.8 Instead, it assesses the degree to which early Ottoman sultans were
able successfully to pursue certain economic policies and the considerations behind

The Ottoman economy in the early imperial age

Figure . Historical photograph showing the covered market (old bedestan) of Istanbul in the
foreground. The core of these buildings was completed during the reign of Mehmed II
between and . Photograph by Ali Rza Paa (d. ). In the collection of the
Library of Congress Prints and Photographs Division (neg. LC-USZ-).


The Ottomans and Byzantines endured a century of uneasy co-existence between the
Ottoman landing at Gallipoli in and the eventual fall of Constantinople in .
The generalized breakdown of law and order in provincial society at the margins was
a matter of concern for both regimes, but it was arguably the Ottomans who achieved
greater success in imposing the rule of law and establishing the order and predictability
necessary for economic stability and growth. Ottoman attempts at regulation in the
early imperial era covered both the rural and the urban spheres. Attention was paid
to the monitoring and control of population movement through the land registration
regime, organized by means of the tahrir system, the periodic imposition of forced
transfers of tribal and other groups to support settlement and development objectives
in newly subdued territories, and the organized transfer of skilled workers to urban
centres (the srgn mechanism). Above all, it was success in governing urban markets
through the strict application of the ihtisab regulatory regime that became a distinguishing feature of the Ottoman regime in the early imperial era.

Rhoads Murphey

Current scholarship tends to reject the consensus view previously held among historians that exaggerated not just the extent of the early modern states dominance in
political matters but also its efficiency and reach in economic regulation.9 The limits
of state power applied most markedly in attempts to regulate economic activity in the
rural agrarian sphere, which accounted for between and per cent of economic
output in most parts of the medieval and early modern Balkans and Middle East.
Testing the limits of state power is problematic and quantifying its actual extent in
economic terms remains elusive. In his study on the statistical reliability of the Ottoman land and population surveys based on a limited sample from western Anatolia,
Michael Cook speculated that, despite its reputation for clarity and comprehensiveness, the Ottoman tahrir system, even in its developed state of the sixteenth century,
seemingly missed (i.e., failed to record) around a quarter of the population.10 As an
instrument even of state fiscal control, let alone regulation of economic practice and
the peasant producers behaviour, the tahrir system was thus far from perfect. Particularly in this latter sphere, of regulating the behaviour of the main agents (peasants,
pastoralists and others) who operated in the rural context, the depth of penetration of
the states legislative and regulatory impact has been convincingly questioned.11 The
weight of custom, tradition and time-honoured practice proved difficult for the state
even to modify, let alone discourage or dislodge, despite the best efforts of the land
surveyors, tax assessors, agronomic advisers and other would-be regulators it deployed
in seeking to alter peasant behaviour to conform with state expectations and administrative orders.
It would be indefensible to ignore or minimize the scale and importance of the
Ottoman states efforts to sedentarize nomads, regulate and tax peasants, and generally
transform the rural landscape by extending its sphere of influence over the pastoralists and cultivators of the soil who made up the bulk of the Ottoman population.
However, in realistic terms, and recognizing the real limits of Ottoman state power in
the fifteenth century, it was from the narrow base of the more closely regulated urban
space and urban markets that the Ottomans launched their first and most effective
efforts aimed at modifying undesired market tendencies such as hoarding and price
speculation, and at creating the basis of a fair balance between mercantile profit and
affordability for average urban consumers. Indeed, they staked the reputation and
legitimacy of the new dynastic regime on the success of their regulatory influence in
urban settings. The primacy and seriousness of the Ottoman commitment to enforcement of fair trading practices in urban markets comes out with striking clarity in
surviving early documents.


By the sixteenth-century high imperial era, order, security, discipline and regulatory
consistency had become the general hallmarks of Ottoman administration, through
the predominance of kanun (sultanic administrative law). As part of this, from the very
outset of the imperial venture the Ottomans had taken special care to lead by example
and to protect their record for good government in cases where the sultans own personal or palace household interests were involved. A sultanic order of , addressed
by Mehmed II to his commissioner in Bursa in charge of special procurements, makes
clear this inward-directed orientation of Ottoman regulatory effort.12 The terms of

The Ottoman economy in the early imperial age

appointment specified for the sultans agent define explicitly the limits of his proxys
jurisdiction and powers, emphasizing that it was both necessary and appropriate that
the appointee be issued strongly worded warnings to observe open market access rules
and other regulatory restraints. Despite the lofty status of his principal and the highpriority nature of his business, the agent is neither to demand nor to expect treatment
on terms other than those offered to any prospective customer in the marketplace.
Particular care was taken to avoid creating the impression that either the sultan or the
agent acting on his behalf operated above the law or laid claim to special privilege or
economic advantage from which the general public was excluded. This self-denying
ordinance applied with particular force with regard to access to basic goods, especially
You are to refrain from seeking preferential terms, favourable treatment or imposing forced sales on those who bring grain for sale at the grain market by invoking
the phrase we are buyers for the state and imperial household. You are to purchase goods on the market on the terms offered to all other prospective buyers and
you should realise your purchases using reliable intermediaries, avoiding any acts
that may cause injury to the general public. In short . . . whether with respect to
cloth purchases, grain purchases or the purchase of any other goods [for the imperial household] you are to execute your purchases and requisitions with absolute
integrity . . . striving to avert either the waste of imperial resources on the one
hand or harm being visited on the public on the other. All purchases and transactions should always be carried out under regular auspices and scrutiny, using the
good offices of the local justices (kadis) and police magistrates (subas).13
The operative word in the Turkish text is dkeli, meaning always or as is the usual
norm,14 and implying that the insistence on procedure and strict adherence to marketplace regulations is either of long-standing application or at least an ideal on whose
reintroduction the Ottoman regime places particular priority. It is worth noting that
the principle of the sultans accountability to the same standards and norms of marketplace behaviour as both his kuls and his other tax-paying subjects is given forceful
expression in a document reliably attributed to perhaps the most temperamentally
autocratic Ottoman sultan of any era.
The details of Bayezid IIs ihtisab kanunnamesi (marketplace regulations) for Bursa,
dated , provide another revealing example of the motivating spirit behind Ottoman attempts to assert rights in the regulation of social relations at the local and
municipal level. Both in the preamble to the ordinance and in subsequent chapter
summations, it is made abundantly clear that what prompted the document was the
sultans awareness of recent departures from previous high standards of marketplace
behaviour and socially responsible trading practice. He was compelled to make a thorough review of the current situation and to lend the weight of his authority to the
prompt restoration of previous good practice, whose exact requirements are revealed
in minute detail in the body of the text. In the preface he says:
It was previously ordered that a careful record be kept of all changes introduced
to the price control regime of Bursa from the time of my accession [] to
the present [] with a view to establishing the specific circumstances of the

Rhoads Murphey

application of the price control regime for each type of good and commodity.
Have the price assessment criteria been applied consistently over that time and,
if not, when and under what circumstances were changes to assessment protocols introduced? In the case of variation, what is the basis of the new system
of assessment? On all of these matters you [i.e., the kadi (judge) and the muhtesib (marketplace inspector)] were previously instructed to keep yourselves fully
informed, omitting no relevant evidence whether large or small and to report
to me as necessary in full detail so that your assessments and evaluations might
serve as the firm basis for establishing the authorized regulations (kanunname)
and be used as a reference and guidance when required. In accordance with the
exulted and most noble order (may [its efficacy] never cease), the senior officials of
each craft and trade were summoned one by one and each questioned concerning
the character of customary assessment conventions for each product and asked
whether these traditional protocols and controls were currently being observed or
had been subjected to change over time. They were further prompted to explain
the reasons behind any changes introduced and the exact dates and circumstances
when these changes were adopted. In response to this summons, the guild officers
responded and replied unanimously saying that for the past five or six years [i.e.,
since approximately ] no trace has remained of the customary pricing regime
previously applied to the regulated crafts and trades. The whole price regime to
its very foundation has been fundamentally altered, changed and corrupted such
that at present virtually no one applies the customary standards observed of old
for determining prices. In view of the current unsatisfactory and confused state of
market regulation in Bursa [around the year ] I hereby order a new inspection of crafts and trades to be carried out starting with the bakers [and continuing
through the ranks of all the other craftsmen and traders present in the city].15
Strongly worded statements of regulatory intent such as those preserved in the documents of and demonstrate Ottoman resolve to achieve consistency in the
enforcement of codes and regulations promulgated by the imperial authority. While
such proclamations cannot serve as proof of Ottoman success in their stated objective
of full regulatory control over urban markets, they provide at least strong evidence
of the firmness of their legislative intent. Apart from such pronouncements of lofty
principle, there is also plenty of corroborative evidence to suggest the governments
close engagement with local economic agents and actors at the municipal level and its
effective intervention to broker compromises between the potentially diametrically
opposed interests of producers and retailers on the one hand and common consumers
on the other.


Just as in the political and administrative spheres it is premature to attribute absolute authority to Ottoman rulers before the capture of Constantinople, it is equally
clear that mechanisms of control affecting various spheres of the economy monetary
policy, control of trade routes, centrally monitored investment strategies had evolved

The Ottoman economy in the early imperial age

only partially by . On the most basic level, connection between the two halves of
the empire in Anatolia and Rumeli was, until the restoration of Ottoman control over
Salonica in , maintained only tenuously throughout much of the period. As late
as , while en route from Anatolia to assume command at Varna on the Danube,
Murad II still relied on help provided by naval allies situated along the shores of the
Sea of Marmara and the Bosphorus in order to avoid the Venetian blockade that had
effectively closed the Dardanelles. The Chronicle of Kemalpaazade (d. ) comments several times on the landlocked position of the Ottoman polity during Murad
IIs reign: Sultan Murad came to Gallipoli where he realised that the presence of the
Venetian fleet barred any possibility of crossing there.16 The Anonymous Chronicle
provides a similar assessment of the sultans position:
Murad arrived at Gallipoli where the fleet of the Frankish infidels had dispatched
a large number of ships to close off the sea passage. In the end Sultan Murad
managed to cross to Rumelia [i.e. Europe] on board a Genoese vessel from
The subsequent unification of Ottoman territories in Europe and Asia by Mehmed II
was of considerable economic as well as political significance. It was achieved through
a series of campaigns, beginning with the conquest of Constantinople in and
finishing with the defeat of Uzun Hasan, ruler of a territorially extensive, rival Anatolian Turkmen state, the Ak-Koyunlu, in August at a major battle at Bakent
on the plain of Tercan near the banks of the Upper Euphrates. This was followed in
by subordination of the Genoese community in Caffa in the Crimea, which
secured Ottoman organizational (if not yet navigational or commercial) control over
the Black Sea maritime zone. Thereafter, the Ottomans occupied a commanding position over an extensive trade network linking the coasts of the Aegean, the Marmara,
and the Black Sea. The strength of the Ottoman economy in this period is reflected
in Jacopo da Promontorios impressive list of Ottoman assets and revenue sources
dating from .18 But in , a mere twenty-five years earlier, this was a world
still undreamed of. In the early fifteenth century, the significantly fragmented Ottoman state was composed of semi-independent or only partially linked regions. One
of the inescapable results of this political framework was an economy characterized
by autarchic norms and practices, against which the centralizing influence of the state
authority and state regulatory mechanisms was only beginning in partial form and
at a gradual pace to make inroads. The chronicler Akpaazade (d. after ) provides the example of the centrally located province of Saruhan in western Anatolia,
where Bayezid IIs attempts to impose export restrictions on salt were met with stiff
resistance and, in effect, ignored: in that district [i.e. Saruhan] there were trading and
transport restrictions imposed on salt but the [indigenous tribes] paid no attention to
these restrictions.19
Governed by this sense of the pervasive parochialism of the eastern Mediterranean
in the late fourteenth and early fifteenth centuries, our account of Ottoman economic
life in the pre-conquest era (i.e., pre-) will place particular emphasis on continuity of practice between the early Ottomans and their Anatolian predecessors, the
Seljukids and the Turkish beylicates.

Rhoads Murphey


The products of the pastoral economy formed a very significant element of the western
Anatolian export profile in the fourteenth century. The trade in horses between the
south-western Anatolian principality of Mentee and the Venetian island of Crete provides clear evidence of the magnitude of this activity contemporary with the early years
of the Ottoman emirate.20 While it is difficult to judge the importance of particular
trading commodities from the limited sources available for the period before , the
few surviving documents provide strong indications that wealth and prosperity in the
early Ottoman state tended to be measured in terms of herd size. Orhan Gazi indicated
his preference for his son Murads succession in a pre-decease disposition of his property drawn up in July by assigning eighteen mares (yund) as his share.21 Even after
the siege of Constantinople and the conclusion of the Veneto-Ottoman war of
, a contract for customs of the market zones of the capital (i.e., Gelibolu [Gallipoli]
and skdar) from reveals that livestock in transit (horses, camels and sheep) still
accounted for a significant part of the nearly , ducat annual revenues included
in the contract. The actual value of the contract was . million akes, equivalent to
, ducats at the rate of akes to the ducat.22 Empire-wide (even accounting for
a certain degree of exaggeration in Jacopo da Promontorios survey report of Ottoman
revenues), the tithe on just three categories of animals horses, oxen and mules and
excluding sheep, by far the most significant revenue-producing category of animal,
yielded five times this amount, or roughly , ducats.23 Ottoman commercial
activity was not limited to animals and animal products, but overlooking these, alongside other sources such as salt and mining works, as important elements in the states
revenue make-up results in a serious distortion of economic realities.
A developed transit trade in luxury goods, from both Egypt and Iran, was a legacy
inherited by the Ottomans from their Muslim predecessors in Anatolia. Bursa, in Ottoman hands for less than a decade when visited by Ibn Battuta in , is described by
that traveller as a great and important city with fine bazaars and wide streets.24 Bursa
remained the economic and political capital of the Ottoman state well beyond the
end of the interregnum in . Court life shifted definitively to Edirne only with the
reinstatement of Murad II on the throne in and the celebration there in of
his son Mehmeds royal wedding. Until then, the preference for Bursa as the place of
residence for the principal governors of the realm, and of the sultan himself, is clear
in near-contemporary historical records.25 Bursa was also favoured by its geographical position close to the most important eastern Mediterranean trading centres of the
Genoese, at Pera, Phocaea (Foa) and Chios in the Aegean. Throughout the period,
from the first GenoeseOttoman commercial treaty in (confirmed and expanded
in ), and despite Mehmed IIs offensives from the mid-s against the Gattalusi
family in the Aegean, the Ottomans worked in close commercial co-operation with the
Genoese. Recent research suggests that, while Mehmeds establishment of control over
the Dardanelles after compromised the integrity of the Genoese trading empire
in the eastern Mediterranean and Black Sea regions, Italian mercantile interests nevertheless maintained and even expanded their pre-eminent position in Ottoman
markets until at least the s.26
The transformation of the economy according to Ottoman precepts and priorities
begun under Mehmed II was demonstrably less comprehensive in some regions of the

The Ottoman economy in the early imperial age

empire than in others. The active participation of Muslim merchants from Anatolia
in the Black Sea trade following the establishment of Ottoman control over Amastris/
Amasra in , Sinope and Trabzon in , and the Crimean port of Caffa in
is well documented in the customs register of Caffa from c..27 It appears, however,
that in the early fifteenth century there was little outlet for such activity. Subject to the
restrictions imposed by their states territorial confinement for much of the period to
Anatolia west of the Sakarya and Bulgaria south of the Balkan range (with only the
most tenuous connection between the two regions), the Ottomans sought before
to develop the economic potential of their patrimonial lands, in which Bursa continued to play a leading role.28 The integration of the empires dispersed economic zones
in Anatolia and Europe was realized in full only during the s.


One of the key variables affecting the growth potential of newly conquered areas, both
rural districts and cities, was the availability of sufficient workers in key industries.
Ottoman expansion into the Balkans after their victory at the first battle of Kosovo
in was assisted by the ready availability of a large cohort of Trkmen and other
Muslim settlers from Anatolia who performed basic tasks in agriculture and animal
husbandry. They contributed significantly to the enhanced productivity of the land
and the sustainability of Ottoman rule.29 Collectively, population movements organized in a planned and carefully calculated manner by the state authority formed one
of the key elements in both rural prosperity and urban growth in Ottoman Balkan
territories.30 Relocation of key workers craftsmen, administrators, municipal officials
and other civilian groups from towns and cities of the Anatolian hinterland secured a
further benefit in contributing to the process of reconciliation, normalization and the
general stabilization of Ottoman rule in the aftermath of military subjugation.
The post- development of Istanbul itself illustrates how srgn policies were
carefully planned to secure the transfer of individuals (whether urban or rural based)
who could offer the particular experience, skills and professional expertise needed to
promote reurbanization and growth after a period of demographic and economic
contraction. Mehmed II was forcefully criticized by contemporary commentators for
the unprecedented scale and vengefulness of deportations from Karaman province
to Istanbul between and .31 Yet, from the perspective of the city dwellers
of Istanbul, who gained immeasurably from the labour and skills transfer and the
creation of whole new neighbourhoods, the process resulted in durable, long-term
enhancement of the new capitals commercial potential.32 There is clear evidence that
similar policies had been introduced following the conquests of Edirne () and
other Balkan cities such as skb (Skopje, in ) on a consistent basis throughout
the century preceding the fall of Constantinople.33
Statistics for skb indicate that the city population in this remote sector of the
imperial borderlands grew very rapidly between the mid-s and the mid-s.34
This occurred despite the fact that, for a prolonged period after until around
, when Thessalonica was restored to Ottoman control, the region experienced
continuous administrative disruption, periods of disconnection with the imperial
heartlands, and the depressing effects of general economic uncertainty arising from
the disturbed political conditions. Following the stabilization of the Balkans during

Rhoads Murphey

the reign of Mehmed II, skbs urban population grew even more dramatically, due
in significant measure to population transfers achieved through the srgn mechanism
in addition to natural migration and settlement. In skb comprised households ( Muslim and non-Muslim) accommodating a population of ,
souls. In addition to its military and administrative staff, a community of this size
required a diverse and specialized civilian and artisanal workforce of sizeable proportions to provide basic market and other services.35 Barkans data extracted from the
third survey of provides important clues about the socio-economic organization
of Ottoman cities in the earlier period.36 While a noteworthy addition in is an
expanded staff overseeing the mint operation (darbhane) based in the city, balancing
that is a numerically still quite significant (perhaps even dominant) older group whose
profession is listed as aknc or cross-border raider. To the urban aknc listed
within the city confines in the mid-sixteenth century must also be added considerably
larger concentrations in the outskirts and rural hinterlands. Their continuing presence
reflects a general pattern in which Ottoman conquest was followed immediately by
large-scale settlement accomplished through strategic (re)settlement of populations.
One explicit purpose of srgn was the removal of specific groups to a distant European frontier as a means of clearing the interior regions of Anatolia of contentious,
previously disloyal and potentially anti-Ottoman tribal configurations who had strong
ties and a long-established political base in inner Anatolia. In the areas where they were
resettled (such as the regions along the Vardar valley, particularly its upper reaches
around skb), such groups then served a dual purpose, both as military specialists to
defend and expand the frontier and as herders of animals, as pastoralists and as a supplementary labour force with the potential to settle, cultivate and stabilize the regions
behind the military frontier. Ottoman chroniclers make consistent reference to the
dual purpose served by these early Anatolian immigrants, as cultivators and producers
and not just conquering heroes or gazis.37 Commenting on the resettlement in of
Saruhan-Beglu Trkmen nomads from the area around Menemen near the west coast
of Anatolia, who were first sent to the Filibe area and later accompanied their leader
Pasha Yighit, the first Ottoman governor of skb, Akpaazade states: it is the law
of sultans that they should order srgn that it might cause the region (il) to prosper
once again.38
Overwhelmingly, it was nomadic and pastoral groups who took part in srgn.
Evidence suggests that it was through a mixture of compulsion, the offer of favourable
terms for taxation and the attraction of other economic incentives such as the easy
availability of under-populated and under-cultivated land that settlers were attracted
to the frontier regions. Once arrived in the newly acquired forward zones of the former
military marches, srgn groups underwent a rapid process of diversification and transition, from an economy based primarily on animal husbandry and raiding to a mixed
sedentary-pastoral economy which eventually led to permanent village settlement and
growth of urban centres. How this process worked in detail in the earliest period
of Ottoman expansion between and cannot be understood from strictly
contemporary sources. However, some sense of the governing principles and incentive mechanisms driving early settlement can be extrapolated from sixteenth-century
provincial regulations (sancak kanunnameleri) surviving from frontier regions such
as Silistre (Silistria). Situated on the lower Danube and home to a mixed society of
pastoralists, raiders, semi-sedentarized former nomads and fully sedentary cultivators,

The Ottoman economy in the early imperial age

Silistre preserved its frontier character well into the later imperial era. The Silistre
sanca kanunnamesi, dating from the reign of Selim I (), indicates that during
the process of settlement, and for groups who still practised seasonal transhumance, a
grace period of tax exemption or leniency was recognized for tribesmen and settlers of
srgn origin. Until these populations had taken up residence (mtemekkin) for three
consecutive years in a fixed place, they were exempted from the usual obligation of
settled Muslim peasants to pay the tithe (r) of per cent and captaincy assessment
(salariye) of . per cent, which constituted the normal tax of one-eighth on grain


Another useful indicator of the level and intensity of Ottoman economic activity in
the pre-conquest era is mining and the related spheres of coin minting and general
monetary policy. By comparison with the post- empire, which needed far greater
quantities of specie to meet the salary payments of an expanding bureaucracy and permanent army corps (the kapu kullar) paid in coin, the cash needs of the fourteenthand early fifteenth-century state were modest. In this early period the states ability,
for the most part, to meet its regular obligations from existing treasury resources is
indicated by the fact that the first major devaluation of the Ottoman silver coinage
based on the ake occurred only in , at the beginning of Mehmed IIs first sultanate. So stable had the Ottoman coinage been in the first century and a half of the
states existence that even this relatively minor adjustment of per cent, reducing
the akes silver content from . to . carats or . to . grams, proved politically
explosive and contributed to the forced abdication of the young sultan in September
, a mere two years after his accession to the throne at his fathers insistence in
August . Mehmeds monetary adjustments can be considered minor given that
the scale on which they were effected remains unclear. It is uncertain whether all of
his accession year coins of AH () were low weight, or only those produced
at selected mints.40 Evidence suggests that he later achieved considerable success in
enforcing monetary controls: a document from addressed to the silver assayers of
Bursa refers to the confiscation of a horde of old (i.e., recalled) akes.41 It nonetheless
remains doubtful that any of his predecessors had enjoyed similar powers or efficacy in
the realm of monetary control.42
Perhaps the best way of judging how far early fifteenth-century sultans were able to
direct monetary policy by manipulation of the coinage is to examine the actual extent
of Ottoman control over mineral resources in the two halves of the empire. There is a
natural temptation to over-interpret scattered references in the chronicle record to the
Ottomans acquiring mines and mineral resources with their conquests in the central
Anatolian province of Rum (the Amasya-Tokat region) and in Rumeli (the Balkans)
to mean that they thereby immediately assumed direct control over the exploitation
of these mines. What seems more likely is that, initially, they either received proceeds
from these mines in the form of tribute or shared the resource with their local allies,
vassals or prospective subjects. This pattern of joint exploitation seems clearest with
respect to the copper mines at Kre, north of Kastamonu.
Retrospective accounts written by Ottoman historians long after the definitive
incorporation of the Candarid principality into the Ottoman realm in portray

Rhoads Murphey

the result of Beyazid Is preliminary incursion into the Kre region in as a conquest.43 However, it seems clear that the coastal zone remained outside the boundaries
of direct Ottoman rule throughout the first half of the fifteenth century, effectively
precluding direct Ottoman influence in these mineral-producing regions. There is no
firm evidence to suggest that the Ottomans were able to control production or distribution at the Kre mine during the lifetime of Isfendiyar Beg (d. ) or indeed at
any time before Mehmed IIs conquest of .
The financial value for the Ottomans of eventual control of the Kre copper mines
is variously reported. Data from indicate that they yielded the treasury an annual
revenue of ,, akes, equivalent (at the then current exchange rate of akes
per ducat) to , ducats.44 The figure of , ducats from the preceding decade,
c., is closely comparable.45 It therefore seems possible, if not probable, that the
Kre mines may have provided as much as , ducats annually in treasury revenues
as early as the reign of Mehmed II. The sum of , ducats indicated in Jacopo
da Promontorios account of is for revenues from all sources (customs, salt dues,
etc.), not just copper mining.46 For copper production alone a maximum figure would
have been more like , ducats.47
Before the Ottoman occupation of Serbia a long-term objective finalized by the
reconquest of Semendire in the Ottomans relied as their principal source of silver supply on the intensive exploitation of a relatively few mining sites in Thrace (Sidrekapsi and Serez) and Macedonia (Kratova and skp). Of these regions, Thrace was
by far the more important. Although the chronicles contain enthusiastic reports of the
capture of the mines at Kratova as the immediate consequence of Murad Is victory at
Kosova in ,48 their full exploitation seems to date from the period not of Bayezid
I () but of Bayezid II (), a century later. Not only was Bayezid
II responsible for a full-scale reorganization of the legal basis of mining in general, he
also opened a mint at Kratova and ordered the striking of accession year coins there to
commemorate the event.49 Bayezid IIs use of revenues from Kratova for the building
and maintenance of the erefeli mosque in Edirne is recorded in a document of
, where the figure of , akes represents the remittances of one tax farmer
(amil).50 However, it is difficult to extrapolate from this limited data any reliable idea
of total production at the mine. Some skp mintings are known from the time of
Murad II, but the active Ottoman mints before were Bursa, Amasya and Ayaslug
in Anatolia, and Edirne and Serez in Rumeli. Of these, only Serez had significant silver
deposits in its vicinity. Belons observations of , noting the operation of between
and active forges at the mine, indicate sustained Ottoman investment in the
development of the Serez site from the late fourteenth century onwards.51 In contrast,
both Ottoman and non-Ottoman sources show that direct Ottoman exploitation of
the Serbian and Bosnian silver mines dates from no earlier than the s. At Srebrenica the Bosnian Kocacevic (Kova-olu) family maintained their dominance over
local affairs, including mining, until the mid-s.52 Practically speaking, conditions
conducive to the full development of Ottoman mining in the central Balkan lands had
to await the pacification of the Bosnian frontier c..
In Serbia the Ottomans position before was roughly similar to that of ambiguous control and shared exploitation witnessed in their relations with the Isfendiyarids, who controlled the copper deposits of north-central Anatolia. According to the
historian Kemalpaazade, the tribute-paying potential of the Serbian kingdom just

The Ottoman economy in the early imperial age

before the Ottoman conquest was ,, akes, the equivalent (at the current rate
of akes per ducat) of , gold coins,53 based largely on the production of the
silver mines at Novobrdo, Trepa and Rudnik. In practice, however, the actual sum
remitted by the Serbian ruler George Brankovi (d. ) and his successors in the
years to fluctuated considerably between the base rate of , and a preconquest high of , ducats.54 In sum, Serbian compliance rates with Ottoman
requests varied dramatically from year to year. Only after could the Ottoman
treasury direct and deploy this resource to match its own economic objectives.
The preceding discussion relating both to Anatolian copper and Rumelian silver
mining raises the implicit question of whether full Ottoman sovereignty was a necessary precondition to ensure the flow of precious metals into the Ottoman lands. On
the whole the answer is probably affirmative. Only after achieving a substantial jump
in their real assets through Mehmed IIs acquisition of the Serbian silver mines and
the Candarid copper mines c. were Ottoman rulers able to exercise meaningful
control over monetary policy. While Mehmeds predecessors particularly his father
Murad II had sizeable royal treasuries, they generally placed greater reliance on external sources such as war booty (ganimet) than on fixed assets and internal sources such
as tithing and taxation. The land registration of Albania in and its centralizing
implications provoked violent reactions, illustrating the difficulties encountered by
the Ottomans in their early attempts to introduce such registration and its corollary,
direct taxation.55 In assessing the potential of the early Ottoman economy it is also
important to recall not just the narrowness of Ottoman territory in the early fifteenth
century compared to the vastly expanded empire resulting from the two transformatory events of the capture of Istanbul in and of Cairo in , but also the tenuousness of Ottoman rule at the margins, both in the northern and western Balkans
and in the central Anatolian heartlands. The drift away from the centre in the aftermath of Timurs invasion of Anatolia in and the re-establishment of a number
of the former Anatolian beyliks profoundly affected the ability of Ottoman rulers to
command and control both the manpower and the resources of Anatolia. Full Ottoman sovereignty over key Anatolian regions was reasserted only with the conquests of
Mehmed II in the s and s.


Certain broader parameters and characteristics of Ottoman attempts to establish
their hegemony over trade and trade routes require brief consideration. Ottoman
expansion westwards to the Marmara and Aegean littoral in the fourteenth century
was accomplished through co-operative arrangements. Wherever possible, the urban
infrastructure was left intact and largely undisturbed. Contemporary sources reveal the
gradualism of the Ottoman advance against the economic predominance of established families in the Meander valley and their general preference for continuity of
control.56 Ottoman trading arrangements with the Italian maritime states were patterned closely on (if not inherited intact from) those established earlier, particularly
by the Mentee and the Aydnoullar beyliks. The transformation of the Ottoman
economy and the creation of a relationship of dependency vis--vis the Italian maritime states, whereby the tables were turned in the Ottomans favour, dates only from
the half century after . Only then could the Ottomans proceed, in step-wise

Rhoads Murphey

fashion, to incorporate all outstanding Genoese colonial bases in the Aegean and
along the Black Sea coast. That this process of transformation was still only partial and gradual is indicated by numismatic evidence which suggests that Ottoman economic sovereignty was not very far advanced until well into Mehmed IIs
reign. Based on his evaluation of the gold issues of Genoese Pera, one scholar has
argued convincingly for the continuing dominance of the Genoese in Black Sea
trade throughout the s.57 Before it was never consistently possible for the
Ottomans to break the semi-monopolistic control of either the Genoese in the Black
Sea or Venice in the trade with Egypt and Syria. It is perfectly correct to attribute a
commercial motive to Bayezid I in his southward expansion against the Karamanids
and to interpret his move, in , to incorporate the hinterlands of Antalya as
prompted by the Ottomans desire for control over inland routes linking that port
with their own commercial capital Bursa in the north.58 Had the plan succeeded, the
Ottomans would clearly have gained broader access to and the potential for control
over trans-Mediterranean trading networks. However, Bayezids plans fell far short
of full realization. The real battle for dominance over eastern Mediterranean trade
was fought not in the late fourteenth century, but in the early sixteenth, when the
Ottomans gained control of Egypt.
In the early fifteenth century the Ottoman state possessed only Gallipoli to protect its strategic and commercial interests throughout the eastern Mediterranean.
The sufficiency of its Gallipoli fleet to challenge Venetian naval dominance was
tested in and found wanting.59 In the naval sphere too it was Mehmed II who
oversaw the development of an effective Ottoman naval force capable of defeating
the Venetians. Mehmed made it one of his chief priorities after to reconfigure
and expand Ottoman naval and maritime capacity through an ambitious construction programme undertaken between the years and and the relocation
of the imperial naval arsenal to Kadrga Liman in Istanbul.60 Control of Istanbul
and, ultimately, of foreign shipping and navigation into and out of the Black Sea
were necessary prerequisites for the Ottoman commercial revolution of the later
fifteenth century. It seems inescapable that the terms of Ottoman commerce before
were essentially a continuation of the pattern of dependency on western naval
support and deference to western commercial interests witnessed during the final
century of Byzantine rule.

Muntaner , Chronicle: II, .
Ibid.: .
Ibid.: .
Ibid.: .
Ibid.: .
Hodgson : I, , .
For a survey of basic features of the Ottoman economy in existence before
, focusing on
the philosophical basis and practical operation of the land regime, see Inalck : .
Ayubi ; Revel .
Cook : , n. .
Tsugitaka : .

The Ottoman economy in the early imperial age

Anhegger and Inalck (eds) : .
Ibid.: , lines .
Tarama Szl (): II, .
Beldiceanu (ed.) : document facsimile on p. , lines .
Kemalpaazade, Chronicle [Murad II fragment], MS, Bibliothque Nationale, Paris, MSTurcs,
Supp. : folio a, and a similar comment on folio b.
Anon. , Chronicle: , lines .
Promontorio .
Akpaazade : .
Zachariadou : .
z (ed.) : .
Tekin (ed.) : text facsimile, folios ab; transcription, pp. .
Babinger : .
Ibn Battuta : II, .
Cf. Anon. , Chronicle: , lines .
Pistarino : .
in Inalck : ; full transcript in Inalck (ed.) .
Inalck : ff.
On the strategic importance of srgn as an engine of growth in the central Balkan lands
from the early fifteenth century onwards, see Barkan : ff., esp. , citing a document specifying the selection criteria for Anatolian population groups considered suitable for
deployment as emigrants and settlers to under-populated districts of a newly conquered frontier province.
Barkan : ; Murphey .
See in particular Kemalpaazade , Chronicle: , and, for a Karamanid perspective,
ikari , Chronicle: .
On the deportations from Karaman and its results, see Inalck : .
For an overview focusing on the southern Balkans, see Kaleshi and Lowry , esp.
, on the development of Serres, based on the Ottoman survey of .
Inba : . See also Popovic : , and Tevfik : .
Inba : .
Barkan : , table . See also Todorov : , table .
E.g., Lutfi Paa ([] ): , re the deportation c. of Aktav Tatars from the Iskilip
region of central Anatolia to the Filibe (Plovdiv) area in southern Bulgaria.
Akpaazade : (line ) and [] : (line ).
See Barkan : , para. ;
. Akgndz : III, , para. .
nalck (ed.) : , n. .
Inalck (ed.) : , text of document no. .
On the wide circulation of foreign coins (especially gold), until the later fifteenth century, see
Pamuk : . On the diverse (and heterogeneous). mix of coins in the Black Sea economic zone, until at least the mid-sixteenth century, see Inalck (ed.), : .
Cf. Hoca Sadeddin [] , Chronicle: I, .
Barkan (ed.) : .
Murphey : .
Promontorio : . .
Cf. Babinger : ; Inalck : , table I: .
Cf. Neri , Cihannma: I, , lines .
An example is provided in Pere : .
Gkbilgin (ed.) : .
Belon : folio a.
Beldiceanu (ed.) : .
Kemalpaazade : .
Babinger : .

Rhoads Murphey

Inalck : xiii.
Zachariadou : ; Foss .
Sahilliolu : .
The argument is cogently presented in Inalck a: , esp. .
. Setton : .
Inalck : .



Colin Imber

he economy of the Ottoman empire rested on peasant labour, and, in an area

extending from the western Balkan peninsula to eastern Anatolia and into Syria
and northern Iraq, the typical peasant tenement or ift was a farm worked usually,
but not inevitably, by a single family. Most tenements belonged to a timar (fief), and it
was the fief-holders who controlled the peasants access to the land. Between the fourteenth and the seventeenth centuries, most fief-holders were sipahis (cavalrymen), who
held their fiefs in return for army service, but fiefs also supported vezirs, governors and
other servants of the sultan during their time in office. From the seventeenth century,
fiefs increasingly went to tax farmers and, in the eighteenth, following the introduction of lifetime tax farms in , came to resemble private estates. However, neither
peasant nor fief-holder owned the land: only what was above the land trees, vines
and buildings could be held as private property. The issue was not who owned the
land,1 but who owned the revenues from it. Private proprietors owned the taxes from
their estates as property which they bequeathed to their heirs; fief-holders owned the
taxes from the fief for the period of their tenure.
This system of land tenure and taxation was an inheritance from pre-Ottoman
regimes, with many local variations within the system.2 Nonetheless, from the late fifteenth century, Ottoman administrators attempted to codify and, as far as possible,
standardize the laws governing relations between peasant and fief-holder. The reign of
Bayezid II (), when there was a need to consolidate the conquests of his father
Mehmed II (), saw the first major effort in this direction: the kanunname
(law-book) for the district of Hdavendigar around Bursa became the model for later
codes. Shortly afterwards, a compiler amalgamated several rather disparate documents
from the early to mid-fifteenth century to produce a kanunname misleadingly attributed
to Mehmed II.3 More importantly, a systematic kanunname, applicable throughout
the empire, appeared in about .4 A new recension of this code appeared thirty or
more years later, following the conquests of Selim I () and Sleyman I (
),5 while the next major revision of the law came after the suppression of the celali
rebellions in , as part of an effort to restore order to the land.
The main concern of early kanunnames was taxation, the clauses on land tenure
tending to deal with exceptional and even random cases. The s, however, saw

Colin Imber

the promulgation of the New Kanunname, the most enduring achievement of the
reforms of the Kprl vezirs. This was a comprehensive land code, made up of
extracts from earlier texts compiled systematically under headings. It remained in use
until the appearance of the land code of during the tanzimat (reform) era. What
follows is an account of the law as it developed to this point.

The term ift means literally a yoke and, in principle, comprised the area of land that
a family with a yoke of oxen could cultivate within a year. In the late fifteenth century, the systematization of the laws of land and taxation created the need for a more
precise definition, and this the compilers of the kanunname for Hdavendigar
a ift is taken as consisting of or dnms of good land, dnms of land
of middling quality and to dnms of poor quality land. A dnm is an
area of normal paces lengthwise and breadthwise.6
This definition, which took into account the size of the holding and the quality of
the land, set the standard for the empire, reappearing in Bayezid IIs kanunname of
c.7 and in a number of local codes. Some later kanunnnames slightly modify this
scheme, but it was the measurements fixed in that the compilers of the New
Kanunname were to adopt in the s.8
In some regions, however, old definitions persisted. In parts of central Anatolia
in the mid-sixteenth century, the popular definition of a ift was not according to
the precise measurements of the land, but rather according to the yield of the soil,
while the system in use in Syria continued to measure land in terms of the work
that a yoke of oxen could perform.9 In these cases, the land surveyors adapted the
term ift to refer to the type of tenement found in the locality rather than attempting
to impose the standard. In south-east Anatolia and northern Iraq, some fief-holders
continued to equate a ift of land with a yoke of oxen, but did so simply in order
to increase the amount of tax that they could collect, by demanding that families
with two or more yokes pay at a higher rate than those with only one. This was a
practice which the land surveyors uncovered and prohibited in Diyarbekir in
,10 extending the same prohibition later in the century to the adjoining province
of Mosul.11
However defined, the term ift always denoted an undivided plot of land and was
the unit which surveyors used to determine a peasants tax liability, fixing the rate
according to whether he held a ift, half a ift or less. In mountainous regions where
parcels of fertile land were too small and scattered to amalgamate into ifts, instead of
calculating the tax per ift, the land surveyors calculated it per dnm on each plot, at
the standard rate of one ake per two dnms for good quality land and one ake per
three for land of poor quality.12 In districts where, by contrast, there was a surplus of
fertile land, the law permitted cultivators to farm additional plots, provided they continued to work the ift for which they were registered. When they did so, following a
principle enunciated already in the kanunname for Hdavendigar, they paid tax
per dnm on the extra land, in addition to the tax on their ift.13

The law of the land

In defining the dimensions of a ift, the compilers of the kanunname also

specified how much land the occupant could set aside as grazing for his animals and as
space for threshing and storage, copying into the kanunname the words of the decree
which they had received in answer to their query on this subject: it is not forbidden
for a peasant to leave a few dnms of land fallow as pasture for the needs of a yoke of
oxen and for a threshing floor.14 The compilers of the c. kanunname copied this
clause verbatim, and it reappears in a number of provincial codes.15


A new entrant to a ift acquired title tapu to the land by paying an entry fee to the
fief-holder, the link between title and payment becoming so strong that the term tapu
usually came to denote the fee rather than the title.
Not all property was subject to tapu. Buildings, vines and cultivated trees were private property. Fief-holders therefore awarded title to open fields, but not to buildings,
vineyards or orchards. Remove the buildings, vines or trees, however, or run a plough
through an orchard, and the soil reverted to its original status as tapu land. Nonetheless, some categories of open land were exempt from allocation by tapu. The
kanunname forbids the appropriation of land used for common grazing: since it is
harmful to the public to sow, fence in or give by tapu meadows which serve as grazing
land for the townsmen and villagers animals, these things have been forbidden.16 The
same clause appears in the general
. kanunname of c. and in later provincial codes,
the kanunname for Yeni Il applying the prohibition also to the land in front of a
persons house.17 The question of how far the front of a house could extend did not,
however, receive an answer until the following century, when the jurisconsult (mufti)
Mehmed Bahai (d. ) and, following him, the New Kanunname, ruled that it
should occupy no more than half a dnm.18
The exemption from tapu also applied to the hassa. This was a ift reserved for
the fief-holder himself, who, in principle, could not allocate it to anybody else. The
basic rule appeared in the kanunname for Bolu: ifts recorded in the register as
hassa may not be given by tapu. The rule, however, was not absolute. An additional
clause gives the fief-holder the right to award title to the ift, but for no longer than
the period of his tenure of the fief.19 A similar rule in the kanunname for Sofia
allocates natural water meadows and reed beds as hassa to the fief-holder, giving him
the right to grant them by tapu during the period of his own tenure only.20

The level of tapu tax

The rate of tapu tax was the amount that unprejudiced persons determined, this
formula applying universally until , when a decree established that the fee payable for a ift which passed to a daughter, brother or sister by inheritance was the value
of one years produce of the land.21 The compilers of the kanunname for the
Morea reduced this rate by fixing the fee at the value of the annual produce of the land
less the value of the one-seventh tithe due to the fief-holder. Anyone other than a

Colin Imber

daughter, brother or sister of the deceased continued to pay what unprejudiced persons estimated.22 On the insecure borderlands of the empire, however, it was difficult
to reach a consensus on the level of the fee, and it was this uncertainty that persuaded
the surveyors in Bosnia, on the military frontier with the Habsburgs, to fix the rate of
tapu tax in the s, not by seeking expert opinion, but instead by holding an auction
among bidders for the land.23
In contrast to what was due for agricultural land, the tapu tax payable for houses
was fixed. While houses themselves were private property, the land beneath them was
not, and anyone building a house paid tapu tax for the site at a rate already in force in
the first half of the fifteenth century and specified in the kanunname of Mehmed II:
ground tapu, for the site of a house: maximum, akes; below this, or akes;
if [the person] is poor, or akes.24 The general kanunname of repeats
the tariff, but adds that it does not apply if a person builds on land which he already
holds by tapu because it is as if he were building on private property.25 A later decree
laid out the regulations in more detail. Before building a house, a person had to seek
permission from the fief-holder. If he did not, the fief-holder could either demolish
the house or grant tapu for the site at the standard rate. However, no tapu tax was
due if the house was on the site of an old building which had fallen into disrepair.
On some new houses, however, the decree imposed a new charge. If a person built on
agricultural land, he became liable to a rent equivalent to the tax that he would have
paid on the field if it had remained under cultivation, allowing the fief-holder to take
an annual income from houses which in most cases would be on land which the
builder already held by tapu and so exempt from tapu tax for the site. It was this law
that was to appear in the New Kanunname in the s.26


The consequence of neglect
No one enjoyed absolute security of tenure on the land. A person lost his ift if he
failed to pay his taxes or if he left it uncultivated, the basic rule appearing already in
the kanunname for Hdavendigar:
if lands which are suitable for farming remain fallow when there is no hindrance
[to cultivation], and if this would be harmful to the fief, then in order to alleviate
the harm, it is permissible by custom to take them from the occupant and give
them to someone else by tapu.27
The general kanunname of c. copies this clause, but replaces the vague stipulation
that the occupant loses the land if this would be harmful to the fief with a precise
regulation. It is land that remains uncultivated for three years that is subject to confiscation. This rule remained in force until the nineteenth century, but with some
qualifications over time, the Hdavendigar kanunname already making an exception
for mountainous and uncultivated lands or land subject to flooding, which cannot
be farmed every year.28 This exemption became generally applicable, but, as the
kanunname for Ktahya stipulates, a farmer might still lose the land if he failed more
than once to sow crops in years when it was cultivable.29

The law of the land

The three-year rule also applied to wasteland which a person intended to clear. If
this lay outside the boundaries of a fief, he had an unrestricted right of occupancy
until the moment of a new tax census, when the surveyor would register the land and
allocate it to a fief. If, at the end of three years, the occupant still had not cleared the
land, the fief-holder could reallocate it by tapu. If, on the other hand, the land due
for clearance lay within the boundaries of a fief, the three-year period began from the
moment of occupancy. If the occupant had cleared it within three years, he received
title, but, if not, the fief-holder could reallocate it to someone else. If the occupant had
cleared the land without the fief-holders permission, the fief-holder could remove
him at any time, although the law enjoined but did not oblige him to grant title to
the occupant as a reward for his efforts. The rules for occupancy of wasteland became
more stringent in the late sixteenth century, with an addition to the standard threeyear rule: in mountainous lands . . . the no obstacle restriction does not apply. In
other words, the occupant of wasteland would lose title if he did not cultivate it within
three years, whether or not he had a valid excuse for not doing so.30
The peasant who faced eviction for neglecting his ift nonetheless had a residual
claim to the land. In a refinement to the three-year rule, the kanunname for Aydn
states: if the holder of the land once again pays the tapu [tax] which an outsider would
pay, the timar-holder should once again allocate [the ift] to him and not to anybody
else.31 This became a general although not a universal rule. A mid-sixteenth-century
statute from Karaman requires the fief-holder to give due warning before expelling a
person who has neglected his tenement for three years, but without, apparently, requiring him to pay tapu tax if he wishes to be reinstated. He simply has to resume cultivation. The statute further restricts the fief-holders powers by requiring any reallocation
of the land to be with the cognizance of the kadi.32 The kanunname for Silistre
similarly enjoins the fief-holder to warn the negligent farmer after three years, but does
not explicitly require him to pay tapu tax if he wishes to keep his ift.33
In the seventeenth century, the compilers of the New Kanunname adopted the rule
as it appeared in the kanunname for Aydn in , their formulation implying that
after three years of neglect a ift becomes automatically vacant, but that the previous
neglectful occupant has the right of first refusal when the fief-holder wishes to reassign the land.34 In answering the question as to whether a person loses his ift when he
leaves it in the safekeeping of another, who then fails to till the soil for three years, they
cite a law which appears in an early seventeenth-century collection of statutes attributed
to the nianc Celalzade (d. ): the fief-holder should reallocate the land, regardless
of whether or not the original occupant had left it in the safekeeping of another person.35 A related question concerned lands left uncultivated by a missing person. In
a petition to the throne elicited the decree that, if there was no news of a person after
three years, then the fief-holder should reallocate the land. This remained the law until
, when a new decree reduced the period to one year. Soon afterwards, a counterpetition complained that the one-year period was too short, and a new decree reinstated
the three-year rule, which remained in place for inclusion in the New Kanunname.36

Double taxation
ifts often lay uncultivated when their occupants left them to work the land in other
places, raising the problem of whether they should pay their taxes to the fief-holders

Colin Imber

with whom they were registered or to the fief-holders on whose land they now worked.
This issue appears already in the kanunname of Mehmed II, which rules:
if a peasant possesses land, but leaves the land . . . uncultivated and instead sows
on another fief-holders land, he should pay tithe twice: once for the land which
he sows and once for the land which he has left uncultivated. But if the fief-holder
has no land to give the peasant and he sows somewhere else, he should give the
tithe, according to custom, only to the [fief-holder on whose] land he is settled.37
The kanunname for Hdavendgar reiterated this rule,38 which then became general, and later kanunnames stipulated that the first fief-holder cannot take an extra tithe
if the farmer cultivates a new plot of land only as extra labour after he has completed
work on his original ift. This is the law that appears in the New Kanunname.39
In addition to a second tithe, anyone who abandoned his ift in order to farm in
another place had to continue paying ift tax on his original tenement. However, the
question remained of who paid the ift tax when a peasant left his village, leaving
someone else to cultivate his land. Different laws, it seems, applied in different regions.
In Diyarbekir and Mosul, it was the person who actively worked the land who paid
both the tithe and the ift tax.40 However, according to the kanunname for Silistre, the person in whose name the ift was registered remained liable for ift tax, even
when another person farmed it and paid the tithes.41

Abandoned land
The law required peasants who abandoned their land in order to take up a trade to
continue paying taxes to their fief-holders, a statute governing the payment due from
a person who takes up the carrying trade appearing already in the kanunname of
Mehmed II: if a person who hires out pack-animals, or a carter who is a carter by
trade, abandons his ift . . ., he should carry salt and other items for his fief-holder and
pay him akes per year.42 The akes is evidently compensation for loss of the
tithes and the corve a substitute for the ift tax.
The corve is an archaic feature, absent from later statutes, but the notion that
akes was the rate of compensation due for loss of tithes lingered. In parts of early
sixteenth-century Rumeli (the Ottoman Balkans), the ift-breakers tax payable for
abandoning the land was akes, the sum comprising akes compensation for the
loss of tithes plus akes for the loss of ispence (a tax of akes payable by Christian
households). The rate was not, however, uniform. The kanunname for Silistre
puts the ift-breakers tax at akes for a Muslim and akes for a Christian, fixing
the rate by compounding, for Muslims, akes in lieu of tithes and akes in lieu
of ift tax and, for Christians, akes in lieu of tithes and akes in lieu of ispence.43
In neighbouring Sofia, by contrast, the rate in was only akes.44 In Srem to
the west of Belgrade it was higher, the differential rates of and akes applying
respectively to rich and poor rather than to Muslim and Christian.45 In the midcentury the rate increased. A statute attributed to Celalzade raises the rate to akes
for a whole ift, akes for half a ift, and akes for less than half, justifying the
increase with a statement that the flight of peasants from the land to engage in trade or
commerce had harmed the income of fiefs.46

The law of the land

A decree of , compelling fugitive peasants to return to their villages following

the defeat of the celalis, confirmed the rate of the ift-breakers tax at , and
akes according to the ancient law, stipulating that, although no arrears were due, this
was to be an annual tax.47 This was the law that was to pass into the New Kanunname,
whose compilers also dispelled the misapprehension that ift-breakers tax was payable
only by Christians: concerning the law of the ift-breaker: Muslim peasants are also

Poverty, old age and infirmity

The double tithe and the ift-breakers tax applied to the able-bodied, but not when
it was poverty, old age, illness or a natural disaster that had forced a peasant from the
land. The kanunname for Hdavendigar already states that it is a blatant injustice to take ift tax from a person who has lost his land in this way. A statute in the
general kanunname of c. refined the law by ruling that, in these circumstances,
the fief-holder should confiscate the abandoned ift and re-register the incapacitated
peasant as landless, and so liable for a lower rate of tax.49 The rule was probably universally applicable. In the province of Diyarbekir, however, peasants who had abandoned their ifts through poverty or natural disaster were liable both to the tax due
from the landless and to rgadiye, originally a corve but often commuted for cash.
This rule appears in a kanunname of 50 and was copied into the later kanunname
for Mosul.51 More important, however, than the imposition of rgadiye was the condition absent from the kanunname of c. that the fief-holder could not remove
the incapacitated peasant from his ift before a period of three years had elapsed. In
other words, the normal rules for confiscating a ift after it had lain idle for three years
applied also in this case.

Forced resettlement
The aim of the ift-breakers tax was to ensure that the fief-holder did not suffer a loss
of income when a peasant abandoned his ift. A set of regulations requiring the fiefholder to bring back and resettle fugitive peasants served the same purpose. The
kanunname for Hdavendigar already states that it is ancient law to bring back and
resettle local peasants who have dispersed, adding that it was forbidden to remove
peasants who had left their land for more than fifteen years. These, it is implied, and
later kanunnames state more explicitly, should pay taxes to the holder of the fief where
they now lived. In answering the question of what should happen to peasants who had
settled in towns, the compilers of the kanunname quoted a decree issued in response
to their query, which ruled that, after they had been resident for more than fifteen
years, they were to be registered as townsmen.52 The same rule reappears in the general
kanunname of c., which reduces the residence requirement for registration as a
townsman to ten years, and adds that change of status was possible only if [the settler]
is not recorded in the register of peasants.53
These rules remained in force throughout the sixteenth century. The kanunname for Vize, for example, states that, if a peasant had been absent for more than ten
years and his ift remained uncultivated, he should pay ift-breakers tax, but that if he

Colin Imber

had been absent for less than ten years, his fief-holder should bring him back with the
cognizance of the kadi.54 However, details of the regulations vary. One recension of
the kanunname of gives a ten-year residence requirement for settlers in villages and a twenty-year period for settlers in towns before they could be registered as
townsmen. This was unless they had settled in Istanbul: the code forbids the removal
of migrants to the capital.55 Another recension of this kanunname, by contrast, requires
a delay of ten or fifteen years before a peasant is safe from removal from a village and
of fifteen years if he has settled in a town. A marginal note, however, emends this rule:
the son of a registered peasant is a peasant. He does not escape this status by being
registered in a town.56 It is nevertheless questionable whether these rules were practical. They do not appear in seventeenth-century formulations of the law, suggesting
that after they went into abeyance.


When a peasant died without heirs and there was more than one applicant for his ift,
the question could arise as to whether an applicant from within his village had priority
over an outsider. The general kanunname of gives precedence to the insider,
but this rule was unknown to the compilers of the kanunname for Yeni Il, who
clearly believed the opposite: on the question of lands which are subject to tapu, it is
forbidden to discriminate between an insider and an outsider.58 In the seventeenth
century the chief mufti Yahya (d. ), on the strength of a decree, settled the matter
by permitting villagers to take back land which the fief-holder had allocated by tapu to
an outsider,59 and this was the rule that was to appear in the New Kanunname:
if a person dies . . . leaving no heirs . . . his lands should be given to members of
the village who are in need of lands and meadows and are applying for tapu. It is
against the law to give them to an outsider.
Only if there were no applicants from the village could the land go to a stranger.

The sale of ifts raised problems. Unrestricted sales could upset the political order
by allowing land to accumulate in the wrong hands, and sale itself presented a legal
conundrum. Peasants did not own their ifts, and so, in principle, could not sell them.
In practice they did, and the law had to accommodate the fact.
The sale of land became an issue during the reign of Bayezid II. A decree of
rules that sale and purchase of land are forbidden by custom (urf ),60 and a kanunname for Aydn from the same period similarly outlaws the practice. The prohibition,
however, is only apparent, as the clause in the kanunname continues by laying down
the conditions which the parties must observe if the sale is to be valid: [The transaction] is permissible if, with the cognizance of the fief-holder, [the buyer] pays a sum for
the right of residence. The fief-holder takes one-tenth of the price which [the vendor]

The law of the land

receives.61 The same formula appears in several local codes and in some recensions of
the general kanunname.
It solved two problems. By requiring the cognizance of the fief-holder, it ensured
that transfers of land did not slip out of control, and it solved the problem of how
peasants could buy and sell land which they did not own by defining the object of the
sale as the right of residence on the land, rather than the land itself. The kanunnames
also imply that title to the land passed to the buyer on payment of the price for the
right of residence, and that the tithe which the fief-holder levied on the price was
compensation for the loss of the tapu tax which he would have received had he granted
title to the vacant ift.
Evidently, few people observed these regulations, and the issue arose again in about
, when, in reply to a questioner who stated that kadis in Rumeli were granting
certificates confirming sales of land, the chief mufti Ebussuud declared that all certificates which described these transactions as sale were void. He objected not to the
transactions themselves but to the terminology: it is absolutely contrary to the sharia
for kadis to define the giving and receiving [of land] by the peasants as sale and
to issue certificates [accordingly].62 This prohibition created the problem of how to
define these transactions if they were not sales. Ebussuuds solution was to invent a
new terminology. In place of sale, he instructed kadis to use the term tefviz, meaning
consignment. The usefulness of this word was that, while it carried the idea of transfer, it was not a technical term in the Islamic law of sale and was therefore available
to describe transfers of ifts. Its adoption allowed Ebussuud to devise a new formula
which he instructed kadis to use when registering these transactions: X, with the
permission of the fief-holder, received so-and-so many akes from Y and transferred
(tefviz) the use of his land to him. The fief-holder also gave the tapu to him.63 This
formula avoided the term sale and made it clear that what was being transferred was
the exploitation of the land and not the soil itself.
The use of the term tefviz to denote a transfer of land continued in legal vocabulary
into the seventeenth century and, through its adoption by the compilers of the New
Kanunname, into the nineteenth. It was not Ebussuuds only innovation. He also abolished the tithe on the purchase price, instead making it a condition that the fief-holder
grant the purchaser the tapu to the land:
if someone wishes to vacate the land which he occupies and, with the cognizance
of the fief-holder, gives a sum of akes for the right of residence and if, when he
vacates it, the fief-holder gives that person [the land] by tapu, this is canonical and
In making transfers invalid unless the fief-holder specifically granted title, Ebussuud
hoped to maintain control over the occupancy of land. He probably also intended the
purchaser to pay tapu tax, with this payment replacing the tithe levied on transfers.
This did not happen. Instead, it became the practice to pay the fief-holder for permission to transfer the land, raising the question of whether it was the purchaser or
the vendor who paid a problem which Yahya settled by confirming that it was the
purchaser.65 What he paid, however, was no longer understood as a tithe, the early
seventeenth-century mufti Pir Mehmed (d.) stating outright: in [the matter of
giving] permission [to transfer land] and [granting] tapu, it is unheard of to take a

Colin Imber

tithe. What an entrant to vacant land paid was tapu tax, while what a purchaser paid
was a fee for permission to transfer the occupancy to himself, at a rate that is deemed
reasonable . . . to satisfy the fief-holder.66
The compilers of the New Kanunname adopted these rules, adding insistently that
tefviz was invalid without the explicit permission of the fief-holder.

Inheritance by sons
ifts passed directly and unconditionally from father to son. The same rule applied,
even when the son was a minor. A clause in the kanunname for Hdavendgar
outlined the procedure to be followed:
tapu [tax] for an orphan is a rejected and prohibited innovation. His fathers
land is treated as if it were his inherited property. If the land descending from
his father remains uncultivated, it should be given to another person, and if, on
reaching maturity, the orphan demands it back, it is once again assigned to the
Despite the law, the practice of reassigning orphans land by tapu continued. A clause
in the Aydn kanunname repeats the prohibition, requiring tapu-holders to vacate the
land when the orphan reclaims it, and to seek reimbursement for the tapu tax from
the fief-holder.68 The same clause reappears in the general kanunname of ,
which also places a limit of ten years from puberty during which the orphan must
claim his inheritance,69 a rule that was to remain in place for inclusion in the New
If the deceased left several sons, these could inherit his ift as a jointly held tenement. The kanunname ruled that, even if one of the sons appeared in the register
as the holder of the ift and the others as landless, they in fact held the ift jointly, and
should divide the tax equitably among themselves.71 Later kanunnames repeat this ruling, but do not address the problem of whether the brothers held the title to the land
collectively, or whether each one held title to his own portion. This became an issue
when one of the brothers died. If they held the title collectively, the surviving brothers would automatically acquire the deceaseds share of the land, and before the early
sixteenth century this was the usual pattern. However, in about , the grand vezir
Hersekzade Ahmed Paa ruled that, when brothers held a ift jointly, each one should
hold title separately for his own portion. Consequently, if one of them died, the survivors could take over his portion only on payment of the tapu tax. Nonetheless, they
retained a prior claim to the land. The fief-holder could give the land to an outsider
only if the brothers could not pay or renounced it voluntarily.72
Hersekzades ruling did not gain immediate currency. A recension of the general
kanunname, copied in , repeats the law allowing brothers to hold title collectively,
but a marginal note attributed to Celalzade comments: this is wrong. The fief-holder
seeks tapu [tax] from his brothers bringing the law into line with Hersekzades
amendment.73 Nonetheless, uncertainty continued. In the first decade of the seventeenth century, Pir Mehmed ruled that, when two brothers held property jointly, a
fief-holder could not demand that the survivor pay tapu tax for his deceased brothers
share. Elsewhere, on the strength of a decree of , he states the opposite, suggesting

The law of the land

that the issue had again become contentious, until finally a decree settled the issue in
favour of Hersekzades original law.74
If one of the brothers occupying a ift died leaving a son, the son inherited his share
of the land without paying tapu tax, a rule which appears already in the kanunname of
Aydn from the reign of Bayezid II.75

Inheritance by brothers and grandsons

If the holder of a ift died without leaving a son, certain of the deceaseds relatives, provided they paid the tapu tax, could inherit the land. Before , the fief-holder should
first of all offer the land to his brother the term brother, Hersekzade explained,
applying only to full brothers or to half-brothers sharing a father.76 This remained the
rule into the seventeenth century, the compilers of the New Kanunname stating that,
in matters of tapu, the term brother applied without distinction to full brothers or
to brothers sharing the same father. Half-brothers sharing a mother were legally not
The rules governing inheritance by grandsons changed during the sixteenth century. Hersekzade stated that, provided his father was no longer alive, a grandson could
inherit a ift from his grandfather without payment, a rule that appeared in the general
kanunname of . However, Celalzades marginal note in the kanunname states
bluntly: land does not descend from grandfather to grandson, adding that the rule
applying to brothers applies also to grandsons: if the grandson pays tapu tax, the fiefholder should not give the land to anybody else.78 It was Celalzades ruling that was to
survive into the seventeenth century, when it was incorporated into the New Kanunname with the proviso that a grandson must claim the land within a time limit of ten
years from his grandfathers death.79

Inheritance by daughters
Before , women could not legally occupy land in their own right. Nonetheless, the
kanunname for Hdavendigar and later codes stated that a fief-holder could not
expel a woman provided she kept her land under cultivation and paid her taxes.80 The
assumption was that any woman in this position would be a widow, but in emending
the relevant clause, which, in one recension of the kanunname of c., opens with
the phrase if a woman on her husbands land does not leave [the fields] uncultivated
. . .,81 to read if a woman, by some means or other, occupies land . . ., the redactors
of the kanunname of accepted that widowhood was not the only way for
women to acquire land.82
Nonetheless, the administration tried to curtail womens access to the land. The
kanunname of copies a clause from a group of western Anatolian kanunnames
of forbidding women from acquiring ifts by paying tapu tax. This prohibition,
as the kanunnname for emigezek makes clear,83 applied to both widows and
daughters. A further restriction followed in , when a decree forbade daughters
from acquiring shares in ifts by becoming partners with their brothers.84 The law
allowed one exception to this rule. To the clause in the kanunname of barring
daughters from inheriting land, Celalzade added this comment:

Colin Imber

if the deceased person cleared the land with his own axe and, in short, expended
cash and effort, and if he died leaving no son or brother, but leaving a daughter, it
is now commanded that it should be given to his daughter.85
The new law probably dating from thes stipulated nonetheless that, in order
to acquire the land, the daughter should pay tapu tax.
A major change followed in , when, for the first time, a decree allowed
the lands of a deceased person who leaves no son to be given to his daughters
who are requesting them, for the tapu [tax] which disinterested Muslims determine.86
This was to raise two problems. First, the decree placed no time limit on daughters claims, with the result that some demanded their inheritance many years after
their fathers death, when the land had already been reallocated several times. In
, a decree following a petition put a limit of ten years from the death of the
father.87 The second question was whether, and on what terms, the daughters
own children could inherit from their mother, the assumption from being
that a son could inherit without paying tapu tax and that a daughter could inherit
on payment of the fee. In , however, a decree stipulated that land could
descend to sons only on payment of tapu tax, raising the question of whether
this applied also to daughters. The answer came in an order issued at some time
before : the deceased [mother]s vacant land descends by tapu, [and] only to her

Inheritance by sisters, fathers and mothers

In , a decree extended the right of inheritance to a deceased mans sister, provided
that she paid tapu tax and was living in his place of residence. In the following year
a petitioner asked for the residence requirement to be lifted, eliciting a decree which
modified the original law by restricting inheritance to sisters living in the region.89 In
another decree, again in answer to a petition, extended the right of inheritance
to the deceaseds parents:
it has been decreed that if the deceased leaves no son, daughter, or brother
with whom he shares a father, but leaves a father and a mother, all his lands
and meadows should be given to his father for the tapu tax which disinterested
Muslims determine. If he leaves no father, they should all be given to his
Between and , therefore, a series of decrees had extended the right
to inherit land to family members other than sons and brothers. The changes had
occurred piecemeal, but nonetheless followed a series of consistent principles. Only
sons could inherit without paying tapu tax. Males took precedence over females in the
same degree of relationship, and descent was patrilineal. The changes had created a
hierarchy among the heirs and, with it, the need for a concise statement of the rules.
The task of providing one fell to the nianc Okuzade, who, in , at the request
of the mufti Yahya, summarized the laws of inheritance. It was Okuzades summary
which appeared in the New Kanunname:

The law of the land

the deceaseds vacant lands are given by tapu to his daughter; if he has no [daughter], to his brother with whom he shares a father; if he has no [brother with whom
he shares a father], to his sister resident in the [same] place; if he has no [sister
resident in the same place], to his father; if he has no [father], to his mother . . . A
deceased mans land descends without tapu only to his son.91

The laws of land tenure served two purposes. The first was to keep revenues
flowing by ensuring that the land remained under continuous cultivation; the second, by controlling access to the land, was to prevent the growth of large private
estates which would have diverted the revenues to private owners. The fundamental
principles of the law remained intact between the fifteenth and nineteenth centuries,
the major change being the extension, after , of the right to inherit land to family
members other than sons and brothers, and to women in particular. It was evidently
demography that forced the change. Before , the population had been growing and there was no shortage of men to till the soil. However, during the wars and
rebellions of the late sixteenth and early seventeenth centuries, there was a flight of
men from the land, and it was to ensure that the deserted ifts would not lie idle that
the administration legalized their descent to the daughters, sisters and parents of the

Who owned the land?
The question of who owned the land itself did not arise until the s. Following
the annexation of central Hungary as an Ottoman province in , the kazasker of
Rumeli, Ebussuud, provided a statement on the laws of land tenure that were to apply
in the newly acquired territory. His statement not only gives a summary of the rules
of land tenure, but attempts to do so within a framework of Islamic law which treats
land as a commodity that can be held as private property. Ebussuud therefore faced the
problem of deciding who actually owned the land. His solution was to fix ownership
in the treasury, using the formula the real substance of the land is the Treasury of the
Muslims.92 In Islamic legal theory, the Treasury of the Muslims is the joint property of the Muslim community. The compilers of the New Kanunname in the s
adopted Ebussuuds statement as the basic document on land law. Elsewhere, however, they make the contradictory statement: miri land that is, land which was not
controlled by a trust or by a private owner is the sultans.93 However, both they and
Ebussuud are careful to avoid using any technical term which would imply that either
the treasury or the sultan held the land as private property. They indicate ownership
by the simple use of the Turkish possessive genitive, which is legally non-specific. In
reality, the question of ownership of the land was irrelevant. The important issue was
who controlled the land or, more specifically, its revenues.

Colin Imber

From the Corpus Juris Civilis to the New Kanunname
Although the Ottoman land regime developed over the centuries to meet changing
needs, its basic principles remained intact, and these, it is clear, were an inheritance
from pre-Ottoman times whose distant origins are plain.
In the fifth century ce, Roman jurists faced the problem of describing a type of
contract which transferred land to be enjoyed in perpetuity by a tenant and his heir
or assignee, provided only that he continue to pay rent to the owner. When the jurists
were unable to decide whether this was a contract of hire or of sale, the Emperor Zeno
established it as an independent contract under the name emphyteusis.94 It was a contract whose terms conform exactly to the conditions of peasant tenure in the Ottoman
In the following century, the Emperor Justinian refined the law by decreeing that, in
the absence of any agreement to the contrary, the owner could eject the emphyteutical
tenant if he failed to pay the rent for three consecutive years.95 The same rule was to
apply to Ottoman peasants. Similarly, the distinction in Ottoman law between fruittrees, vines and buildings, which could be held as private property, and the soil itself,
which could not, seems also to date from the time of Justinian. It was this emperor who
decreed that an emphyteutical tenant could, provided he had received permission from
the owner, sell any improvements which he had made to the land.96 Since it is impractical to sell improvements to the soil itself, this rule could apply only to things above the
soil, most obviously to fruit-trees, vines and buildings. Since Justinian still requires the
tenant to obtain the permission of the owner of the land before selling the improvements, these are not yet in the fullest sense the tenants private property as they were
later to become in Ottoman law. Nonetheless, the separation in Ottoman law between
the soil itself and things above the soil clearly has its origin in Justinians ruling.
It is possible also that the Ottomans inherited two further rules of late Roman
law. First, when an Ottoman peasant sold his right to occupy a ift, his fief-holder
levied a charge, fixed officially at one-tenth of the price realized. This seems to reflect
a late Roman practice: Justinian attempted to curb abuses by landowners by limiting the sum which they could collect when an emphyteutical tenant transferred his
land to another party, to one-fiftieth of its price or valuation.97 Second, Ottoman law
required a new entrant to a ift to pay an entry fee to the fief-holder. From the late
fourth century ce, Roman law required a tenant who obtained land on the imperial
domain by emphyteutical right to furnish sureties against the land being abandoned.98
By Ottoman times, this payment had lost its character as a security, but the principle
that occupancy of the land was conditional on the payment of an entry fee remained
in place.
The Roman contract of emphyteusis, with some variation in its precise form, survived in the Byzantine empire and its successor states down to the Palaiologan era.99 It
is this contract that the Ottomans inherited and which was to become the basic form
of peasant tenure in the Ottoman empire. It was the contract, in its Ottoman guise,
that Ebussuud was to describe in summary in the s and which the New Kanunname of the s was to regulate in detail, more than a millennium after Justinian
had laid out the basic rules in the Corpus Juris Civilis.

The law of the land


See Postscript , p. .
See Postscript , p. .
Barkan : no. ; Kraelitz-Greifenhorst [] .
Facsimile of Koyunoglu MS, in Akgndz : II, .
Heyd : .
Barkan : .
Akgndz : II, .
MTM : .
Barkan :
Ibid.: .
Ibid.: .
Ibid.: .
Ibid.: .
Ibid.: .
Akgndz : II, .
Barkan : .
Ibid.: .
MTM : .
Barkan : .
Ibid.: .
MTM : .
Barkan : .
Ibid.: , .
Ibid.: ; Kraelitz-Greifenhorst [] : , .
Tuncer : .
MTM : .
Barkan : .
Ibid.: .
Imber .
Barkan : .
Ibid.: .
Ibid.: .
MTM : .
Ibid.: .
Barkan : ; Kraelitz-Greifenhorst [] : .
Barkan : .
MTM : .
Barkan : , .
Ibid.: .
Ibid.: ; Kraelitz-Greifenhorst [] : , .
Barkan : .
Akgndz : VI, .
Barkan , no./: .
Akgndz : VII, .
MTM : .
Ibid.: .
Akgndz : II, .
Barkan : .
Ibid.: .
Ibid.: .

Colin Imber

Akgndz : II, .
Barkan : .
Akgndz : IV, .
Ibid.: IV, .
Ibid.: III, .
Barkan : .
MTM : .
Sahin and Emecen : .
Akgndz : II, .
MTM : .
Barkan : .
MTM : .
Ibid.: .
Barkan : .
Ibid.: .
Akgndz : III, .
MTM : .
Barkan : .
Akgndz : II, .
Ibid.: III, .
MTM : .
Akgndz : II, .
Ibid.: II, .
MTM : .
Akgndz : III, .
MTM : .
Barkan : .
Arif : .
Akgndz : III, .
Barkan : .
Ibid.: ; Beldiceanu and Beldiceanu-Steinherr : ; MTM : .
Akgndz : III, .
MTM : .
Ibid.: .
Ibid.: .
Ibid.: .
Ibid.: .
Barkan : .
MTM : .
Justinian : .
Corpus [] : XIII, .
Ibid.: XIII, .
Ibid.: XIII, .
Ibid.: III, .
Weiss ; Zakythinos : .



Sofia in the seventeenth and early eighteenth centuries

Rossitsa Gradeva

t is virtually impossible to overestimate the importance of kad courts in provincial

life in the pre-tanzimat period. This was an institution with which for a variety of
reasons nearly everyone in the Ottoman empire came into contact Muslims, Jews
or Christians, reaya or askeri, villagers or urbanites, tribes, visitors from abroad and
individuals, as well as groups with different professional, religious or social profiles.
Kads were approached either as judges or as the local administrative representative of
Ottoman authority in the respective district, or both, as these functions were closely
intertwined. In the past forty years, much has been written on the institution and its
(mainly judicial) activities, for various parts of the empire and vis--vis various groups
of Ottoman society.1 However, many of its local specifics, its evolution over time,
and its relation with other Ottoman institutions of authority and with local populations have yet to be studied and interpreted, and these continue to attract scholarly
attention. While the number of publications on the functioning of the court in the
seventeenth century increases, with regard to the Balkans, publications dedicated to
the earlier centuries are practically non-existent, and those for the eighteenth century
are still relatively few.
This essay provides an overview of the role, functions and structure of the kad
court, taking the example of Sofia during the seventeenth and early eighteenth centuries. It shows the kad court as a judicial institution but also one at the heart of the
administration of Rumeli, a centre of social life and public space where all types of
Ottoman society met. In many respects, therefore, this study presents the principal
features of kad courts throughout the empire. However, it is difficult to describe the
kad court in Sofia during this period as entirely typical. This is due partly to the process of transformation taking place during the seventeenth century within most of the
so-called classical Ottoman institutions, and for the Ottoman polity in general, and
partly to the fact that Sofia was the seat of the governor of Rumeli. The latter obviously added to the usual obligations of a kad, putting him in the precarious position
of ranking highly in the judicial hierarchy, but at the same time being exposed to the
direct effect of the valis presence. This largely explains some of the discernible specifics of the kad court in Sofia.2

Rossitsa Gradeva


By the first decade of the seventeenth century the kadlik of Sofia had been permanently integrated into the mevleviyet (higher) grade of the judicial hierarchy indicative of its growing importance. Sofian kads originated from all parts of the empire,
including Baghdad, Jerusalem, Ayntab, Bursa, Crimea, Georgia, Istanbul, Edirne and
Sofia itself. Some belonged to eminent ulema families, others had strong ties with
dervish brotherhoods sometimes both but there are also striking exceptions to this
model men of very modest background. Most followed the established pattern of
career-building. Before becoming a judge they would have normally taught at highranked medreses (colleges of Islamic law), though rarely at the Sleymaniye in Istanbul
or in other major colleges of comparable importance. Others, however, had only a
very brief teaching career, often in lower level medreses. The Sofia mevleviyet was somewhere in the middle to lower ranks of this judicial grade, but apparently was a lucrative
position. Not a single seventeenth-century Sofia kad became kadasker, although one
fifteenth-century kad became grand vezir (Piri Mehmed Paa, d. , who was kad
in Sofia before ). The normal positions a kad would attain before and after Sofia
were those in Diyarbakr, Kayseri, Manisa, Belgrade, Bosna Saray (Sarajevo) and, in
exceptional cases, Baghdad, Filibe (Plovdiv) and Izmir.3
The majority of seventeenth-century sharia judges in Sofia had very short terms in
office, from just over two years down to two to three months, with the average about
a year. Several spent more than one incumbency and hence more years in Sofia; some
chose to retire there, others died while in post.4 In contrast to sixteenth-century kads,
who left a lasting trace both in the topography of the city and in the memory of their
fellow citizens, few seventeenth-century judges developed any personal relationship
with the city. They not only donated less to local welfare, they are also not mentioned
in Christian texts surviving from the period.5 The latter might be due to the lack of
major clashes between Muslims and non-Muslims comparable to those of the neomartyrs in the sixteenth century in which the kads played a major role.6 Although the
seventeenth century was far from peaceful, the spectacular events in the life of Sofians
destruction of churches in the s and s,7 the banishment of local prostitutes
and even, by leave of the sharia and for the reform of the world, the hanging of some
of them (between December and July )8 were initiated not by the kad but
by the current governor. One kad challenged the governor who in executed several people in Sofia without the legal sanction of the judge and then tried to secure the
latters ex-post-facto approval.9 Despite the theoretically important role of the sharia
judge in such cases, nowhere in these events, except for that of , does the city kad
appear as a major actor. This can probably be attributed to the growing role of the
governor in provincial administration and adjudication at the expense of the judge.


Despite the prominence of the governor, in the seventeenth and early eighteenth century the kad court in Sofia remained a focal public space for its citizens and for the
inhabitants of the kaza (judicial district). The kad constituted one of the pillars of
Ottoman authority in the province. The scope of the work of the kad court can be
assessed using the surviving sicills (registers, record books), despite concerns about

A kadi court in the Balkans

their reliability.10 Principally, registration of documents issued by the kad or reaching him was not mandatory but depended on a variety of circumstances, including
the readiness of the interested party to pay the necessary fee.11 Although registration
in the sicill provided security in the case of loss of the original document or the reopening of the case, it was not always resorted to. The same applies also to documents
issued by the central authorities, especially those in response to petitions of individuals or groups. One may suspect also specific areas of under-representation such as
non-Muslims and villagers, who probably approached the court less frequently than
did Muslim inhabitants of the city.12 Nor do the registers include the outgoing correspondence of the kaza. We only learn about (some of) this from the incoming orders
registered in the sicills, from the records of the divan the mhimme and the ikyet
defters (registers of important affairs, registers of complaints) kept in the imperial
council or from single documents. Finally, it seems that many courts kept parallel
registers one for general issues, the main judicial and administrative activities of the
court, and others more specialized for example, for vakf documents,13 for inheritances and related judicial acts (tereke, muhallefat),14 for incoming orders,15 and sometimes for marriage contracts,16 which might cover several judgeships. What survives
is often just one of these parallel sicills. Yet, despite these recognized drawbacks, sicills
offer a unique insight into the wide scope of kad court activities and the complex
negotiations and relations this implied. They continue to be the most comprehensive
source available for study of these activities.

A judicial institution
The kad court was first of all a judicial institution. All its other functions emanated
from its primary and traditional association with adjudication on the basis of the sharia.
Its early integration within the Ottoman bureaucracy, and its authorization to apply
the kanun and other legal edicts issued by the sultans, further enhanced the courts
role. The Sofia sharia court reached the peak of its importance in the sixteenth and
seventeenth centuries, but towards the end of the seventeenth century possibly began
to become less significant on account of the increasing authority of the provincial governor. This trend continued until the second half of the nineteenth century, when the
kad court returned to being a judicial institution treating mainly the (broadly) religious problems of Muslims. Late seventeenth- and eighteenth-century records attest
to the frequent appearance of the provincial governor and his staff at hearings in the
Sofia court; the fact of the joint sessions of the kad with the provincial divan; and
the overwhelming presence of the ayan-i vilayet (provincial notables) in the work of
the court as parties in a significant number of cases, as witnesses to the event in even
more, and as a forum which decided on many issues of all-kaza importance, which
were then recorded in the sicill.17
In its judicial role, the kad court functioned in complex interaction with several
other institutions which were similarly not exclusively judicial.18 At the highest level,
in the heart of the empire, was the imperial divan acting in the name of the sultan. Primarily a political body, it also received applications and grievance petitions
from all parts of the empire from individuals and various communities, Muslims,
Christians and Jews, men and women, citizens, villagers and tribesmen and responded
to them, if only to instruct the officials concerned to review the case. Applicants

Rossitsa Gradeva

complained about the malfunctioning of local administration and oppression and

taxation issues, but also about private legal cases related to inheritance, murder, debts,
family conflicts, problems among trade partners, or others. Sometimes they approached
the divan following an unsatisfactory (from their point of view) decision of the local
kad; on other occasions they simply petitioned the divan. The orders in response,
issued in the sultans name, are reflected in the mhimme 19 and, after , also in the
ikayet20 series of the divan or, from , in the regional ahkam defters.21 Seventeenthcentury mhimme and ikayet registers rarely include more than ten judicial cases per
year from Sofia, which suggests that the kad court remained the focal judicial institution for Sofians, and only a few approached the divan for private reasons.
Locally, the kad court interacted in a variety of legal and administrative cases with
the governors divan, another institution of mazalim jurisdiction in the Ottoman
state.22 Cases were sometimes brought first to the vali, where the kad was also invited,
and then to the kad court, where members of the governors retinue participated as
witnesses to the event; sometimes cases went in the opposite direction, first to the
kad and then to the vali; in yet other cases they also reached the central divan, from
where they were transferred to the vali. It is not clear whether these were all the cases
taken to the governors court, and, if not, why the kad was involved in these in particular or why he would record them. There are indeed many questions around the
functioning of the provincial divan and the division of prerogatives and functions
with the respective kad. Surviving registers do not support the obvious hypothesis
of a division between sharia and kanun (administrative law) cases: the governor was
approached on inheritance disputes and the kad with grievances about irregularities
in the collection of taxes.
Much has been written on the attitudes of non-Muslims, mainly Orthodox Christians and Jews, to the kad court. Scholars who have analysed these on the basis of the
kad sicills are unanimous that members of all non-Muslim communities turned to
the kad court not only for cases that were mandatory (mixed that is, with Muslims
or criminal), but also in cases regarded as religious (such as marriage and divorce).
Some see the explanation for this phenomenon in the actual absence of alternative
non-Muslim courts; others indicate the awareness of non-Muslims of the possible
advantages that turning to the Muslim court would bring them more favourable
stipulations of the law, a higher chance to win their case, a better chance to have the
verdict/decision procured, lower fees which led them to overlook the prohibitions
issued by their religious and communal authorities. Although their status was not
very clearly defined, non-Muslim religious and communal courts (not necessarily
overlapping) did exist and interacted with local kad courts. Their importance seems
to have grown from the seventeenth century onwards as part of a complex evolution
of the Ottoman polity and the developing role within this of communal structures,
especially concerning taxation.23 However, only in the late eighteenth/early nineteenth century did the Ottoman authority officially and explicitly grant the Orthodox religious authorities/communities a sort of monopoly on religious cases. But
this was never fully enforced, perhaps partly because of the preferences of non-Muslims themselves, who continued to take advantage of more favourable prescriptions
of the sharia.24
Even less is known about the communal courts based on customary law, which
functioned among Orthodox Christians and some groups of Muslims. This is a highly

A kadi court in the Balkans

controversial issue, in the first place because, with the exception of those in the Aegean
islands, they have left documentation only rarely.25 Nationalistic Balkan historiographies have tried to present communal structures as a major alternative to the alien
courts, which thus contributed to the preservation of the national spirit. While this is
probably an exaggeration, the role of these institutions may not be simply discarded.
Their existence is apparent, even where no documents have been preserved, in explaining the relatively low percentage of Christian villagers attending the sharia courts, and
probably also the ecclesiastical courts. It is yet another question what law these institutions applied. Very often one may suspect a complex interaction of the dominant
sharia, tribal and pagan traditions, canon law and pre-Ottoman legal practices.26
All these institutions, and several others of lesser importance not considered here,
which functioned in a complex interaction and hierarchy, offered many opportunities
for forum shopping for those concerned, as long as they knew of their specifics and
advantages and were prepared to go against prohibitions.27 For different groups of
Ottoman subjects in the kaza of Sofia, it is difficult at this stage to evaluate the respective weight of these judicial institutions. Some kad sicills are available from the midsixteenth century, but mainly date from the beginning of the seventeenth century and
even then with significant gaps. The contemporaneous series of mhimme and ikayet
defters, which reflect activities at the imperial divan, also contain significant gaps. The
other components of the judicial system the governors divan, the religious courts
and the customary law courts remain an unknown variable for Sofia district in the
period under consideration. At this stage we can claim only that they existed. The very
low number of court records which involve Jews, Armenians and Latins (Ragusans
and Catholics) in the kad sicills of Sofia, in comparison with their relative weight
in the citys population, suggests that the respective communal institutions received
some of the obligatory kad court cases.28 However, the abiding impression is that the
kad court continued to function, as both a legal and an administrative body, as a focus
of life for all the groups indicated.

The notarial function

What made the kad court unique, as compared to other official Ottoman judicial
institutions, was its notary function. No divan recorded documents of this type. On
the other hand, there is sufficient information to show that religious and communal
courts did provide these services to their co-religionists, although they were not officially assigned to do this. Notarial documents are recorded in the extant codices of
metropolitanates, preserved from the seventeenth century onwards; Jewish responsa
also refer to the recording of various agreements and transactions. Bearing in mind
the limitations of these courts in serving only the respective religious group, and the
undisputed higher security provided by the kad court in the case of subsequent contestation and as the court for mixed cases, it is not surprising that the latter contains
the bulk of the notary records. In the extant Sofia sicills for the period discussed here,
notarial documents make up between and per cent of the legal entries. They
include mainly records of debts, contracts for transactions of houses, (work)shops,
mills, vineyards or other roofed or landed property, sometimes in places in Anatolia or
other distant settlements in Rumeli; transfer of rights of usage of landed plots; loans;
suretyships (kefalet); and proxies (vekalet). Only very rarely do we come across records

Rossitsa Gradeva

of transactions of moveable property, even of relatively expensive items such as slaves

or cattle. However, the kad did issue documents related to the manumission of slaves,
conditional or unconditional.
Some transactions and agreements were contracted in loco, in the court, while others, probably the majority, seem to have been negotiated outside it and recorded sometimes years later, when one of the parties needed a legal document or when the due
sums were fully paid. Very rarely, contracts were registered between a community
or mahalle (city quarter) and a person hired to provide certain services. Suretyships
were recorded mainly in relation to debts and collective liability, but also when an
offender was being released from prison and a guarantee was needed to ensure his or
her appearance in court when required. As elsewhere in the Ottoman empire, vekalets
were usually registered when a party to a judicial case needed a representative in court
or an official, as a deputy to carry out (part of) his or her obligations. The normal case
was for a specific vekalet to serve in a strictly determined case, for a clearly specified job,
such as a sale or other transaction, or in arranging divorce and post-divorce relations.
Muslims and Orthodox Christians (but rarely Jews or Armenians), men and women,
appear in court as both the authorizing and the authorized party.

Contested cases comprise a significant component of the Sofian kad courts judicial
activities, between and per cent of the legal documents in extant records in this
period. These include disputes over landed and other property, or over the borders
of a common, between villagers and townsmen. They also include cases concerning
the payment of taxes and dues, about their amount and the method of collecting,
between tax collectors and taxpayers, or between taxpayers (groups and individuals)
only. Closely related to the kads broad prerogatives in the control and regulation
of urban economic life is the judges role in cases going beyond the competences of
the respective guild bodies: disputes between esnafs (artisan guilds); disputes between
an esnaf and individual craftsmen who were not its members or who were, but the
offence was beyond their jurisdiction or considered of high public importance; on
breaches of production regulations; on provisioning with raw materials; on prices, and
conflicts with market officials; and on the way esnafs performed their obligations to
the state in peace and war.

Marriage and divorce

A significant number of documents recorded in the Sofian sicills concern family relations.29 These range from registration of an agreement between (former) spouses
and the allocation of alimony or the appointment of guardians to disputes related to
inheritance or difficult divorces. The courts issued special permission (nikah tezkere,
izinname), bearing the judges seal, to couples intending to marry. Such a document
informed neighbourhood imams that there were no sharia obstacles to the conclusion
of the marriage.30 Marriage contracts (nikah akdi) which fixed the size of the dower
due to the wife in the case of the husbands death or divorce (mehr-i meccel), and
occasionally provide details on specific issues and conditions of cohabitation, are quite

A kadi court in the Balkans

rare. Very occasionally, kads also registered Christian marriages, perhaps reflecting
problems regarding the Christian legalization of the marriage. Documents related to
divorce crop up more often. Because of the specifics of Muslim divorce, most divorces
registered in court were initiated by the wife (hul divorces), were conditional, or followed some agreement between the spouses. Information about divorce initiated by
the husband comes mainly from documents concerning the settlement of disputes
between the former spouses. Christians, particularly women, also registered divorce,
either directly in the sharia court or after the dissolution of the marriage at the ecclesiastical judicial body. Many questions arise in such cases which for the time being
remain speculative: how knowledgeable Christians were about the stipulations and
intricacies of Islamic law; where they learned about it; whether the mehr-i meccel
which women demanded in court or which occasionally crop up in terekes were actually pre-arranged and in what format, or were just a strategy after divorce or the death
of the husband; what it cost Christian women to raise property claims that fit sharia
law but were in contravention of canon law; and so forth.31 Issues of public morality
of both Muslims and Christians and the marital status of women in particular were
under scrutiny, and the court was involved to secure this.

According to the Ottoman kanun, in cases when there were no legally recognized heirs,
or when their due shares according to the sharia were insufficient to exhaust it, agents of
the state treasury (beytlmal) were to visit the house of the deceased and decide whether
the state had any interest in the estate. Many such cases would eventually reach the sharia
court, at the demand of the agent or if contested by the heirs. The kad court intervened
when legally recognized heirs were absent, when a wife was pregnant, when there were
under-age heirs in short, when there were weak legatees who needed protection. It
could also be approached when the heirs could not agree about the division of the estate
among themselves or for any other legal problem that arose in relation to it. The court
also decided disputes between heirs and agents of the treasury, between heirs and other
interested parties, and among heirs only, often caused by transactions effected before
the death of the legator. Indeed, eventual problems with the beytlmal or among heirs
may have been the reason for the rather frequent transactions with real property between
spouses recorded in the kad sicills. Disputes between manumitted slaves and the heirs
of their former masters or the agent of the state treasury were not rare, and this explains
the frequency of the requested issuance of relevant documents by manumitted slaves.
Anticipated problems with the beytlmal were probably the reason for an otherwise rare
practice to which Ottoman subjects resorted only in exceptional circumstances to
declare and register gifts to their potential heirs or to indicate the names of the heirs in
front of a member of the court. In Sofia, this was done mostly by temporary residents in
the town, mainly Armenian merchants. Except for cases of donation for pious purposes,
wills were declared also within the family, in the presence of friends and neighbours
and probably also religious functionaries of the respective faith, and only disagreements
among the heirs or some transaction would reveal them in the sharia court.32 Although
these problems were regarded as a prerogative of the religious courts of each community,
we often see members of the non-Muslim communities, mostly Orthodox Christians,
settling family problems according to sharia law.

Rossitsa Gradeva

Criminal cases
Only around per cent of documents in the Sofia kad sicills relate to criminal cases
principally manslaughter, battery, highway robbery, theft, adultery, burglary, counterfeiting and forgery. Other documents, such as suretyships and divisions of estates,
reveal that many other crimes occurred in the district of the Sofia kad but were not
the subject of court intervention. The main reason for this is probably the fact that,
in offences against man, Islamic law leaves much space for negotiation between the
parties involved, including reconciliation outside the court. The court was approached
usually with regard to potential problems for the territorial community where the crime
had taken place, a village or urban neighbourhood whose inhabitants were bound by
collective liability and who were held responsible if the culprit remained unknown.33
In order to avoid protracted litigation and oppression from police officials and the relatives of the victim, even when the culprit was identified, the community still tried to
secure a document from the kad court absolving them from responsibility. This was
why the inhabitants of a mahalle or village were often the first to inform the sharia judge
about a death that was violent or had occurred in unusual circumstances, a suicide or a
wounding and to demand an inquiry. Sometimes the victim or his or her relatives stated
the fact of the criminal act, or simply of death, thus absolving the community from
payment of compensation and occasionally also identifying the culprit. A specific but
frequent case is when the inhabitants of a village or neighbourhood sued members of
the community, women of easy virtue or bandits, usually demanding their eviction. In
such cases the testimony of the community against them was deemed a sufficient proof
even when the defendant denied the accusations. The kad court tried people accused
of highway robbery when (as in the cases of immoral behaviour) charges were raised by
large groups of local people, whose claim usually decided the case.
A very brief overview and comparison between the early and late sicills does not show
any significant shift in the nature of legal cases reaching the kad court. What does
change, however, are the court sessions (meclis-i er) held in the presence of the vali,
in the divan-i Rumeli (council of Rumeli) attended by the kad of Sofia, but also by a
divan efendisi.34 At present it is only possible to say that these elevated sessions concerned a wide range of issues and were often either preceded or followed by sessions
of the kad court per se, sometimes supported by emr-i erif (imperial orders). They
concerned problems with the collection of taxes, sometimes for distant villages in the
province or following clashes between former vali and the non-Muslims of Sofia.35
Other topics include the contested sale at auction of inherited property between highranking askeri in Sofia, a contested timar in the sancak of Vidin, a trial of slaves who
murdered their aa, and other matters which were probably taken to the vali because
of their high public importance.36 The interaction between the kad court and the
governors divan also awaits further investigation.


As a judicial and religious authority the sharia court controlled the functioning of
the pious foundations (vakfs), appointments or temporary replacements of Muslim

A kadi court in the Balkans

religious functionaries (imam, duaguy, hatib and others), and the various levels of statedefined aspects of the structures of non-Muslim communities. Thus kads oversaw the
vakfs their establishment, financial situation and transactions the appointment of
trustees (mtevelli), and repairs to their material components.37 Mtevellis were also
replaced in or through the court, a procedure usually started by the local Muslim community, the immediate recipients of their services.38 Occasionally problems related to
non-Muslim vakfs were also brought up in the sharia court. The usual documentation in Sofia sicills as far as non-Muslim religions are concerned is registration of the
berats of metropolitans, which defined the limits and scope of their functions allowed
by Ottoman authority,39 and documents relating to permission for the restoration of
Christian and Jewish places of worship. Kads were involved at various stages in the latter procedure. Following an application from the respective group and an order from
the sultan, the court organized an inspection of the building. The exact dimensions
and a description of the old structure were recorded to serve as the basis for monitoring
the actual work done, also the prerogative of the kad. The Sofia sicills contain documents reflecting both stages of the procedure,40 as well as the increasing intervention of
the governor and other high-ranking officials in control over non-Muslim structures
and taxation and in the preservation of the dividing lines between Muslims and the
non-Muslim communities.41


The kad court was in theory the eyes and ears of the sultan in each provincial district,
in terms of control over and feedback about the functioning of the various institutions
and systems constituting the local administration, the urban and district economy, the
taxation system, maintenance of order and roads everything that was broadly within
the interests of the central authority. This general principle had specific dimensions
in particular kazas on account both of location and of changing demographic and
economic realities. The following summary presents the types of problem the kad of
Sofia oversaw as reflected in one sicill for .
The kad court, and the sicill, continued to be the place where appointments were
announced and recorded: of a new vali, of the kad himself, of the mtesellim (tax collector), of the kaymakam of the seyyids (local administrator for matters concerning the
descendants of the Prophet) in the kaza, of the kassam-i askeri (divider of askeri inheritances), of the mimarba (the city architect), and of city police officers and Janissary
garrison commanders. The kad registered permissions to seek and extract silver in the
district; he registered the representatives in situ of high-ranking military officials, the
metropolitan, low-ranking (Muslim) religious functionaries, and many others. Berats
of officials as well as of the various tax collectors were not simply read in the court but
were also regularly recorded in the ledger, a procedure which probably signalled the
official launching of their activities in the kaza.
The kad court continued to play an important role in overseeing the collection of
taxes and state deliveries, and was instrumental in the resolution of conflicts that arose
between the parties involved. The sicill contains berats for the collection of various
taxes, levies and tithes payable in cash; for deliveries of carts for a military campaign,
hunters for the imperial hunt, and provisions for the armoury; for the taxes of various groups, including the kefere taifesi (that is, the Orthodox Christians), Jews and

Rossitsa Gradeva

Armenians, gypsies, yrks (nomads) and voynuan,42 inhabitants of the city mahalles,
or simply the whole reaya of the kaza, and sometimes even those in smaller adjacent
With regard to disputes over taxation, the sicill of shows a new development as compared to the first half of the seventeenth century. Several entries note
that, due to changed circumstances in the kaza caused by the dispersal of villagers
and the impoverishment of those remaining, the local ayan-i vilayet had gathered for
the purpose of reassigning the respective tax obligations demanded in the imperial
defter. Although no places are specified, it is clear that these gatherings were outside
the court, and there is no indication that the kad attended the meeting. Yet, the new
defter was taken to the court and recorded in the sicill, probably in order to give it legal
authority.43 The same procedure the gathering of local power-holders in one place
to carry out the distribution (tevzi) of the obligations was followed also for various
deliveries. The obligation to send , hunters for the imperial hunt near Edirne
was allocated among the respective ayan and village zabits in the kazas of Sofia and
Breznik.44 In these and other cases concerning the obligations of particular communities or professional groups, it seems that the kad merely registered and sanctioned a
decision reached elsewhere.
Closely related problems concern the timar system.45 Alongside the registrations of
berats for timar holders, sicill entries signal already known trends and developments:
the gradual incorporation of timars into the tax-farming system and consequent tax
collection matters; the appointment of representatives; the delineation of borders
between timar villages; and widespread attempts at cheating, such as reporting the
death of or the failure to perform obligations by a titular spahi by a candidate for the
respective timar or zeamet. Many orders urge the kad to investigate and ascertain the
correct situation.
Traditionally, kads oversaw local mobilization for military campaigns.46 These
orders were read in court, and the kad was broadly authorized to see that all concerned
fulfilled their obligations. In , local commanders and timar holders from the
sancak were summoned to Isakcea on ruz-i hzr (St Georges day, the traditional
first day of spring campaigns), while Janissaries engaged in the defence of the serhad
(frontier) and of Babada were called to their place of duty.47 Also traditional is the
role of the kad in fighting the omnipresent brigandage and oppression. He not only
investigated specific cases but also compiled lists of the hayduds (brigands) and their
accomplices, probably to be sent to the capital city or to the vali. He was expected to
ensure the proper functioning of the systems for maintenance of order, placing guards
in dangerous places and restraining those engaged in the maintenance of order from
committing oppression themselves.48
Interestingly, this sicill contains very few documents concerning the functioning
of the esnaf and the market in Sofia, or in general. No fixed prices or changes in esnaf
membership are recorded, although contemporary sicills from Ruse contain these. The
reason for this is unclear, as the kad was clearly involved in various orders concerning the campaign obligations of some esnaf and in decisions about communities, or
fixing the salaries of the tellaks (attendants) in a local hamam. He was approached
also with regard to monopolies in the sale of various products, including alum from

A kadi court in the Balkans


The kad was assisted by a number of specialized officers, staff and other more or
less informal participants in the judicial process. Among the first group, apart from
the various police and military officials, were the naib, the muhtesib (market inspector), the kassam-i askeri and the katib (scribe), as an intermediate figure between the
specialized and the ordinary staff. Naibs appear in two typical roles, as deputies of
kads of all ranks in kazas and nahiyes and as assistants of a sharia judge.50 In Sofia
they often figure as the representatives of the court in commissions sent to investigate
criminal acts, in civil and family law cases or about non-Muslim places of worship,
which implies a certain level of legal qualification. Sometimes naibs figure among the
instrumental witnesses to cases. In Sofia, as in the empire generally, the muhtesib was
attached to the kad court. His competences included in the first place ensuring the
proper functioning of the market and the imposition of fines on craftsmen and tradesmen in loco for minor offences, while more serious transgressions of market law and
discipline were taken to the kad. Another obligation of the muhtesib was collecting
due debts as well as finding sureties for insolvent debtors. Muhtesibs also had prisons
under their administration, mainly for debtors but also for brigands.
The kad was also assisted by scribes, court attendants and gate-keepers. In the Sofia
court, as elsewhere in the Balkans, katibs were charged with a variety of tasks which
exceeded the ordinary scribal work. Apart from their obvious role of compiling documents and recording them in the sicills, the several scribes frequently led investigating commissions, which obviously required some level of legal qualification from at
least some of them. Often they are listed among the instrumental witnesses as well.
Muhzrs were among the kads assistants most frequently mentioned. Their primary
obligation was summoning defendants who had failed to appear in court. They also
accompanied muhtesibs when collecting debts. Very often they served as instrumental witnesses as well. Throughout the period under consideration they had a chief in
Sofia, a muhzrba, but it is not clear whether he had any specific tasks different from
those of ordinary muhzrs. Emins of the court were most probably engaged in collecting court taxes. They too appear among the witnesses. Here we should probably list
also the interpreter (tercman) who occasionally appears in the Sofia sicills, usually
among the instrumental witnesses, and who mediated between non-Ottoman-speaking zimmis and the court.51
Local inhabitants took part in the work of the court in a variety of capacities. Apart
from witnesses providing evidence for specific cases, they were employed as experts,
who gave opinions in disputes as muslihun (mediators) and instrumental witnesses.
The institution of the muslihun is closely related to the Ottoman judicial system.52
Reconciliation was a widespread mechanism of conflict resolution throughout the
empire, officially recognized and accepted by the judicial authorities, who often merely
recorded an already reached agreement. Mediators intervened in disputes between
individuals, between an individual and a territorial community (mahalle or village),
between Christians, between Muslims and in mixed cases, and in a wide range of
disputes, including homicide. This, and the fact that only a very few cases regarding
minor offences reached the court, raises the issue of the real role of this institution in
the legal system and of the law applied in it.
One of the most debated topics concerning the judicial systems of Islamic states is

Rossitsa Gradeva

that of the instrumental witnesses (uhud ul-hal) whose names are listed at the end of
kad court documents.53 Their presence presumably rendered legality to the document
issued by the court, and they could serve as potential witnesses to the correctness of
the procedure in case the decision was contested. They might also have advised the
kad about local practices. In Sofia and other parts of the Balkans the instrumental
witnesses were neither a limited circle of reliable people nor bearers of specialized
legal knowledge. Very often we see outstanding townsmen and members of the court
among them, including former kads of Sofia and kads currently of other places, naibs,
scribes, muhzrs, Janissaries and other military ranks, teachers, mosque functionaries
and mahalle imams, but also craftsmen and traders, people without any identification, and members of the entourage of the provincial governor, especially in the later
sicills. In some cases these witnesses obviously had an interest in the outcome, as with
inhabitants of the same mahalle, or even neighbours, in the case of property transactions or criminal investigations; or when they had similar professional or confessional
affiliation with at least one of the parties. This, however, is not always the case. An
important change in the membership of the witnesses to the event during the seventeenth century is the significant presence in the latter part of the century of the Muslim
leadership people who can be identified as ayan and their dependants.
Finally, another group drawn from the local population who participated in the
work of the kad court consists of the so-called experts (ehl-i hibre, ehl-i vukuf, ehl-i
hiresi) and the commissions sent to establish the circumstances in a variety of cases,
which occasionally included specialists. They gave opinions on vakf expenditures,
church reconstructions, transacted objects and construction, in criminal cases and in
disputes between esnafs. The opinion of these specialists usually provided the basis for
the courts decision. Despite the significance of their advice, their names and professional characteristics often remain unknown to us, hidden behind the generic term
experts. Unfortunately this makes it impossible to draw systematic conclusions both
as to what kind of people were called upon in different cases and as to who would
initiate such an invitation the parties concerned or the kad. This body nevertheless
represented a complex of local interest and knowledge working in co-operation with
the local sharia court and policing functionaries.

The sharia judges and the court continued to be approached in a wide variety of judicial cases and to oversee a number of military and administrative officials, as well as
many others in between. The kad court served as the focal judicial institution for the
local population of all groups, although it was the court primarily of Sofian Muslims
and much less for other religious communities or for villagers. It issued documents
which served as legal proof of property rights, legalized agreements reached outside
and inside the court. It was a body of authority whose decisions were most likely to be
enforced. No doubt this had made it attractive for non-Muslims when they were in a
position to choose between their own and the alien court. Kads adjudicated mainly
in contested cases, while in criminal cases they acted in close co-operation with officials involved in securing public order; in the field of kanun, especially in questions
related to the timar system and to taxation, they co-operated with the vali and his
divan. In Sofia towards the end of the seventeenth century, an even closer collaboration

A kadi court in the Balkans

developed between the governor of Rumeli and the kad, with the former attending
some sessions of the sharia court when it acted as an appellate institution, in cases
involving important local people or deemed of public importance. On other occasions, the kad of Sofia participated in the dispensation of justice at the valis divan
and then recorded the minutes in his sicill, both developments reflecting the further
integration of kads in the administrative system and the growing importance of the
governor as part of the judicial system. In the long run this would result, by the midnineteenth century, in the gradual transformation of the kad court into a purely sharia
one, which served primarily, but not yet exclusively, the religious needs of the Muslims
in a kaza, and in the loss of its administrative prerogatives.
This process was paralleled by a gradual decrease of the role of sharia judges in the
lives of the empires non-Muslim subjects. It should be pointed out that, despite the
overwhelming majority of the Christians in the district of Sofia, relatively few cases
between Christians only are recorded in the kad sicills as compared to those between
Muslims. Jews and Armenians in Sofia, though an urban population, were much less
attracted by the possible advantages of the sharia court and approached it mainly
in mixed cases i.e., with members of other religious groups, Muslims in particular. While Orthodox Christians approached the sharia court for a variety of family
and inheritance law disputes, even to get married and divorced, or for the division of
estates in cases of intra-family discord, this is not the case with Jews and Armenians.
To explain the decreasing interest of non-Muslims in the kad court, we should bear
in mind that cases between townsmen are also more numerous than those involving
villagers, and in the seventeenth century most of the towns in the Balkans, and Sofia in
particular, were already inhabited by a prevalent Muslim population. In the case of villagers, major obstacles were the dangers of the journey and transport difficulties, as well
as the expenditure implied by a stay in town. Very important in shaping the attitude
of non-Muslims to the kad court is the fact that it based its decisions on the sharia,
while non-Muslims could turn to alternative institutions Orthodox Christian, Jewish, Armenian, and Catholic religious and self-rule courts whose importance seems
to have been growing from the seventeenth century in parallel with the strengthening
of communal structures. Sharia courts nevertheless remained throughout the Ottoman period, until the reforms of the nineteenth century, the major judicial institution
in the provinces for Muslims in particular but also for all Ottoman subjects, and an
administrative heart of any district.

For an overview, see Agmon and Shahar for the territory of modern Turkey and the Middle East, and Gradeva for Balkan historiography.
Gradeva a.
Gradeva a.
E.g., Sencar Muizeddin Mehmed Efendi, who between and , when he died in the
city, spent more than seven years in six appointments.
Probably the most prominent monument founded by a sixteenth-century kad is Seyfi Efendis
Banebai/Molla Camii in the city centre, built in the s by the school of the Ottoman chief
architect, Mimar Sinan, and still functioning. The other components of Seyfi Efendis vakf
a dersiye (schoolroom), a han and a kervansaray have long since disappeared, as have most
of the other monuments of Ottoman culture in Sofia. Several sixteenth-century kads donated

Rossitsa Gradeva

funds for the construction and maintenance of mektebs and medreses, for mosques and tekkes,
and their names were commemorated in the names of city neighbourhoods.
Gradeva b.
Gradeva : .
By Melek Ahmed Paa (Dankoff : ).
Sencar Muizeddin Mehmed Efendi resisted the governors demand, rode on horseback to
Istanbul and reported the case to the sultan. The oppressive paa was summoned, interrogated
in front of the kad and executed (Uur : ). While numerous grievances against
provincial kads have been found in seventeenth-century registers of complaints, these were
usually against judges of lower rank or naibs (deputy kads), who often stayed for very long
terms or were even locals. So far, I have found only one against a kad of Sofia, in (AD
4, sira 197/doc 1613, September 1670).
The sicills used here are S bis (), S (), S (tereke defter, ), S
(), S () and S (), all housed in the Oriental Department at the
Sts Cyril and Methodius National Library in Sofia. On the sicill collection in Sofia, see Ivanova
. For concerns about the reliability of such records generally, see Zeevi ; Gradeva
: ; Ergene : ; Agmon and Shahar : .
Despite prescriptions for the keeping of sicills (Akgndz : ) it seems that recording
very much depended on the personal wish of the parties.
Gradeva .
See, for example, R , of . Cf. Raduev et al. : doc. , . Another
register from Ruse (Ruse Historical Archive, no. , dated ) suggests that this may have
been the local practice. However, in Sofia, vakf documents were registered in the ordinary
sicills: S contains evidence of both new and continuing vakfs.
S contains exclusively divisions of estates and documents relating to inheritance problems.
See, e.g., Dimitrov ; Gradeva c: .
Muji .
For similar observations on an eighteenth-century court in Anatolia, see Ergene : .
On the theoretical framework of legal pluralism/diversity, weak and strong, and as an attempt
to deduce Ottoman realities, see Shahar .
Heyd : .
Majer : .
While much remains to be discovered about the specifics of each of these series, the ahkam have
attracted least scholarly attention. There were also various estate courts for ulema and for
the military which drew these groups partly out of the kad jurisdiction, especially in criminal
cases. There is little research on their role and functioning within the judicial system, or for the
respective group.
Ursinus ; Gradeva a. My usage of the term mazalim adheres to Heyds the removal
of wrongs (: ).
Hacker ; Gradeva ; Kermeli and her essay in this volume, chapter .
nalck : ; Gradeva c: .
Slot : ; Kermeli : .
Pantazopoulos [] : ; Gerber ; Gradeva : ff.; Kermeli :
Shaham ; Shahar .
Gradeva d. Elsewhere, however, the situation could be very different; cf. Veinstein .
Ivanova , .
S bis, p. , docs IXV. These, however, are unique in the sicills in Sofia. They might have
continued to be issued but were not registered.
Laiou : .
S , f. v, doc. I, of .
Ivanova : .
Recorded in S and S . The meaning of the term divan efendisi is uncertain here.

A kadi court in the Balkans

E.g., nahiye of Razlok, S , /I, /I; for clashes, S , /I.
S , /II and /I; S , /I; S , /II resp.
E.g., S contains several vakfiyes, documents related to the running of Mehmed Sofu Paas
foundation in the city, changes of mtevelli, inspection of repair work, and financial loans.
S , /I.
E.g., S contains several documents relating to the dismissal of the current metropolitan;
none have yet been found regarding the Jewish community.
Gradeva .
S , /I, /I, /II, /I.
Non-Muslims serving in the early centuries as auxiliaries in the Ottoman army and, from the
seventeenth century onwards, mainly as grooms in the sultans stables.
S , /III, for bedel-i celepkean; /I, for avariz; and others.
S , /I.
On the timar, or dirlik, system, see the essays by Metin Kunt and Gbor goston elsewhere in
this volume (chapters 7 and 15 respectively).
Gradeva 2004c.
S , /I, /II.
S , /I, /II, /III.
S , /I, /I, /III, /III.
Veinstein .
Veinstein ; iek .
Tamdoan .
Jennings : ; Canbakal : .



Amy Singer

verse from the Koran, inscribed over the gates of more than one Ottoman imaret
(public kitchen), reads:

And they give food in spite of their love for it, to the needy, the orphan, and the
captive [saying]: We feed you only for the Face of God; we desire no recompense
from you, no thankfulness.1
With this message displayed so prominently, it is no surprise that imarets were long
considered comparable to modern soup kitchens, feeding modest free meals to the
poor, including widows, orphans, the aged, the sick and the infirm, as well as students
and ascetic mystics. However, the range of clients eating daily, free of charge, at Ottoman imarets was much broader economically and socially than this list suggests, as
were the menus, service and settings.2 The following discussion explores Ottoman
imarets in the context of Ottoman charitable practice, material culture, consumption
habits and architecture. It considers imarets through the prism of their buildings and
locations, their clients and the food they served.
Ottoman charity was framed by the principles and practices of Muslim charity.
Annual zakat (alms) payments are one of the five cardinal obligations of all Muslims. In addition, the Koran and the hadith (sayings about the words and deeds of
the Prophet Muhammad) frequently recommend giving sadaka (voluntary charity).
Sadaka donations can be as small as a blessing or as large as an endowed complex of
buildings offering social and welfare services to thousands in the form of a mosque,
school, hospital, hospice, caravansaray, bath house, public kitchen or public fountain. Imarets are therefore physical evidence for the charitable donations of individual
All imarets were established as vakfs, charitable endowments sometimes known as
sadaka cariye (ongoing/enduring charity). Vakfs were the usual means by which social
and welfare services were established and maintained throughout the Muslim world,
beginning perhaps from the earliest days of Islam. Many of these services were those
provided today by government welfare offices (education, health, poor relief), but they


Figure . Inscription above the entrance to the imaret of Mihriah Sultan (d. ), mother
of Selim III, at Eyp, Istanbul. Koran : : We feed you only for the sake of God;
we desire no recompense from you, no thankfulness. Authors photo.

also included ritual facilities (mosques, ablutions pools, cemeteries) and infrastructure projects (water supply, roads, bridges). Provisions for building maintenance are
included in vakfiyes (endowment deeds), signalling that the complexes were imagined
as enduring through time, a permanent feature in the socio-economic life and built
fabric of cities.4
Beneficent food distributions existed in the Muslim world before the Ottomans.
Among the oldest is the simat al-Khall, the Table of Abraham, in Hebron. Ascribed
to Abraham and attested by the eleventh-century Persian traveller Nasir-i Khusraw, it
provided daily to anyone coming to Hebron a loaf of bread, a bowl of lentils cooked
in olive oil, and raisins.5 Under the Fatimids and Mamluks in Egypt, the sultans distributed food on special occasions such as Ramadan; in times of hardship they also
engaged members of the elite to contribute donations of food aid. Purpose-built kitchens contemporary with the earliest Ottoman principality existed elsewhere. In Tabriz,
the foundation of Rashid al-Din () included separate kitchens and menus for
travellers and for orphans, students and sufis.6 However, it was under the Ottomans
that imarets became a widespread feature of multi-purpose endowments, serving meals
daily to defined groups of beneficiaries.

Amy Singer

Imarets were founded across the Ottoman empire, throughout its history. The first
was built in the s, the last perhaps in , and at least one, at Eyp in Istanbul,
is said to have operated continuously since the late eighteenth century. They were
located mostly in the towns and cities of Anatolia and the Balkans, more rarely in
rural areas or villages, and were less prevalent, too, in the Arabic-speaking provinces of
the Middle East and North Africa. The kitchens numbered in the hundreds, but they
were not all founded at the same time and did not all remain open without interruption. More were founded in the first three centuries of Ottoman history than in later
periods. They ranged in size from neighbourhood or small provincial facilities offering
meals to one or two dozen people, once daily, to the large imperial kitchens that could
feed many hundreds twice a day. The largest clusters of imarets were in Istanbul, perhaps as many as fifty, and in the former Ottoman capitals of Bursa and Edirne, which
had eight and eleven, respectively. Other large or important towns also had several
imarets Amasya (eight), Iznik (ten), Manisa (seven), Salonica (seven) and it was not
uncommon for a sizeable town to have at least two or three (Afyon, Aksaray, Belgrade,
Damascus, Dimetoka, Kastamonu, Mecca, Skopje, Trabzon).
There is no official list of all the Ottoman imarets, and one is difficult to compile.
The word imaret itself can be misleading. The original Arabic word imra means
habitation and cultivation or the act of building, making habitable. In Ottoman
Turkish, the word became imaret, and was used with two additional meanings. It
signified a construction project of one or more buildings, a meaning derived from the
use of the complexes to found and develop new neighbourhoods. Complexes also contributed to the creation of new settlements or fortified outposts by providing the core
of basic physical infrastructure and services needed for a new Muslim community or a
stopping place. More narrowly, imaret was the public kitchen, one of many elements
in a complex, sometimes specified as imaret-i darz-ziyafet, the building for feasts,
or imaret-i darl-itam, the feeding building.7 This chapter uses the word imaret
only to refer to the public kitchens and their subordinate facilities: storerooms, cellars,
refectories and baking ovens.
The fact that several institutions distributed food created additional confusion about
imarets. Especially in the first two centuries of Ottoman rule, it seems that zaviye-imarets and imaret-mosques overlapped functionally. Moreover, free meals might be had
in a sufi lodge (tekke, takiyya, hanegah or zaviye), a caravansaray, an imperial palace,
or the home of a wealthy person. It is unclear whether these institutions more often
competed or co-operated with each other, especially in their urban settings. It is also
unclear how individuals chose among one or more institutions, though such strategies
may have been an integral part of urban survival for some. The collective impact of
these institutions on the populace is not easy to calculate, but it has been estimated
that, in the sixteenth century, up to per cent of the population in Edirne and Istanbul received food daily at imarets.8


The discrete history of imarets was mostly neglected until the end of the twentieth century. When mentioned, they were conceived as largely static and unchanging, because
they were studied primarily from their prescriptive endowment deeds (vakfiye). Also,
they were usually portrayed as auxiliary appendages to the mosques and medreses that


took centre stage in Ottoman complexes, serving as the showcase elements of Muslim
faith and Islamic learning. The first edition of the Encyclopaedia of Islam (published
) has only a brief paragraph on the term. The second edition (published
) has no entry at all, and imaret is covered there primarily in the general
article matbakh (kitchen). Godfrey Goodwin repeatedly mentions imarets in his classic History of Ottoman architecture and defines imaret as a soup kitchen in his glossary,
but it is not in the index. However, his Appendix II is an extended discussion of The
vakfiye and imaret system, where imarets are complexes, not kitchens. Both mer
Lutfi Barkan and Halil nalck have asserted the importance of this vakf imaret
system for the development of the empire.9
As for somewhat longer works on imarets as kitchens, two on the second Ottoman
capital, Edirne, offer an extended catalogue of local imarets and an analysis of their
collective impact.10 A more recent publication is this authors monograph Constructing
Ottoman beneficence: an imperial soup kitchen in Ottoman Jerusalem, which investigates
the foundation process and early operation of the imaret of Haseki Hurrem Sultan (d.
), wife of Sleyman I.11 The same imaret was studied extensively by Peri for the
insights it afforded on eighteenth-century vakf foundation and management in Jerusalem.12 Broader discussions of imarets include an entry in a multi-volume collection
on Turkish history.13 Most recently, the fifteen wide-ranging articles collected in Feeding people, feeding power define the problmatiques of imaret research, identify relevant
source materials, and delineate directions for future study.14
There is no shortage of source materials about public kitchens. Thousands of Ottoman endowment deeds survive, for large and small foundations alike. One deed could
run to thousands of words, describing in greater or lesser detail the intentions of the
founder; the institution to be established; the services provided; the staff, salaries,
equipment and furnishings; and the endowed revenue-yielding properties. For imarets, the deeds often stipulated the menu, ingredients, kitchen staff, budget for foodstuffs, equipment and salaries. Vakifyes provide blueprints of founders intentions,
expectations and ideology, all of which were culturally shaped and calibrated to the
specific size and place of the individual foundation. Other relevant written records
include tapu tahrir defterleri (Ottoman revenue survey registers) and local kad sicilleri
(judicial court records), as well as evidence found in Ottoman chronicles, and the
reports of Ottoman and foreign travellers. Many buildings remain as evidence, some
in good repair or even in use. Some kitchen equipment (grain storage chests and mills)
has been found, and it is possible that more cauldrons, ladles or bowls may yet be
Annual or periodic muhasebe defterleri (incomeexpenditure accounts registers)
exist for many of the imarets. They contain detailed lists of foodstuffs purchased and
stored, the names and positions of people employed in or benefiting from the foundation, and the sums and sources of annual endowment revenues. The account books
are useful companions to endowment deeds and literary texts for understanding the
changing capacity and functioning of imarets. More detailed registers listed food and
bread recipients, cash payments in lieu of meal allocations, daily storehouse revenues
and expenditures, and detailed registers of both people who received cash stipends but
no free meals and employees who received salaries.15 Altogether, the account books are
an unparalleled source about foodstuffs and imarets, as well as about consumption and
nutrition norms.16

Amy Singer

Together with zaviyes, imarets were a settlement mechanism of the Ottoman conquests
westward across the Balkans and further south and east in Anatolia. Each institution provided food and shelter to wandering dervishes, travellers, students, merchants
and the poor. They were one aspect of a general policy that aimed to create a stable
Ottoman presence in each town and city of the empire, and represented a significant
Ottoman investment in the provinces.17 No clear architectonic genealogy of Ottoman imarets has been established. Scholars point to the Seljuk caravansarays as one
likely model for public kitchens, but the earliest imarets were small institutions. The
caravansarays, even more modest ones, included sleeping, stabling and storage spaces
together with a kitchen, mosque and bath, enclosed by an external wall; they provided
free, secure lodging and food at regular intervals between the towns of Anatolia.
The T-shaped mosques that housed zaviye-imarets and imaret-mosques in the first
two Ottoman centuries were multi-functional. Their central space was reserved for
prayer and sufi rituals, while secondary spaces like the side iwans (rooms) were for
sleeping and perhaps eating. Cooking was probably done in a separate building or
outside in the courtyard. This flexible space accommodated the various practices of
Islam favoured by the gazi fighters and their sufi shaykhs, who also provided spiritual
guidance for the semi-settled nomads and new converts in the Turkish principalities of
western Anatolia and the newly conquered Balkans. Examples include the fourteenthcentury mosques of Orhan I and Bayezid I in Bursa, as well as the slightly later ones of
Bayezid Paa in Amasya and Mahmud Paa in Istanbul, all of which had imarets.18
By the sixteenth century, the term imaret was more often used exclusively to mean
public kitchen (modern Turkish, ahane). The narrowing of the meaning seems to
have paralleled the architectural evolution of Ottoman mosques into extensively articulated complexes in which each function was housed in a separate structure. This is
reflected in the chronicle of buildings designed by the consummate Ottoman architect Mimar Sinan (d. ), which included imaret as a separate category or building
type.19 In her study of Sinan, Necipolu translates imaret as hospice, emphasizing the
fact that many imarets operated in close proximity to accommodation for guests, such
as a caravansaray, han, tabhane or tekke.20 In such cases, the imaret also fed the guests
and sometimes provided fodder for their animals.
Lowry claims that the earliest known Ottoman imaret-like institution was the
hanegah built in by Orhan Gazi (), in the village of Mekece, east of
Iznik (Nicaea). No explicit provision is made in its vakfiye for serving food, only that
the manager of the endowment should expend what is in the interests of the traveling
dervishes, the poor, the strangers and mendicants, and those in search of knowledge.
Lowry makes an undocumented but informed leap to call the place an imaret, assuming that food and shelter were in the interests of this group.21
The fifteenth-century Ottoman chronicler Akpaazade (d. after ) wrote that
Orhan Gazi established the first imaret in Iznik in , shortly after the besieged
city surrendered to the Ottomans in . By the time Akpaazade used the term
imaret in the later fifteenth century, he could apparently do so unambiguously. Orhan
reputedly inaugurated his imaret personally by lighting its first kitchen fire.22 Another
imaret was founded in Iznik by Murad I () in , in honour of his mother,
Nilufer Hatun. By the time of the survey register, Iznik had five imarets and


town quarters named for two of them.23 More than one of these imarets was founded
by members of the andarl family of early Ottoman commander-vezirs.
Among other prolific founders of early Ottoman imarets were Evrenos Bey (d. ),
one of the most successful Ottoman commanders in the Balkans, and his descendants.
They founded at least twelve imarets across northern Greece during the fourteenth and
fifteenth centuries, together with other buildings such as mosques and sufi tekkes. As
in Mekece and Iznik, these institutions probably served many local Christians as well
as the Muslim followers of Evrenos, encouraging people to remain in newly conquered
areas. Whether appreciated for their practical benefits or as reflections of the value
placed by Muslims on charitable endeavours, the endowments may also have played a
role in attracting people to convert to Islam. Lowry counted imarets altogether in
the Ottoman Balkans, mostly dating from the first years of Ottoman rule. Considered according to contemporary national boundaries, they include Greece (sixtyfive), Bulgaria (forty-two), Albania (nine), countries of the former Yugoslavia (twentynine), Romania (two) and Hungary (two). Alongside the imarets, there were over
zaviyes in northern and central Greece, such that institutions offering food and shelter,
even if modest, must have been ubiquitous in the early Ottoman period.24
Bursa, the first Ottoman capital, had many imarets, but, unlike in Iznik or in the
small towns of the Balkans, they were mostly part of large imperial mosque complexes.
These complexes, funded by the spoils of conquest, were built by Orhan Gazi, Murad
I, Bayezid I (), elebi Sultan Mehmed () and Murad II ().
That of Orhan Gazi, destroyed in , was the earliest in the city and the first imaret
to exist as a free-standing building. It was erected in close proximity to the founders
mosque and its other attendant structures. The earliest surviving free-standing imaret
building belongs to the Yeil complex built by elebi Sultan Mehmed I. Compared
to later imarets, its two rectangular rooms of brick and uncut stone, originally roofed
in wood, seem casually planned, plain and functional.25 Surviving fourteenth-century
structures such as the Nilufer Hatun imaret in Iznik or the Evrenos imaret in Komotini/
Gmlcine afford little sense of how food preparation and distribution were managed.
Murad IIs Bursa complex shows the structural direction taken by imperial imarets
in the mid-fifteenth century, with huge indoor hearths for cauldrons and refectory
spaces. Unfortunately, the kitchens of Murad IIs two complexes in Edirne the
erefei and the Muradiye have disappeared, as has that of Fatihs enormous complex
in Istanbul. What imperial imarets had become by the later fifteenth century is readily
observed, however, in the imarets of Bayezid IIs extant complexes in Amasya (),
Edirne () and Istanbul (), all free-standing stone buildings within the
general perimeter of the complex.26 In , the Amasya imaret was in use as a soup
kitchen, while that of Edirne was restored in .
The largest imarets were built in the sixteenth century. Over twenty were constructed under the direction of the imperial chief architect Sinan, of which only eight
survive. Five of these were buildings with their own independent courtyards (e.g.,
Sleymaniye in Istanbul, Muradiye in Manisa), spaces that integrated cooking, storage and dining facilities, housed separately from the other activities of the complex.
Three others, however, shared their courtyards with the mosque of the complex (e.g.,
Sulaymaniyya in Damascus).27
Eighteen imarets functioned in Istanbul during the sixteenth century, all connected to mosque complexes. Nine had been built by sultans, nine by vezirs or other

Amy Singer

high-ranking persons. Yet, large as they were, these facilities could not have fed everyone in need of a meal, whatever the source of their need. Yerasimos has identified an
additional endowments that provided food in Istanbul, all of them smaller and
based in neighbourhood mosques (mescid), zaviyes or schools, mostly feeding local
poor people. These distributions were not daily, but occurred primarily on religious
holidays, and complemented those provided by endowments for clothing, water supply, urban maintenance and tax assistance.28
Some imarets were constructed by incorporating older structures associated with
previous rulers. The Salimiyya mosque and imaret in Damascus took over the Mristn
al-Qaymar while the Haseki Sultan imaret in Jerusalem took over the house of a
prominent Mamluk woman. This strategy afforded the founder a space in an existing
urban fabric, essentially introducing an Ottoman presence while erasing the monuments of previous rulers and their affiliates. The buildings and their activities then
became a focus of Ottoman identity and attention. Selim I gave his Damascus imaret
preferential treatment in distributing foodstuffs, and perhaps for this reason it was
closed during a local revolt against Ottoman rule in .29
Recalling the Seljuk hans, some complexes with imarets were also built in more
isolated places, specifically in order to ensure the safety and comfort of travellers on
the main roads of the empire. A very early example of this is the imaret built at one
end of Uzun Kpr, the long bridge erected by Murad II and part of a project to
secure the marshy, bandit-ridden road south of Edirne. According to the sixteenthcentury historian Mustafa Ali, Murad himself served food there when it opened. Imarets were included in the defensive complexes built by Sleyman I at Belen and Sokollu
Mehmed Paa at Payas () on the road between Adana and Aleppo.30 In Syria,
imarets seem to have been tied more to pilgrimage routes. In Damascus and Jerusalem,
they are both disproportionately large, emphasizing their role in sustaining pilgrims
to the holy sites.31
Kitchens in Ottoman imarets and sufi tekkes had similar elements: kitchen (matbah), refectory (mekelhane or taamhane) and pantry or storeroom (kiler). Tekke kitchens also often had a spiritual dimension lacking in imarets, since cooking food was
sometimes part of sufi rituals and the person in charge of the kitchen held a high rank
in the orders hierarchy. The kitchen could also be a place of study and initiation,
notably for the Mevlevis, who were taught to whirl there. For the Bektais, the person
responsible for educating novices was called a dede or ser tabbah (head cook). The
mythical black cauldron (kara kazan) symbolized the order itself, and was also adopted
by the Janissaries, representing their particular attachment to the Bektais, as well as
signalling revolt when overturned.32

Contrary to early twenty-first-century expectations, not all or even most of the
people who ate in Ottoman public kitchens were impoverished. Rather, the right to a
meal was a function of social or economic status, employment or profession. Imarets
imposed different kinds of restrictions on their clients: who could eat, how, what,
where and when. All of these conditions point to a system conceived not only to assist
people or distribute to them food they deserved, but also to control them and to reinforce existing socio-economic hierarchies.


Figure . Illustration of the vakf complex built by Sleyman in the s at Belen (Bakras), a
halting station on the route from Anatolia into Syria, through the mountains north of
Aleppo. The complex included two caravansarays, a mosque and a dervish convent
which provided food for travellers. From Seyyid Lokman, Hnername (),
Topkap Saray Library, H. : b.

Amy Singer

In Istanbul, the imaret of Fatih (the conqueror) Mehmed was perhaps the largest
anywhere in the empire. By the mid-sixteenth century, approximately , people
were fed there twice a day. Among its regular clients were visiting dignitaries, travellers, scholars and students at the colleges attached to the mosque; the doorkeepers and
guards of these colleges; the students of three other nearby colleges and four nearby
dervish lodges; student candidates and their eight proctors; fifty-six members of
the imaret staff; forty-seven hospital staff members; and fifty-one other functionaries
of the complex, including employees serving at the mosque and tombs. When all these
people had been fed, leftovers were distributed to the indigent poor.33
The Sleymaniye imarets list of around , diners resembled that of Fatih,
although the vakfiye noted that orphans and children of the poor who were present in
the primary school (mekteb) on any given day should also be served.34 At the smaller
establishment of Haseki Hurrem Sultan in Jerusalem, two meals per day were served
to people described in the endowment deed as poor and pious, together with the
kitchen staff of fifty, whoever was staying in the fifty-five guest rooms in the complex,
and the local sufi Shaykh Ahmad al-Dajjani, as well as sixteen of his followers.35 Still
more modest was the kitchen of Fatma Hatun in Jenin, which was to feed approximately fifty people once a day.36
The vakfiyes often stipulated that people were to eat in a specific order, which
reflected a social hierarchy among the diners. The conditions in which their meals
were served and the type or amount of food each person received reinforced the order.
At Fatih, the guests ate first, followed by the college scholars, students and staff; after
them came the students from the nearby colleges, the resident dervishes, and the
candidate students and their proctors. Next to eat were the staff of the imaret and the
rest of the Fatih complex. At the Sulaymaniyya in Damascus, people staying in the
guest rooms were served privately in their rooms, twice a day. In Jerusalem, people also
ate in shifts: imaret employees, then caravansaray residents, and finally the poor, who
were too numerous to eat together, so the men were fed first, then the women.37
Some imarets forbade people from removing food, with specific exceptions. At the
Sleymaniye no strangers could remove food in buckets. However, food could be
taken to poor scholars, the descendants of the Prophet, the blind, the paralyzed and
the sick.38 In Jerusalem, Shaykh Dajjani and his sufis collected food from the imaret
and brought it back to their residence across the city.39 Everyone else ate in the imaret
refectory.40 Even in the last-known imaret planned in Istanbul in , the removal
of food was explicitly prohibited, except in the case of someone sick at home who had
the right to eat there. Such cases, however, required approval from the local imam and
the muhtar of the quarter.41
Imarets often served medrese students (suhte/softa), like a university cafeteria in the midsixteenth century. Students mobility and potential for disruption were nuisance enough
to have them occasionally denied access or for guarantors to be demanded for their
behaviour, since complaints were filed that they plundered the kitchens and threatened
the staff. A certain standard of decorum was expected in imarets. However, according to
Mustafa Ali, the quality of food in Istanbul imarets was enough to raise a protest.42
Christians, local and foreign, seem to have been regular clients at some imarets,
becoming targets of the ideology and policy they embodied. At the end of the fourteenth century, the Bavarian Johann Schiltberger noted that the city [Bursa] contains .
. . eight hospitals [spitler] where poor people are received, whether they be Christians,


infidels [Muslims] or Jews.43 Visiting Bursa several decades later, Bertrandon de la

Brocquire wrote: There are very nice places, like hospitals. In three or four of these,
bread, meat and wine are distributed to those who want to take them in Gods name.44
The serving of wine may indicate that Christians were welcome guests. According to
the English traveller George Wheler, Bursas imarets were still open to all comers in
the later seventeenth century.45 Theodore Spandugnino, writing in the early sixteenth
century about the Fatih Mehmed complex in Istanbul, described
a superb building . . . the hospital is open to all, Christians, Jews and Turks; and
its doctors give free treatment and food three times a day. I have seen men of the
upper class and other grand persons lodging here, their horses being cared for.46
In the early eighteenth century, the imaret of Haseki Hurrem in Jerusalem served
bread and soup made with olive oil and some vegetables to each poor person who
asked for it.47
No first-hand account in a Jewish source has yet confirmed that Jews also availed
themselves of food in imarets and, if they did so, how this accorded with the dietary
laws of kashrut. However, Jews were a minority population in the empire, and so were
not a factor comparable to Christians in Ottoman political considerations. Moreover,
Jewish communities traditionally organized food distributions as part of a broader
poor relief system.48 Free meals worked as Ottoman propaganda to persuade newly
conquered populations that Ottoman rule did not aim to create hardship; nor was
Islam a miserly religion. Rather, imarets were part of Ottoman istimalet (good will or
accommodation), which also encouraged assimilation and conversion.49
One key question with only sporadic answers is how people might acquire the right
to eat at an imaret if they were not included among its originally defined beneficiaries.
Anecdotal evidence suggests that the right to eat could be claimed for a variety of reasons. Poor neighbours of an imaret widows, hacis, dervishes were sometimes granted
food shares through certificates.50 Formal petitions could be submitted to the goverment
requesting food from an imaret. Thus, the kad of Konya appealed to have the dervishes
at the tomb of Mevlana Celaluddin Rumi included in distributions from the newly
built adjacent imaret of Selim II.51 In Istanbul at the end of the sixteenth century, one
Ahmed, a crippled soldier, petitioned for meal rights because he could not work.52
Such rights could be willed or inherited, even divided in the next generation.53
Another phenomenon remains to be further investigated. In some imarets, individuals received food distributions far in excess of their daily needs, either because they
were assigned a large quantity of bread or because they held more than one job, each
one entailing meal rights. In at least one case, receiving multiple meals was expressly
forbidden. However, in others, the excess food was a matter of record. Meiers research
on eighteenth-century Damascus shows people who received entitlements authorized
by the kad and comprising combined distributions of bread from the two Damascus
imarets of Selim I and Sleyman I, as well as from what was called hinta al-fuqara
(wheat of the poor) from the village of Daraya. Some recipients were from wellknown local families, whose entitlement could be inherited from one generation to the
next.54 One can speculate that, in cases like these, people either fed other members of
their families or redistributed food to those needier than themselves, thereby becoming patrons and benefactors as a result of the charity they themselves received.

Amy Singer

Even from the small amount of research on imarets in different periods, it is clear
that the population eating in them varied over time. At Sleymaniye, for example,
the number of employees seems to have decreased while the number of beneficiaries
rose.55 The most significant change seems to have occurred during the nineteenth century, as the clientele gradually included more indigents. The change probably resulted
from several factors: the reorganization of vakf administration that accompanied the
tanzimat reforms; changing ideas about how to care for the poor, their proper role and
place in society; and the dislocations of population, which became more marked in
the latter half of the nineteenth century as a result of successive wars and gradual loss
of territory.56
Earlier changes in the way imaret beneficiaries were defined, as well as larger variations in local practice than those we currently understand, may also be revealed with
further research. In , Istanbul had fifteen functioning imarets. In April ,
all but two were closed by official order after the Young Turk revolution of , in
which older institutions, identified with Abdlhamid II, were eliminated and replaced
by modern ones. However, the substitutes were not adequate to distribute food to the
poor, and by the imarets of Fatih, Sleymaniye, Nuruosmaniye and Atik Valide
were allowed to reopen.57

The focus on what were perceived to be the more noble elements of an urban complex
of buildings is mirrored by the historiographical attention to foods of elite households and especially of the imperial kitchens (matbah-i amire) of Topkap Palace. The
imperial palace kitchens, although funded by the state budget and not by charitable
donations, operated in part like an imaret, and were denoted thus by Evliya elebi.58
The palace kitchens fed the sultan, the imperial family, the entire palace staff, visitors
to the palace, and soldiers when they came to collect their salaries. Foodstuffs and
prepared foods were also distributed from the imperial palace and from the grand
vezirs residence at the Sublime Porte (bab-i ali) to a wide variety of clients in Istanbul, notably poor and needy people in the city on special occasions. In the nineteenth
century, these included the sultans accession, royal birthdays and religious holidays.
While Topkap was the model for distributions, these occurred in a similar way from
the imperial palace in Edirne, from the Old Palace in Istanbul, and later from the
nineteenth-century palaces of Dolmabahe, ra an and Yldz.59
In terms of capacity, Topkap in the later sixteenth century fed over , people
daily, and on special occasions as many as ,. To accomplish this, the palace
operated nine kitchen units, employing sixty cooks, assistants, and a total staff
of , to ,. In the nineteenth-century, palace distributions of cooked food
rose significantly, and under Abdlhamid II the palace kitchens functioned as imarets
for their neighbourhoods. The situation reached a point where people were said to
be moving to the palace vicinity in order to save money on food. This situation was
seriously addressed under Mehmed V Reat (), after the deposition of Abdlhamid II.60
The menu at the palace kitchens was mostly rice pilav, meat, bread and sweet saffron rice (zerde); special meals included rice soup cooked with meat and parsley and
accompanied by baklava made with honey and walnuts.61 The quantity and quality of


ingredients in imperial palaces surpassed those in imarets, as did the variety of foods
served to the imperial family and important guests. But, for the staff and visiting subjects, meals may have resembled those served in imarets.
Meals were mostly served twice a day in imarets, except during Ramadan, when only
one was served. They always included bread. The regular cooked dishes were starchy
and mostly savoury. The most basic and common food was cracked wheat (bulgur) or
rice soup, with or without meat, perhaps enriched with salt, parsley, onions, cumin,
pepper, chick peas, squash and sour grape or yogurt and chard, plus bread. On Fridays
holidays and festivals, or for diners of higher status, richer foods such as dane (meat,
chickpeas, butterfat, salt and rice) were served, as well as sweet dishes such as zerde,
zrba (starch, honey, dried fruit, saffron), aure (a pudding cooked with dried fruit
and nuts, and whatever was available) or baklava. The special fare of dane and zerde
constituted holiday and ceremonial staples, expected and so placed on every table, no
matter the rank of the guest. At the circumcision feast of Sleymans sons Bayezid and
Cihangir in , dane and zerde also appeared on the tables of rich and poor alike.62
At the Fatih imaret in Jerusalem, approximately , loaves of bread were baked
and distributed daily. Travellers who came to stay at the caravansaray of the Fatih
complex were served honey and bread at the imaret immediately upon their arrival, to
revive them after their journey. The Fatih imaret could accommodate about highranking guests per day served at tables (sofra) laid for four. These guests received daily
meals of dane and sometimes zerde as well, the dishes that most others ate only once a
week. Sometimes guests had meat stew with plums and fresh fruits. Visitors of higher
status, such as the eraf (descendants of the Prophet Muhammad), had sheeps trotters (paa) served for breakfast as a great delicacy, as well as a dish made of pumpkin,
honey, jam, cinnamon and cloves, and generous portions of meat and rice. In contrast,
every two children were to share a bowl of soup, a portion of meat and two loaves of
bread. The standard serving for the majority of staff, college students and scholars,
though unspecified, was probably one serving of soup apiece and a loaf of bread.63
Everyone at the Haseki Sultan imaret in Jerusalem ate the same wheat and rice
soups with bread, similar to those served at Fatih, with larger servings specified for the
staff and travellers. Employees at the Jerusalem imaret received one ladle of soup and
two loaves of bread per meal, the guests one ladle and one loaf, while the sufis and the
largest category of the poor received one half-ladle and one loaf each per meal. On
Fridays, each person received dane and zerde, but the poor had to share a piece of meat
between every two of them, while the others had a whole piece each.64

The earliest imarets seem to have been relatively small institutions, founded by the
sultans, vezirs, and Ottoman frontier commanders such as Evrenos. As the empire
expanded, the buildings and their capacities also grew larger and more magnificent,
reflecting the increasing power and prestige of the Ottoman dynasty. Since imarets
and other imperial endowments were tied directly to the prestige of the dynasty,
their ability to function could reflect well or badly on the sultan and his government.
An inactive or badly run kitchen, as well as criticism like that written by Mustafa Ali,
could work to undermine its legitimacy.65 Yet it was no simple matter to maintain the
large kitchens. Although they may not have been either the most expensive structures

Amy Singer

to build in any complex or the most prestigious, imarets often seem to have consumed
a larger share of annual endowment revenues than any other single institution in the
complex. Not only did they require wages and maintenance, like every other building,
they also required regular supplies of food, firewood, water, and cooking and serving
equipment to accomplish their role, which included sustaining all the other employees
of the complex as well as guests, students and some indigents.
The imarets served an additional function, beyond those of distributing food and
contributing to the legitimacy of the dynasty, as noted above. In the fact of their
endowment, their names, their numbers in the empire, and in the manner of their
functioning, including their menus, imarets were distinctly Ottoman. Far from the
capital and the major cities, they served similar food, every day, at the same time of
day, to a predictable list of clients, with roughly identical variations on holidays and
festivals. Within this uniformity, however, variety did exist. The shapes of buildings
were modified by local materials and artisans. While the basic types of soups, festive
dishes and bread unified the imarets, local variations of additives, local varieties of
wheat, and traditions of what to add to aure pudding affected what was served, as did
local or temporary scarcity or availability. However, the overall tone was reflected in
the vakfiye instructions for two imarets in Syria, which said that food should be served
as it was established in other imarets and takiyyas. This deed also had to define explicitly dishes that needed no description in the central Ottoman lands, a fact emphasizing
that the Ottomans had extended their reach beyond the boundaries of a shared food
culture. Imarets were a stable fixture in the Ottoman landscape, not only a tool of conquest, settlement and perhaps Islamization, but also a means of Ottomanization. They
helped create a commonality of experience among Ottoman subjects in the Balkans,
Anatolia, and the Arabic-speaking provinces of the empire.66

* This research was supported by the Israel Science Foundation, grant # /
Koran : .
Meier (: ) also notes this contradiction and points out the political agenda evident in
the list of imaret beneficiaries.
Zysow ; Weir and Zysow ; Singer .
For a general introduction to vakf (Arabic: waqf ), see Peters et al. ; on renovations, see
Ergin .
Nasir-i Khusraw : .
Sanders : , , , ; Singer a:
. ; Zarinebaf-Shahr ; Kiel .
Ergin : ; Ayverdi : ; Inalck b: ; Singer : .
Barkan a; Gerber ; Yerasimos : . The rough calculations done by Barkan
and Gerber for imarets in Istanbul and Edirne, respectively, have recently been challenged in
Orbay a.
Huart ; Waines et al. ; Goodwin : ; Barkan a; Inalck b.
Kazancgl ; Gerber .
Singer a.
Peri , .
Singer b.
Ergin et al. .
The most extensive studies of these registers were begun by Barkan (b, and )
and have recently been continued in depth by Orbay (, a).


Faroqhi : .
Lowry : , ; Lowry : ; Kiel : .
Necipoglu : , ; Goodwin .
Tanman : ; S Mustafa elebi : , .
Necipoglu : ; Neumann : .
Lowry a: ; also Lowry , .
skpasazade : .
Ibid.: ; Lowry : .
Lowry : , , , .
Ergin et al. : ; Goodwin : ; Tanman : .
Tanman : .
Ibid.: .
Yerasimos .
Meier : ; Singer a: ff.
Necipoglu : , .
Meier .
Tanman .
nver : , which is a distribution list from /.
Krkoglu : .
Singer a: .
Meier : .
Singer a: ; Meier : .
Krkoglu : , .
Topkap Saray Arsivi, Defter /.
Stephan : .
BOA, Y.PRK BSK / Muharrem ( Sept. ), described in the catalogue as:
Hamidiye Camii ile Ertugrul Tekkesi arasnda padisah tarafndan yaplacak imaretin nizamnamesi ile msveddesi.
Neumann : ; Tietze : II, , .
Schiltberger : , cited in Lowry : .
Kline : , cited in Lowry : .
Wheler : , cited in Lowry : .
Spandounes : .
Morison : .
Ben-Naeh .
Lowry : .
Orbay a: ff.
Konyal : .
Topkap Saray Arsivi, E./.
Orbay a: .
Orbay a; Singer ; Meier : .
Orbay a: .
Ener ; zbek .
ztrk : ; zbek : ; Pakaln : II, .
Faroqhi : ; Evliya elebi : a.
Ertug : , .
Ibid.: , .
Ibid.: .
Faroqhi : ; on the circumcision feast, Tezcan .
Krkoglu : .
Stephan : .
Neumann : .
Meier : , ; Neumann : .




Derin Terziolu

he word sufi evokes timeless images of whirling dervishes and an ancient but
equally timeless lore of spirituality with only the most tenuous links to Islam.
Thankfully, there is now a substantial scholarly literature which shows that sufi
thought and practice also has a history, and that this history is inextricably connected
with the history of Islamic piety at large. As the etymology of the word sufi (wearer of
wool) reminds us, the early sufis had emerged out of the ranks of Muslim renunciants
in Baghdad between the late eighth and early tenth centuries. What distinguished this
group from ordinary renunciants was their inward turn, their search for a more intimate knowledge of God through a variety of techniques of self-transformation. Initially
a relatively marginal group, sufis had succeeded by the eleventh century in spreading
across a vast Islamic world, and, notwithstanding the controversies surrounding some
sufi beliefs and practices, normative sufism had become largely accepted as a legitimate
part of what had by then been established as Sunni Islam.1
However, it was only during the next four centuries a period which began with
the resurgence of nomadic tribalism under the Seljuks and ended with the formation of more durable empires by the Ottomans, Safavids and Mughals that sufism
became truly popularized. The organizational features of sufi communities were also
transformed during this period. Increasingly, it became usual for an aspirant sufi first
to find a perfect master and submit him or herself to his authority. Over time, these
communities of masters and disciples became organized into sufi orders or tarikats
(literally, way), each distinguished by a distinctive teaching method and a silsile, an
authoritative chain of spiritual transmission that traced each living master to masters
of the past.2 Nevertheless, the institutionalization of sufism was not an orderly affair,
and the late medieval period also saw some contrary trends, such as the rise of deviant,
nonconformist dervish movements3 and the transfusion of sufism with elements of
extreme Shiism (guluvv) throughout the eastern lands of Islam.
These transformations were still in progress when towards the end of the thirteenth
century the Ottomans first made a bid for political power in the borderlands between
a diminished Byzantine empire and a crumbling Seljuk sultanate of Rum. In this
particular setting, where Muslims were initially a minority, the institutionalization of
sufism took longer than in the more established lands of Islamdom. It was also inti

Sufis in the age of state-building

mately connected with two parallel processes: state-building and confessionalization.

The latter term refers here to the initiatives taken by Ottoman religious and political
authorities during the sixteenth century to refashion the attitudes and behaviours of
the empires Muslim subjects in conformity with the principles of Sunni Islam.4
Historians have generally approached the institutionalization and confessionalization of sufism in Ottoman lands through a conceptual framework which was first
formulated by Kprl in the early twentieth century and extensively utilized by Ocak
more recently. Generally, historians subscribing to the KprlOcak line assume
that the Seljuk and, after them, the Ottoman authorities (and the urban elites under
their rule) had always subscribed to Sunni Islam, but they also postulate the existence
of a popular or folk Islam, which continued the pre-Islamic beliefs of the Turks
beneath a sufi veneer. Accordingly, it was only in the sixteenth century when a rival
Shii dynasty, the Safavids, commanded the loyalty of certain discontented elements
among the Muslim population under Ottoman rule that this bifurcation attained a
new sectarian dimension and led to a series of confrontations between the Ottoman
state and some sufi circles.
Recently, however, some scholars have questioned key elements of this narrative by
showing that the essentialized, dichotomous categories of high vs. low, as well as
orthodox vs. heterodox Islam or sufism, do little justice to the complexity and fluidity of religious (and social and cultural) affiliations in late medieval and early modern
Ottoman Anatolia and the Balkans. They argue that the appeal of no sufi group or
movement was restricted to a single social, political or cultural milieu, and that the
social and religious profile of the adherents of any one group/movement/order could
vary from region to region and from period to period.5 They have further emphasized that the religious transformations of the late fifteenth and sixteenth centuries
entailed not only the Shiitization of some sectors of the Ottoman population and the
emergence of the Kzlba (modern Alevi) community, but also the Sunnitization of
the Ottoman ruling elites and a broad cross-section of Ottoman society. This essay
builds on these findings to present a more complex picture of the institutionalization
and confessionalization of sufism in the Ottoman central lands during the first three
centuries of Ottoman rule. It also strikes a middle ground between two co-existing
tendencies in current scholarship by seeing Ottoman confessionalization not as driven
by political expediency or by religious concerns alone,6 but rather as a complex process
with significant socio-cultural, political and religious dimensions.
The late fifteenth-century legend about a prophetic dream seen by Osman (d.
c.), the founder of the dynasty, represents Ottoman imperial success as the result
of a joining of forces between Osman and a sufi, the Vefai dervish Edebal.7 This
dream narrative is almost certainly apocryphal, but evidence from Ottoman land surveys confirms that Edebal was one of several Vefai dervishes already present in the
Ottoman principality in the first half of the fourteenth century.8 The hitherto almost
completely neglected history of the Vefais in Anatolia presents an apt point to begin
discussion, because it challenges so many of the assumptions of earlier historiography.
The Vefai order had originated not in Central Asia, the presumed homeland of the
Turks, but in neighbouring Iraq. Its founder, Ebul-Vefa (d. ), was a seyyid, a
descendant of the Prophet, who had spent most of his life among the Kurds in northern Iraq, and who had attracted disciples from many different ethnicities and social
backgrounds. By the late twelfth century Vefai dervishes were found in eastern and

D e r i n T e r z i o g l u

east-central Anatolia, and some had moved further west to the borderlands where the
early Ottomans were based after the suppression of the Babai revolt (). One
of the instigators of this Turcoman-cum-sufi-led messianic revolt against Seljuk rule
was a Vefai sheykh by the name of Baba lyas.
It is difficult to find a blanket term to categorize the religious and sectarian orientation of the Vefais from their eleventh-century origins down to their extinction as a
distinctive group sometime in the mid-sixteenth century. There is general scholarly
consensus that Ebul-Vefa must have been at least nominally a Sunni, although he and
his dervishes also indulged in practices that contemporary Sunni authorities found
objectionable, such as holding religious ceremonies where both men and women
were present. Nevertheless, four centuries later some Vefai adherents in eastern and
east-central Anatolia came under the Shiitizing influence of the Safavids and gradually became part of the nascent Kzlba community. Evidence for this has recently
been found in a rich body of archival material preserved by Alevi dede families in that
region.9 Those Vefais who headed further west towards the Byzantine frontier in the
late thirteenth century seem to have evolved in several different directions. Evidence
suggests that a good many Vefais in the early Ottoman principality became part of
the abdals of Rum, a loose association of itinerant dervishes of many different silsiles,
who were known for their strange and scant clothing, their consumption of illicit
substances such as hashish and wine, and their open disregard for social and religious
norms in general.10 Between the late fifteenth and seventeenth centuries many of these
abdals, in turn, became absorbed into the Bektashi order, which had a Shiitizing
tendency similar to the Kzlba. During the same period, however, other Vefais, or
at least their descendants, seem to have come closer to an emerging Ottoman orthodoxy along Sunni lines.11 We find the perspective of Vefais of the second kind in the
chronicle of Akpaazade (d. after ), who was also one of the sources of the dream
narrative mentioned above. The great-great-grandson of Baba lyas, Akpaazade was
deeply proud of his familys Vefai past, but he was also a sheykh of the Zeyni order,
which was known for its Sunni orientation and was popular at the time, especially
with the ulema.12 It must also have been Vefais of the second kind whom the Ottoman
scholar Takprzade (d. ) knew in person, and whom he had in mind when he
represented the Vefais in the early Ottoman state even more forcefully as shariahabiding or orthodox dervishes.13
Ultimately, the transformation of the Vefais in the central Ottoman lands cannot
be understood in isolation from the transformation of the early Ottoman religious
milieu generally. The latter process may have started earlier than presumed, but it
was still rather gradual and took some two centuries to be completed. It has been
argued that a defining feature of the early Ottoman religious milieu was metadoxy,
defined as a state of being beyond doxies, a combination of being doxy-naive and not
being doxy-minded.14 This point may be best illustrated by considering the religious
institutions then existing. The endowment deed for the earliest medrese to be built on
Ottoman soil, in Nicaea/Iznik, dates from , but the institution might not have
been operational until .15 When the Maghribi scholar Ibn Battuta passed through
Iznik and Bursa circa , he wrote no word on the medreses of those towns, but
dwelled rather more on the dervishes and especially the ahis, the quasi-mystical young
mens associations he found wherever he travelled in Anatolia. Particularly striking is
his description of a gathering he attended in the Ahi emseddin zaviye (convent) in

Sufis in the age of state-building

Bursa on the night of the Aure. Religiously normative practices such as Koran recitation, the performance of the canonical prayers (namaz) and preaching were mixed
with (from a Sunni point of view) more controversial practices such as sema (a ritual
that entailed the audition of music and poetry) and dance (raks) all in the presence
of the public and of religious and political leaders.16 Such multi-functional convents
were a characteristic and popular building type in the early Ottoman period and attest
to a social and cultural milieu where different religious and social groups mingled, and
where little need was felt to separate the ritual practices of the sufis and ahis from the
prescribed rituals for all Muslims.17 Religious fluidity might also have been facilitated
by the fact that the earliest ulema found in the Ottoman lands were often themselves
immersed in the sufi tradition. This was the case with the first scholar said to have
been appointed mderris in Iznik, Davud- Kayseri (d. ), one of the best-known
commentators on the works of the great Andalusian mystic Muhyiddin bn Arabi
(d. ).18
According to the early Ottoman chroniclers, writing in the late fifteenth century
but also drawing on earlier oral and written traditions, this seemingly idyllic relationship between the sufis and the ulema (or rather the proto-ulema, identified in these
early texts as fak from fakih, jurist) first began to show some signs of stress during
the reign of Bayezid I (). In this era the Ottomans made an early bid at
empire-building and at centralizing power at the expense of the former Anatolian beys
and the frontier warriors (gazis). Sufis were also affected by this development because
they had long-established links with the gazis as well as with other Anatolian beys, and
because they resented the emerging ulema, most of whom came from the Anatolian
interior and staffed the medreses which were then being built at an unprecedented
rate.19 To judge by the anecdotal evidence provided in the chronicles, sufi resentment
of the ulema stemmed partly from the roles the latter played as agents, allies and beneficiaries of a centralizing state and partly from a clash of values and mores between the
peoples of the marches and the hinterland.
Sufi discontent with the changing Ottoman regime first surfaced in a messianic
rebellion organized in the name of the sufi and scholar eyh Bedreddin in , fourteen years after the Ottoman defeat by Timur and three years after the reunification
of the Ottoman sultanate. While modern scholars dispute the religious and political
ideas of eyh Bedreddin, and even his role in the rebellion, there is more consensus
on the social, political and cultural profile of the rebels.20 It is highly relevant that the
rebellion broke out in the former territories of the Aydnoullar in western Anatolia
and later spilled over to the DeliormanDobruja region in the Balkans. In western
Anatolia nonconformist itinerant dervishes played a particularly active role as instigators of rebellion, mobilizing poor Christian peasants alongside Turcomans and recent
converts to Islam, while in the Balkans Bedreddin made common cause with discontented frontier warriors. Such a heterogeneous group presumably held heterogeneous
ideas about the nature of their endeavour; nevertheless, contemporary accounts suggest that, at least in western Anatolia, the rebels looked to Bedreddin as their saviour,
who would eliminate injustice, abolish all forms of private property and bring together
Muslim and non-Muslim alike under his messianic faith. It has been argued that the
Bedreddin affair represented the key moment of transition between the unbounded
order of multiple forms of worship to the austere world of institutionalized religion,21
but it is questionable how radical a break the event actually constituted in Ottoman

D e r i n T e r z i o g l u

religious culture. To begin with, the messiah figure behind the rebellion, eyh Bedreddin, was himself a man of multiple worlds and affiliations. While on the one hand,
as the son of a Muslim warrior (originally of the Karesi principality) and a Byzantine
noblewoman, he was a typical child of the frontier, he was also a highly respected
scholar who had studied jurisprudence and theology in Cairo before turning to the
sufi path. Although his master, Hseyin Ahlati (originating from or otherwise affiliated with Ahlat in eastern Anatolia), was considered highly controversial and even
suspected of heresy by the Sunni ulema of Cairo, this had not prevented Bedreddin
from being appointed kazasker (chief military judge) by the Ottoman prince Musa
upon his return to Anatolia, a telling sign of the cultural distance that separated the
Ottomans from fellow Muslims in the Islamic hinterlands. Perhaps even more telling
for Ottoman religious attitudes in this period is the fact that, for all the rather bold
ideas attributed to him and his followers, ultimately Bedreddin was executed as a rebel
and not as a heretic. However, this is not to deny the importance of the rebellion as
the first major event which alerted the Ottoman authorities to the dangers that sufis,
particularly the nonconformist itinerant dervishes, could pose to public order.
eyh Bedreddins revolt was actually part of a broader wave of messianism that
washed across the eastern lands of Islamdom (and beyond) in the late fourteenth and
the fifteenth century. The adherents of another messianic group, the Hurufiye, subjected the Koran to a radically esoteric interpretation and denied many fundamentals
of Islamic belief such as faith in the afterlife. Hurufis had originated in Iran in the
late fourteenth century but, after running into trouble with both the Timurids and
the Karakoyunlu rulers there, had directed their activities towards Anatolia in the
early fifteenth century. There they apparently found a receptive audience not only
among some sufis, but also in the person of Mehmed II during his first period of rule
(). By then, however, several of Mehmeds high-ranking administrators were
sufficiently alarmed by the religious ideas and/or the political threat posed by these
dervishes to persuade the young ruler to undertake a purge of Hurufis.22 Finally, as
part of the same broad current of messianic fermentation, the Safavids were transformed during the fifteenth century from a nominally Sunni sufi order into a militant
messianic movement with an extreme Shii ideology. Although Ottoman rulers faced
the full implications of this development only at the beginning of the sixteenth century, several sources suggest that they had begun to worry about the growing influence
of the Safavid sheykhs in Anatolia already by the mid-fifteenth century.23
It would be simplistic to assume, however, that the religious attitudes of the Ottoman ruling elites changed only in response to actual or perceived threats from messianic sufi circles. More broadly, the growth of Ottoman state power, the revival of
trade and other economic activity under their rule, and the growth of Ottoman towns
had created a much more attractive environment both for the ulema and for the representatives of an increasingly institutionalized, normative sufism. Some of these scholars and sufis flocked to Ottoman lands from the more established Islamic centres
of learning in Iran, Iraq, Egypt and Syria, while others were native-born sons, some
of whom had also been educated in these more distant centres. All these individuals
were instrumental in conveying the norms and values that were deep rooted among
the urban elites of the Islamic heartlands, and must also be taken into account when
considering the changing religious landscape of the fifteenth-century Ottoman

Sufis in the age of state-building

In modern scholarship, most of the new orders in Ottoman lands during this period
notably the Bayramis, Zeynis, Mevlevis, Halvetis, Kadiris and Nakibendis are
labelled Sunni. This designation is not so much wrong as in need of qualification.
First, in the absence of a neatly defined church to dictate right belief, the term Sunni
was necessarily more ambiguous and open to dispute than the term Orthodox in
Orthodox Christianity. Second, even the institutionalized sufi orders were considerably less institutionalized than, say, religious brotherhoods in medieval and early
modern Europe. Important differences of belief and practice could exist even among
the adherents of the same order.24 Third, just as fifteenth-century Ottoman society
was becoming socially and culturally more differentiated, considerable religious and
doctrinal fluidity still characterized sufi piety in this crucial transitional period, poised
between the relative metadoxy of the fourteenth century and the strident sectarianism of the sixteenth. Consider the case of Erefolu Rumi (d. ?), the founder
of the principal Ottoman branch of the Kadiri order and a self-identified Sunni. In his
Tarikatname, a manual written for fellow sufis and sufi novices, Erefolu explicitly
declares the Sunni faith to be superior to others, but he also writes positively of the
Alevis, a term he uses to mean sometimes the descendants and sometimes the followers of Ali.25 Further, he asserts that the perfect master (eyh-i kamil) must descend
from the House of Muhammad and Ali, that all must obey him (ulul-emrlik anun
hakkdur) and that those who oppose him may be considered Kharijis26 and legitimately killed.27 Rapprochement between sufism and Shiism manifested itself also in
certain beliefs and practices of the self-professedly Sunni sufis: in their silsiles, nearly
all of which went back to Ali, in their admiration for the descendants of the Prophet
in general and for the Twelve Imams in particular, as well as in their use of symbols
commonly associated with Shiism (twelve-cornered headgear, black turbans, etc.).28
While some of these features of late medieval sufism survived the confessionalization
of the sixteenth century, others did not, at least not without modification.
Likewise, the institutionalization of sufism must also be understood as a process
begun but by no means completed in the fifteenth century. The Bayrami order is
one example. Even though Bayrami sheykhs such as the Yazczade brothers Mehmed
and Ahmed Bican exercised a formative influence on Ottoman religious culture in
the fifteenth century, the early history of the Bayrami order is curiously shrouded in
mystery.29 There is little definite information about its founder, Hac Bayram (d.
): he left no writing of his own, while his followers provided divergent images of
him and his order. The earliest available records suggest that Hac Bayram lived with
numerous followers in a semi-rural setting outside Ankara, supporting themselves
through agriculture and sporadically by asking for alms like common beggars rather
than living off income from pious endowments like most representatives of institutionalized sufism. This indicates a group touched less by institutionalized sufism and
more by the teachings of ftvvet (the code of manliness and chivalry that guided the
ahi groups) and melamet (the path of blame, initially a separate path of mystical piety
that emphasized hiding ones spiritual state while outwardly living a life of anonymity
and sometimes even of seeming profanity). Also, Hac Bayram had been the disciple of
Somuncu Baba, whose silsile derived from the Safavid sheykhs and who had also been
a close associate of eyh Bedreddin. The question then arises: how was this seemingly
unwieldy group of dervishes transformed into an institutionalized sufi order? The
answer may lie in an early brush between Hac Bayram and the Ottoman authorities.

D e r i n T e r z i o g l u

Apparently, at some time during the reign of Murad II (), Hac Bayram was
called to the Ottoman court in Edirne to be questioned about rumours that he was
claiming worldly as well as spiritual dominion (saltanat is the word used). By all indications, however, the sufi master was able to allay these suspicions, and he returned to
his lodge in Ankara with many favours. This early encounter with the Ottomans may
have opened the way for various disciples of Hac Bayram to establish themselves in
the Ottoman Balkans and to gain adherents among the powerful. It also seems that
closer relations with the Ottoman authorities led to some fine-tuning and adjustments
by these dervishes. These adjustments and accommodations were probably at the heart
of the split evident in Bayrami ranks some time after the death of Hac Bayram. One
group, under the leadership of the learned Akemseddin (d. ), evolved into the
emsi branch of the order (generally considered the Bayrami order proper), developed
close links with Ottoman ruling elites and spread rapidly westwards. The other group,
led by an artisan, mer Dede the Cutler, resisted institutionalization and persisted
on the path of melamet, rejecting the distinctive paraphernalia and rituals as well as
the dependence on charity which characterized tarikat sufism. This second group of
Bayramis appeared in major Ottoman cities to the west only a hundred years later, in
the sixteenth century, where they faced persecution and hid their Melami affiliation
under different guises.
As the Bayrami example illustrates, the institutionalization and domestication of
sufi groups were closely related processes. However, the representatives of institutionalized sufism should not be considered simply as agents of a centralizing state. Rather,
when Zeynis and Bayramis and, slightly later, Halvetis and Nakbendis established
convents in major Ottoman cities such as Edirne, Bursa and stanbul, and enjoyed the
patronage of leading members of the ulema and the political elite, they continued to
exercise a considerable degree of autonomy. The Bayrami sheykh Akemseddin was,
on the one hand, the tutor and a confidante of Mehmed II and a supporter of his campaign to take Constantinople and, on the other hand, profoundly ambivalent about
the latters decision to adopt the infidels city as his capital.30 A similar ambivalence
about Mehmeds imperial project is evident in the chronicle of Akpaazade, who
was both a Zeyni and a descendant of prominent Vefai sheykhs.31 In both cases, the
ambivalence reflected not only the personal concerns of the sheykhs in question but
also the collective concerns of their social circles. The latter included in both cases the
frontier warriors, who, like other members of the political elite and contrary to what is
often assumed, were patrons not only of heterodox dervishes32 but also of seemingly
more Sunni-oriented sufis.33
In Ottoman historiography, the reign of Mehmed II is usually taken as the turning
point that finally transformed the Ottoman state from a frontier polity, in which the
sultan was only first among equals, into a centralized bureaucratic empire with an
absolutist ethos. In many ways, however, it fell to Mehmeds much less charismatic
son and successor Bayezid II () to sort out the institutional arrangements
that would sustain this newly forged empire. It was also during Bayezids reign that
the institutionalization of the sufis gained pace. As his epithet Veli (saint, or friend
of God) indicates, Bayezid was known to be pious, with a personal attraction to
sufism. His patronage of sufis nevertheless also had important policy dimensions.
In his early years, Bayezid was concerned especially to enlist the support of the
major power groups within the empire in his struggle to secure the throne against

Sufis in the age of state-building

a challenge from his brother Cem; influential sufi sheykhs provided him with
crucial contacts. Throughout his reign, Bayezid was also concerned to win over various
groups alienated by Mehmed IIs state-building and centralization, including the sufis.
This concern gained particular urgency after the Safavids under Shah Ismail replaced
the Akkoyunlus as the rulers of western Iran in .
Bayezid supported sheykhs from a wide variety of sufi orders.34 Interestingly, the
chief beneficiaries of his patronage were not the Nakbendis and Zeynis, who were
reputed to be the most Sunna-oriented of the sufi orders and were particularly favoured
by the ulema, but the Halvetis. Bayezids close relations with Halveti sheykhs went
back to his days as a princely governor in Amasya, when the latter had supported him
in his struggle for the throne, but their attraction for Bayezid probably went beyond
this early alliance. Halvetis, like Bayramis, had issued from the same silsile as the Safavids and exhibited strong forms of Alid loyalty even while they remained within the
orbit of Sunni Islam.35 This must have made an alliance with them particularly valuable for Bayezid (and later Ottoman rulers) in his attempt to reach out to social and
religious groups potentially receptive to Safavid propaganda.
It may have been with similar considerations that Bayezid viewed the up-andcoming Bektashis. Historians have long puzzled about the considerable overlap
between the religious beliefs and practices of the Kzlba communities, who were
persecuted as heretics, and the Bektashi dervishes, who were viewed with suspicion
but still largely tolerated by the Ottoman authorities. It has been argued that Bektashi
dervishes were initially sent by the Ottoman authorities as missionaries to properly
Islamize the Turcoman tribal population in Central Anatolia but that in the process
they had themselves been acculturated into the heterodox, syncretistic beliefs of that
population.36 This argument, however, seems problematic for at least two reasons.
First, because it (at least initially) attributes an overriding agency to the state when in
fact it was only one of several different actors involved and when the mores and predilections of the ruling elites themselves were quite heterogeneous. And, second, because
it overlooks the diversity of both the Bektashis and the Kzlba in terms of geographic
and ethnic provenance. Others, following Kprl,37 have explained the doctrinal
and ritual similarities between the Kzlba and the Bektashis with reference to their
common religious affiliations, particularly their connections with the abdals of Rum.
It is now generally accepted that the cult of Hac Bekta, an enigmatic Vefai dervish
of the thirteenth century, had initially been the patrimony of the abdals. Accordingly,
the Bektashis proper had emerged out of the ranks of these itinerant dervishes in the
late fifteenth century, and continued to absorb many of the remaining abdals into
their ranks during the sixteenth and seventeenth centuries. Other abdals, however,
seem to have resisted Bektashization and instead became part of the nascent Kzlba
The reign of Bayezid II was a key period for the early institutionalization of the Bektashi order. The writing down of the hitherto oral traditions about Hac Bekta, the
renovation of the lodge of Hac Bekta,39 and the reorganization of the administrative
structure of the order by the sheykh Balm Sultan all took place around this time. Even
though Bayezids role in each of these developments remains unclear, his conciliatory
policies may have laid the foundation for the subsequent accommodation of various
nonconformist sufis under the Bektashi umbrella. However, the growing prominence
of the Bektashis was owing also to their links with other socio-political actors. Among

D e r i n T e r z i o g l u

Bektashi patrons were the old gazi families of the Balkans, whose powers had been
considerably reduced by Mehmed II but who still continued to serve the Ottomans as
provincial governors.40 In addition, the Bektashis had established a foothold among
the Janissaries, the elite military corps whose ranks were filled with devirme recruits
from among the Christian population in the Balkans. It has been speculated that the
Ottoman rulers had encouraged the spread of the Bektashi dervishes among the Janissaries because the syncretistic beliefs and practices of the former were found to be a
particularly effective way to Islamize these Christian-born soldiers.41 If indeed the state
elites had played a role in this process, they may also have aimed thereby to draw the
nonconformist dervish milieus closer to the imperial centre at a time of growing tension between the two. It is also possible, however, that the Bektashis (or, before them,
the abdals) had gained a foothold among the Janissaries independently of any state
involvement. The fact that some Janissaries turned Kzlba in the sixteenth century
or that the tie between Janissaries and Bektashis survived efforts by the central state to
replace it with another the Mevlevi should indicate that state control over this elite
military corps was not as absolute as has been generally assumed.42
Contrary to the prevalent image of Bayezid as a softie, archival records indicate
that Ottoman surveillance of and punitive actions towards suspected supporters of the
Safavids had already begun under his rule, around the time Shah smail had set himself
up as an independent sovereign in .43 Still, Ottoman religious policies did harden
considerably following the pro-Safavid ah Kulu rebellion, which erupted among
the Turcoman tribes of the Taurus mountains in and which was joined by a
wide variety of disgruntled groups, from timariots to ahis. Within a year, Bayezid was
deposed by his hawkish son Selim I (), who adopted a much more aggressive
policy vis--vis both the Safavids and Ottoman subjects suspected of being Kzlba.
Modern historians have rightly emphasized the pivotal role played by the OttomanSafavid conflict in transforming the religious landscape of both empires during
the early modern period. Some changes were quite dramatic, like the persecution of
thousands44 of suspected Kzlba supporters and the destruction of scores of dervish
lodges and convents that harboured them during Selims march through Anatolia,
first to suppress the ah Kulu rebellion and later to confront the Safavids ().
Similar waves of persecution, though smaller in scale, were repeated during later proSafavid rebellions and OttomanSafavid campaigns.45 At the same time, however, the
Ottoman state did not and could not afford to forget its old policies of containment
and accommodation. Even in times of war, the authorities often dealt with Kzlba
communities in a differentiated manner, executing individuals identified as agents
(halife) of the Safavids while subjecting others to surveillance, banishment and a variety of other, lesser, forms of punishment.46 Moreover, persecution or threat thereof
was only one component of the Ottoman policies of Sunnitization. The state also
sought to reshape the behaviour of its subjects by distributing posts and benefits and
by conferring or ratifying claims to privileged status. Recently, Canbakal, analysing
the Ottoman registers of seyyids, or descendants of the Prophet, has suggested a link
between the states attempts to regulate the appropriation of this status and its policies
of containment directed at the Kzlba, for whom descent from the Prophet was of
particular significance.47
In addition to these carrot-and-stick policies directed at specific religious communities, the state pursued more positive measures which required time to take effect,

Sufis in the age of state-building

like the campaign to build neighbourhood and Friday mosques wherever Muslims
lived in Anatolia and the Balkans. This policy was closely linked to attempts to enforce
communal performance of the five daily canonical prayers by the regions Muslim
population. Strikingly, the mosques built in this period were no longer conceived as
multi-functional spaces that could accommodate sufis, ahis and travellers alongside
the congregations for the Friday service.48 Rather, guest rooms were now separated
from the mosques and built as free-standing dependencies. Along with monasteries
and churches, many of the remaining multi-functional convents were also converted
into Friday mosques, while some of the dervish lodges affiliated with heterodox sufi
groups were converted into medreses.49
Significantly, such changes corresponded with the onset of a more discriminatory
attitude among the Ottoman elites towards sufism. This appears most clearly in religio-legal literature, which contains, for the first time, a sustained critique of a wide
variety of sufi beliefs and practices as bidats, inadmissible innovations introduced
into Islamic tradition after the lifetime of the Prophet. Recent scholarship has tended
to downplay sixteenth-century attacks on sufi bidats either as the work of a tiny minority among the Ottoman ulema or as campaigns directed specifically at heterodox and
especially Shiitizing sufis such as the abdals. In fact, however, not only hardliners
like ivizade Muhyiddin Mehmed (d. ) but also moderates like Kemalpaazade
(d. ) issued fetvas declaring the use of ritual music and dance during sufi gatherings illegitimate.50 Notwithstanding the nuances between their positions, it proved
exceedingly difficult to draw the boundary between the sharia-abiding and deviant dervishes on this issue. It is true that the central government primarily targeted
Shiitizing itinerant dervishes (k), instructing local authorities to prevent them from
playing musical instruments and dancing in public.51 Apparently, however, some local
kadis drew a different conclusion from the legal debates and wanted to interfere also
with the Halveti devran (zikr ceremonies in which dervishes moved rhythmically in a
Attempts to negotiate between these groups appear in the fetvas of Ebussuud
(d. ). The son of a Bayrami sheykh, Ebussuud is widely regarded as the architect
of a new consensus between the Ottoman ruling elites and sufis during his long tenure
as eyhlislam (). We lack detailed studies of the many compilations of Ebussuuds fetvas that might clarify the development of his thought on the subject. Nevertheless, when viewed as a whole, his fetvas indicate that even this known compromiser
was unable openly to defend the sufi devran and dance (raks) as had Ali Cemali
(d. ), another eyhlislam from a prominent sufi family, half a century earlier.
Rather, Ebussuud entered into hairsplitting discussions about how it is better to
practise zikr sitting down rather than standing up, and, if standing up, then at least
not to move ones waist and head, and, if moving ones waist and head, then at least
not to move the feet. Only when confronted with a question about the permissibility
of calling the Halveti sheykhs and disciples as well as all who befriended them infidels because they practised devran, did Ebussuud respond by saying that there were
valuable people among them, and that it was wrong to slander people and make false
attributions about matters of which one has little knowledge.53 As much as he may
have strived to protect valuable people among the Halvetis and their sympathizers,
his fetvas, like those of other sixteenth-century mftis, point to a religious milieu in
which such sufi groups as Halvetis, Mevlevis, Kadiris and Bayramis were also coming

D e r i n T e r z i o g l u

under pressure, if not to give up their controversial practices then at least to modify
them, and certainly to practise them more discreetly, away from the public eye.
Much the same can be said about the controversies surrounding the teachings of
Muhyiddin Ibn Arabi, whose posthumous title Sheykh Akbar (Greatest Master)
nicely captures his high standing among the sufis. While hardliners like ivizade were
no doubt in the minority among the Ottoman ulema for their refutation of this great
sufi, it is striking that even defendants of Ibn Arabi such as Kemalpaazade cautioned
that his highly complicated texts were not fit for the eyes and ears of the common
folk, who could easily misunderstand them and be led astray into heresy.54 Hence it
was perfectly possible for the religious and political authorities to criticize some sufis
for expounding on the teachings of Ibn Arabi in improper ways and contexts, while
protecting and patronizing others from a wide variety of orders who did the same more
discreetly. This is also what got some Bayrami-Melami and [Halveti]-Gleni sheykhs
into trouble and led to their persecution as heretics during the sixteenth century, even
as other sufis from the same orders attracted adherents among the ruling elites.
It is important to emphasize, nevertheless, that sufis were not only at the receiving end of Ottoman confessionalization policies. They, too, shaped the Ottoman
religious landscape in various ways, whether by rebelling, dissimulating, negotiating
and compromising or by acting themselves as agents of Sunnitization, a role which
became more important in the later sixteenth century. It was not uncommon for sufi
masters to acquire a solid grounding in the religious sciences even in the fifteenth
century, while dervish lodges had long been major sites of literacy and book learning
alongside the medreses.55 Zeyni, Nakibendi and some Halveti sheykhs had enjoyed
particularly close relations and partly overlapped with ulema circles during the late
fifteenth and sixteenth centuries. During the second half of the sixteenth century,
these existing relations were intensified when, on the one hand, increasing numbers
of ulema frustrated by the congestion in the learned establishment left their careers as
judges or mderrises for the sufi cloak56 and, on the other hand, a more pronounced
Sunna consciousness emerged among the representatives of institutionalized sufism. It
is also in this period that the state began to appoint sufi sheykhs, most notably those
equipped with religious learning and presumably the appropriate Sunni consciousness, as preachers. Many served as preachers in the neighbourhood mosques adjacent
to their dervish lodges, while the privileged few preached in the large imperial mosques
in stanbul, Edirne and Bursa. Sufi sheykhs who occupied these influential posts used
them not only to preach their brand of Sunni piety tempered with sufism but also to
denounce others who deviated from it, whether the Kzlba or other heretics, like
the drisis and Hamzevis (derogatory terms used for Bayrami-Melamis).57 Continuing trends which had originated under Bayezid II, Halvetis once again dominated the
most influential of these preacherships during the late sixteenth and early seventeenth
Did this rapprochement between the self-professedly Sunni sufis and the Ottoman
ruling elites then signify an end to the -year-long process of institutionalization
and confessionalization of sufi groups in the Ottoman empire? This is usually the
impression given by the existing literature, which has defined classic Ottoman religious culture as a welding together of Sunni (Hanefi) legalism and sufism. However,
no sooner was this synthesis achieved than it was also challenged. During the seventeenth century, what constituted the Sunna, the normative example of the Prophet,

Sufis in the age of state-building

and who could be considered as part of the Sunni community continued to be hotly
debated. Now, however, the main protagonists in this debate were no longer the ulema
elite, but preachers, represented by the militantly Sunni puritanical preachers known
as the Kadzadelis, on the one hand, and sufi (especially Halveti) preachers, on the
other. In the secondary literature, the Kadzadeli challenge to institutionalized sufism
has generally been treated as a temporary aberration from the norm, explicable only in
the context of the crisis of the central state and its elites in the seventeenth century.59
Historians have tended to overlook the extent to which this movement was also a
sequel to, or an episode in, a long history of Ottoman Sunnitization. Throughout the
seventeenth century both the Kadzadelis and their adversaries continually made references to sixteenth-century debates on bidats in order to bolster their position.60 Both
sides also played out their differences before a broad urban public, whose ways and
habits had been significantly moulded by some years of Sunnitization. Significantly, the seventeenth-century controversies testified not only to the tremendous success of that process, but also to the continuing ambiguities and counter-trends among
the broader lay public. Not all who were disenchanted with institutionalized sufism
turned Kadzadeli; others embraced an increasingly Sunnified Melamism, while representatives of institutionalized orders such as the Halvetis, Celvetis, Nakbendis and
Mevlevis also continued to attract adherents among the urban public.61
Finally, it should be added that neither the institutionalization nor the confessionalization of sufism were processes that were concluded in either the sixteenth or the
seventeenth century. Early nineteenth-century attempts to recentralize the empire
brought about important sequels to both developments. Also, while this essay has
concerned itself primarily with institutionalized sufism, there was at all times more
to sufism than a set of more or less formalized sufi groups and whatever they believed
and practised. In fact, the spread of the sufi orders in Ottoman cities during the late
fifteenth and sixteenth centuries also helped open up new channels for sufi beliefs and
practices and sufi-tinged attitudes and sentiments to be disseminated across the social
board and to inform a wide variety of cultural practices from the shadow theatre to
poetry. It awaits future studies to take account of these less authorized but no less
important and creative uses of sufism.


The author gratefully acknowledges having been a research fellow at the Research Center for
Anatolian Civilizations during the writing of this article. She also wishes to thank Cemal Kafadar and Ayfer Karakaya-Stump for reading over the text and making a number of valuable
Karamustafa .
Trimingham ; Meier .
Karamustafa .
For a more extensive discussion of the terms relevance in the early modern Ottoman context,
see Krstic : and : .
Karamustafa : ; Le Gall : .
Generally speaking, Ottomanists proper tend to privilege explanations based on political expediency and reasons of state, while specialists on Ottoman sufism and particularly the Turkish
scholars of sufism, trained in schools of theology rather than in history departments, tend to
give primacy instead to religious doctrine.

D e r i n T e r z i o g l u

Askpasazade : ; cf. Anon. : , where the dream is ascribed to Osmans father,

Kafadar : , .
Karakaya-Stump : .
The first scholar to note the link between the Vefais and abdals of Rum was Glpnarl :
, . Later, the connection was made more categorically by Ocak : ; cf.
Ocak : , esp. .
: .
Inalck ; ngren : .
Taskprzade. : , ; for similar assessments of the early Ottoman Vefais by modern
scholars, see Inalck : ; Karamustafa : .
Kafadar : .
Bayrakdar : .
Ibn Battuta : , esp. .
Eyice ; Wolper : esp. , , .
Askpasazade : , .
Kafadar : ; Darling : ; Ihsanoglu, : , .
Balivet ; Ocak : ; there are some alternative views in Yksel : .
Barkey : .
Babinger : , based on Taskprzade (see edn: ).
Yldrm : , , .
Le Gall ; Glpnarl .
Esrefoglu Rumi : , .
Literally, those who go out an Islamic sect whose origins lay in a group of Alis supporters
who deserted their leader and eventually formed a distinctive sect considered beyond the pale
of Islam by both Sunnis and Shiis.
Esrefoglu Rumi : .
See Karamustafa : x, on the relationship between Sufism and Shiism as a thorny question among scholars. On Alid elements in Mevlevi beliefs and practices, see Glpnarl :
Curiously, much less has been published on the Bayrami order proper than on its Melami offshoot. On the former, see Kissling ; Bayramoglu ; Bayramoglu and Azamat ; and
Yurd and Kaaln . While Glpnarl remains a standard reference, for more recent
views see Ernsal and . Cf. also Imber ; Ocak : ; and Clayer et al.
and Kaaln : , .
Inalck ; Kafadar : .
Kiprovska ; Yrekli : .
Umur : .
Kafesioglu : , ; ngren : .
Kissling .
Mlikoff .
Kprl : .
Karakaya-Stump : .
Yrekli .
Ibid.: .
On the related argument about the role of Bektashi dervishes as agents of Islamization in the
Ottoman Balkans, see Kprl [] and Barkan ; there are more recent views
in Krstic : .
Yrekli : .
Sahin and Emecen : xxivxxv.
According to some Ottoman chroniclers, , people were killed during the years ,
though this estimate appears exaggerated in view of much more modest numbers cited in the

Sufis in the age of state-building

Ottoman mhimme registers (Savas : ). However, mhimme registers record only

those dealt with by administrative-judicial proceedings; numbers of those killed in battle are
Altnay ; Imber ; Savas ; Sener .
Savas : .
Canbakal and .
Allocation of the mosque exclusively to normative religious practices, such as the canonical
prayers and preaching, and relocation of all other functions to separate dependencies had
started earlier, with the imperial mosque built by Mehmed II, but his example was not followed
by either Bayezid II or lesser patrons of architecture: Kafesioglu : , , .
Necipoglu : .
ngren : ; cf. Terzioglu : .
Altnay : ; Savas : , .
ngren : .
Dzdag : .
ngren : .
Faroqhi a: .
ngren : ; Niyazioglu : .
Clayer : .
Zilfi : ; Zilfi : .
Zilfi , ; Baer .
Terzioglu : , .
Terzioglu : .





Metin Kunt


ouseholds were the building blocks of the Ottoman political edifice. Therefore
the basic unit for the study of the sociology of political life is the household. In
a dynastic empire the role of the ruler is rightly accorded central place, and the royal
household is the keystone of the edifice. While paying due attention to this imposing
keystone, we must not neglect humbler bricks in the walls: the sultan maintained a
huge personal staff and retinue, but he also insisted that all his officers did the same,
proportionate to their rank and allotted revenues.1
The Ottoman military-administrative system was founded upon the principle that
an officers income would support a retinue appropriate to his office. Income, in this
context, meant tax revenues, fines and fees generated in a locality and assigned to an
officer. The Ottoman term for such revenue grants was dirlik literally, living. The
holder of a revenue grant was expected to reside on location and to collect his share of
agrarian produce as well as ground rent from the peasants. In towns there were market
dues on trade and crafts production in addition to small-scale agriculture, orchards
and vegetable patches. Larger cities on international trade routes, such as Salonica,
Bursa, Aleppo and Cairo, had considerable customs revenues.
At the lowest end of the dirlik allotment pyramid was the cavalryman (sipahi), with
a basic revenue grant, commonly a village. There he was to live; at harvest time he
received his tithe in kind, and he kept the peace in his village, collecting fees and fines
for transgressions. When called to military duty, he presented himself to the district
commander, ready for campaign with his horse, arms and equipment, all paid for out
of his revenue grant. With some seniority and good service on campaign, he might
achieve a rise in his basic revenue from an additional source; if his income increase
was as much as double the basic grant, he was required to bring with him on campaign an extra fighting man, whom he would also maintain out of his revenue grant.
The revenues of a town, as opposed to a village, would be assigned to a commander,
both because his income should be higher and because the administration of a town
required a more experienced officer; it was rare that town revenues were divided up
into smaller livings. Some of the commanders of a district also had duties beyond

Metin Kunt

their town, such as overseeing the proper presentation of all revenue grant-holders
in the district while on campaign, ensuring that they were well equipped and with
the correct number of extra armed men. Commanders might have incomes ten times
more than the basic revenue grant and therefore have personal retinues of ten men or
so. Each district had a governor-commander in the chief city, with revenues at least
a hundred times the basic grant; a district governor might thus be required to have a
personal military household of to men on the basis of his official revenues. A
district governor who made more out of booty would have a retinue much larger than
the expected number. All in all, a district might have to basic revenue grants
and ten to twenty town commanders. Therefore, when the district cavalry force was
mobilized for campaign it might field , men or more, most of them not revenue
grant-holders themselves but serving in a commanders personal retinue. Several districts, perhaps a dozen, made up a province, about twenty in number throughout the
realm in the mid-sixteenth century. The governor-general, who held the chief city and
district of the province, might have revenues five times that of a district governor and
his own personal army also approaching , men.
All these numbers are deliberately vague, since there was considerable variation both
in geography, from district to district, and in history, from age to age; they merely
provide a sense of overall magnitude and proportion. Revenues were estimated, based
on past produce and on administrative regulations for each district specifying local
tithing and ground-rent rates, as well as enumerating other taxes and fees. They were
expressed in akes, the small Ottoman silver coin called asper in European writings.
While the basic revenue grant was a few thousand akes a year, a governor-general was
allocated a million or so. When moving from the provinces to the centre, as a governor-general might do when elevated to the imperial council as a vezir, in the sixteenth
century the seven vezirs were each allocated around million akes; each therefore was
required by regulation to have a household of several thousand men. At the apex of
this system of livings and retinues was the sultan, with revenues (which he allocated
to himself) of million akes or more and, in the early sixteenth century, a royal
household of about ,.
We may make some general observations based on this sketchy outline. The Ottoman revenue grant system developed in circumstances similar to feudalism in Europe,
when a limited amount of precious metals allowed only a limited supply of coin. The
difficulty of transporting large amounts of grain and other agricultural produce, added
to a limited market, necessitated that revenues in kind should be collected on the spot
and consumed locally. But Ottoman livings and retinues were fundamentally different
from European feudalism: the grants were not hereditary, nor were they of fixed location. An Ottoman officer moved about the realm holding different offices in different
districts; his heirs, sons and daughters, inherited his private estate but not his official
status. However, in the fathers lifetime, sons themselves may have advanced in the
system. It was therefore allowed that, at least in the case of higher revenue grants, one
or two sons may be provided with a small grant at the death of the father, especially
if he died in battle.2 As the rank of the officer advanced and the size of his revenue
increased, so did the proportion of revenues in cash, because his living included towns
with markets and cities with customs revenues. Increases in precious metals, in coin
in circulation and in trade, both regional and international, allowed a greater amount
in cash to be allocated to higher revenue grants, especially to the revenues reserved for

Royal and other households

the sultan. Greater revenues meant larger households; larger households meant greater
influence and political power; concentration of the largest revenues and households
in the centre, in vezirial establishments as well as the imperial palace, meant greater
centralization in the polity.
In the first half of the fifteenth century the definitive conquest of Smyrna and
Salonica, the main outlets for Anatolia and the Balkans respectively, and the capture
of Serbian silver mines helped Murad II () strengthen his position as ruler,
richer and more powerful than his frontier warlords. The conquest of Constantinople
in gave Mehmed II enduring fame as the conqueror par excellence throughout
the Islamic world as well as in early modern Europe; it also allowed him to garner
much greater revenues personally. He then captured the copper mines of north Anatolia by incorporating the Candarolu principality of Kastamonu. His grand plan also
involved the elimination of the Genoese from the Black Sea, taking over trading posts
not only on the Anatolian coast but also in Crimea, by the capture of Caffa, the main
port for export of the grains, furs and dried fish of Eastern Europe.3 Selim Is capture
of Syria and Egypt in laid the foundations of the magnificence of his son
and successor, Sleyman. The sultan received a major share of the trade revenues of
Aleppo and Damascus, and from Egypt millions of gold ducats directly into the imperial coffers.4 Their treasury thus enhanced, the sultans were able to invest in a brand
new navy and also increase their households.


Ottoman households were originally military establishments, made up of troops loyal
to their master. In the Muslim tradition of the rulers household, the original impetus
may have been to create a troop distinct from the dominant ethnic element, tribal
Arabs, who had supplied the main fighting force in early Islamic expansion.5 By this
means the caliph raised his political power above tribal rivalries in an age when Islamic
society was also facing the challenge of becoming truly universal, no longer identified
by Arabness. Caliphs recruited captives from the Eurasian steppe, from among peoples
whose lifestyle on horseback made them amenable to training as a personal guard.
During the several centuries before the emergence of the House of Osman, these
mamluk (literally, owned) guards served their masters, became involved in dynastic
politics, founded their own rule, and even established states without dynastic rule but
with acclaimed sultans.6 A second distant tradition came from the steppe heritage
itself. There young men displaced from their own tribes were adopted by the leader
of another tribe as his own men, termed nker. The eponymous founder of the Ottoman state, Osman Bey (d. ), himself had various men in his modest household,
some as kul servitors (the Turkish for mamluk) and some as nker. The Ottoman beys
household grew during the fourteenth century by absorbing more displaced young
men from other lands and more owned servitors.7
An important development came in the second half of the century, when the beys
advisors began to claim a fifth share of the spoils gathered by frontier lords in the Balkans and sold in Anatolia, either as a quantity of coin or as a number of men. Scions of
Byzantine aristocracy in captured towns were also incorporated into the Ottoman rulers palace. Around , an Ottoman innovation in recruitment for the palace was
devirme (literally, gathering) that is, the forcible removal of adolescent village boys

Metin Kunt

for training and education. This peculiar Ottoman practice may have first come about
in the frontier zones, where frontier lords could not always distinguish or did not
choose to do so between what had already been incorporated into Ottoman lands
and what was outside and still open to raiding; otherwise it is very difficult to explain
the devirme phenomenon.8 Islamic tradition, which the Ottomans elsewhere followed
scrupulously, assigned a specific, protected, if second-class, status to non-Muslims
submitting to Muslim rule as zimmis: they paid a head tax (cizye) per adult male, and
in return could practise their religion relatively freely, albeit with certain restrictions
and impositions. Taking children away from law-abiding zimmis was never a practice
condoned or even contemplated before the Ottomans. This is in contrast to rebellious
non-Muslims or those conquered by force: in those cases, the enslavement of persons
was considered as legitimate as the plundering of property. The ambiguity and fluidity
of a frontier zone may have provided the initial impulse and instances of devirme.
Even if this peculiar practice was started by frontier lords, it soon became a royal
prerogative. Other households sometimes listed zimmi volunteers among their members. Was this a euphemism to disguise what might be termed private devirme?
Probably not, since such lists also specified other sources for the masters retinue.9
Contemplation of non-Muslims volunteering for service in a household at the cost
of conversion to Islam leads to the bigger issue of how devirme was perceived by the
zimmi population in general. Even after the practice was regularized, with certain rules
to soften its impact on the non-Muslim peasantry, there can be no doubt that most
families felt that they had lost a son not just to an unknown fate but to total alienation.
Families hid their eligible sons when a palace official approached their village. His purpose was to choose one in forty boys from the locality, excluding those already married
and only sons, and not taking more than one child from each family. At least, such
were the established rules. However, it is also true that Bosnian Muslims, conversely,
saw devirme as a privilege that they wanted to share. The palace official on devirme
duty was ordered to ensure both that all eligible non-Muslim boys were presented for
inspection and that only Bosnian Muslims who were granted the right to devirme
(and no other Muslim boys) should be taken for palace service. Bosnian Muslims
were specified as those who had been living there at the time of the Ottoman conquest
in the mid-fifteenth century but not later Muslim settlers. The recruiting (collecting)
officer had to ensure that, with the exception of true Bosnians, only non-circumcised
boys were drafted.10
Voluntary or not, incorporation through conversion into any great household, not
just entering the sultans service, was a sure way of changing status from that of subject
to one of the ruling stratum. What, to many, may have seemed the loss of a son was
evidently for some a great future for the child. While joining a grandee household did
not necessarily guarantee great fortune, it certainly offered the possibility of one for
those whose brains matched their ambitions. Newly recruited servitors became Muslim and gained a new identity, served their masters, and rose in political rank under
their aegis. Those who served an illustrious master, sultan or pasha over a long period
remained associated with him even after leaving his household and gaining independent office. As the sultans sisters and daughters were married to vezirs and pashas
trained in the royal household, so too the loyal servitors of a grandee were further
rewarded by being married into his family. An Ottoman House was thus joined at
the top with its household.

Royal and other households

Boys drafted into royal service or joining a grandee household were separated by ability and inclination. In the case of royal recruits, this selection was done in several
stages: in the first instance, brawny but not necessarily brainy boys were apprenticed
(as acemiolan) to the janissary household infantry; the more promising were then
sent to various royal establishments for education and training. For these selections,
officers used physiognomy (ilm-i kiyafe), the knowledge of inner qualities apparent in
the visage. The initial stage in creating a new person, after circumcision and formal
conversion, was the naming: a Muslim should not carry a non-Muslim name or patronymic. They all became sons of Abdullah Abdullah being a common first name
but also having the meaning of slave/creature of Allah, and by extension any person or A. N. Other. For the personal name, as well as the common Muslim names
(Mehmed, Ahmed, Ali, etc.), newly converted boys were often given the Muslim form
of biblical names (Davud for David, Yahya for John, Isa for Jesus, Sleyman for Solomon, etc.) or heroic names from the great Persian epic ahname, the Book of Kings
(particularly Behram, Rstem, and Iskender for Alexander). Abdullah could never
be the first name of a convert, devirme or otherwise, although it was often used as a
given name in Muslim families; thus not all sons of Abdullah should automatically
be considered converts.11 A boy who might have been entered in the devirme register
as Dimitri, son of Petro, Albanian, blond, round-faced, tall, mole on right cheek
now became Hasan, son of Abdullah, but with all the other attributes recorded in
the janissary register. If a Halil used to be Vassilis, was the rhyming coincidental?
Was a new Hizir, the green man of the Koran (who found the elixir of life) and was
conflated with St George in Muslim lore, originally a Giorgos or a Gyrgy? Did the
naming officer have a sense of humour, or a sense of history, or an understanding of
trans-cultural connections?
The most select of the sons of Abdullah came to the central royal establishment,
the Topkap Palace. There they were entrusted to the white eunuchs of the imperial
harem, the inner sanctum. The mens section of the harem was more commonly called
enderun, the inner service, to distinguish it from birun, the outer service of the palace,
where special guard units as well as royal troops were attached. In the enderun the boys,
now palace pages, were at first without any special duties but received basic education under the supervision of the white eunuchs. Later they were divided into several
chambers named for the services they performed for the enderun itself and for attending their master, the sultan. Pantry Chamber boys received the meals cooked in the
great palace kitchens in the birun and served them to the eunuch tutors and classmates;
the Campaign Chamber boys were in charge of laundry duty, especially on imperial
campaigns; the Treasury Chamber boys, older and wiser, cared for the valuables in
the imperial collections precious objects, books, luxury cloth and special garments.
While they performed these services their education and training continued, under
outside tutors as well as the white eunuchs, and including not only book-learning but
also palace etiquette. The elite of the palace pages, those who stayed in the enderun
longest and were therefore the best educated and trained, were admitted to the Privy
Chamber, charged with the care of holy relics of the Prophet Muhammad and his
companions brought from Cairo in by Selim I, and in general attending to the
wishes and pleasures of the sultan.

Metin Kunt

All boys acquired the rudiments of an education; with the exception of a few who
had little interest in books, most also pursued studies in languages, literature and history. Ottoman humanities was based on Persian classics. The ahname, the source of
names for some of them, remained the most valued reading, with its tales of heroism
as well as wise kingship, of good rule based on justice. Persian verses by Sadi and Hafiz,
the revered poets of Shiraz, were the exemplars Ottoman gentlemen had to learn well:
not only Ottoman sultans but also many officials dabbled in poetry themselves, a few
with commendable results. Penmanship and calligraphy were encouraged as much as
archery and horsemanship. As at any good school, most of the boys balanced booklearning and field pursuits, some gaining a reputation for knowledge of the arts, while
others excelled in sports.


The late sixteenth-century writer Mustafa Ali, deliberating on what made the Ottomans great, considered the blending of races to create a new Osmanl man to be the
secret of Ottoman success.12 Although commoners in Ottoman society, in Anatolia
and Rumeli, also experienced a certain degree of blending through intermarriage, voluntary or otherwise, what the writer had in mind was the Ottoman ruling elite. The
household system was the cradle of this new elite. Its language was Turkish, but the
elite were not, and did not refer to themselves as Turkish. The Turkish ethnic element in the realm was only one among many other ethnicities, and at least until the
seventeenth century Turks were deliberately excluded from the royal household. It
was cause for scandal that, when Selim II ascended the throne in , he enrolled
thousands of Turks from his inordinately large princely army into the imperial Janissary corps. Servitors of the royal household were meant to be a blend, certainly not
The educated Osmanl was a Turkish-speaker and a Muslim and therefore a Turk
in European eyes, but his written Turkish for literature and history was a cultural
blend and considerably removed from ordinary spoken Turkish. Persian and Arabic
phrases and expressions, even grammatical constructs, infiltrated written Ottoman
Turkish, bureaucratic and literary, more so than Latin and French expressions did
English. By around , however, not only had literature completely shed the Persian shadow but even religious treatises came to be written in Turkish rather than in
Arabic, the language of the Koran and therefore traditionally of religious scholarship.
As in Europe, where the title of a book may be Latin while the content is in the vernacular, so Ottoman books often carried Arabic titles and Persian sub-headings but the
text would be in Turkish, albeit a flowery one distinct from the language of the Turks
among the subjects. The Ottoman elite may have seemed Turkish to the outsider, but
they had a distinct new identity.
Even the Islam of these new Ottomans could be considered to contain some idiosyncratic elements. Indeed, Turkish settlement in the land of Rum (i.e., Anatolia as
the Byzantine Rome) from the eleventh century had given rise to syncretic impulses
both in popular Turkish Islam and in courtly culture, expressed mainly through mystic movements coalescing into brotherhoods. The urban culture of thirteenth-century
Konya, under Turkish Seljuk rule, was the original setting of Mevlana Jelaleddin-i
Rumi, whose descendants and disciples developed his approach to universal godhead

Royal and other households

through artistic expression, in poetry and music, in the Mevlevi brotherhood (or way
in Islamic terminology). Meanwhile, social and political protest movements in rural
areas of Seljuk Anatolia had a leadership with a religious tinge and were influenced by
pre-Islamic inner Asian practices. By the early sixteenth century, such movements and
their dervishes came together under the rubric of Bektashi. These two main brotherhoods, Mevlevi and Bektashi, and others, became active in South-East Europe under
the Ottoman aegis and facilitated conversion in both rural and urban Balkan society. But in addition to the conversion of Anatolian and European Christians to Islam,
in Anatolia there was the phenomenon of many Turkish speakers as members of the
Armenian and Greek churches, either as Turks converting to these churches or as some
Greeks and Armenians converting to Turkish speech. All these various types of conversion, religious and linguistic, as well as intermarriages across ethnicities, helped to create
an Ottoman society in its Byzantine-Roman core areas which developed a character
distinct from older Muslim areas of West Asia.
What was true of Ottoman Rumi/Roman society as a whole was also true for the
Osmanl ruling class and its new converts. For the janissary corps, the Bektashi dervishes became chaplains and guides, promoting an easy-going, tolerant version of
Islam appropriate for new converts. The Bektashis even offered an alternate trinity
in the form of AllahMuhammadAli, and incorporated in its symbolism the number
twelve, significant both for Christians and Shiis. In the palace itself the cultured Mevlevi way held sway; many court poets, musicians and sultans themselves were adherents.
Other brotherhoods were also represented, most with music and hymns and rhythmic
movement in their services. In the seventeenth century there is even mention of people
referred to as hub-mesihi literally, Christ-lover.13 Yet the Osmanl could also be a
very devout Muslim. The powerful grand vezir Sokollu Mehmed Paa (d. ) invited
the strict and austere scholar Birgivi Mehmed Efendi to Istanbul, and honoured him.
The grand vezirs piety is attested by his apologist Feridun Bey, although there might
have been a political motive, to use the puritanical scholar as a counterpoise to the
equally great and influential legislator eyhlislam Ebussuud Efendi. Two generations
later, in the mid-seventeenth century, Birgivis moral descendants and the followers
of his disciple Kadzade Mehmed Efendi divided the palace people as much as they
did Istanbul society. When the fundamentalist Kadzadeli movement threatened a
massacre of those they saw as lax sufis, they were suppressed by the government, but
then the sultan, the palace and the whole ethos of the Ottoman establishment turned
quite conservative, at least temporarily. The palace was big enough to include austere
Muslims as well as Christ-lovers.14

The sultans purpose was to train and educate his own men to become the commanders
and administrators of his palace and his empire. Pages looked forward to graduation
(kma literally, leaving) from palace service, in the first instance from the inner
training ground to functional service in the outer courts. Lesser recruits had already
been sent to the janissary infantry companies, but palace pages left to fill the elite
ranks of the six sipahi, household cavalry, regiments. Daily wages were low, but they
were still the sultans own troops and as such their master still fed and clothed them
in a style befitting the glory of the head of household. Pages who had spent longer in

Metin Kunt

enderun service graduated to office in various special guard units (the left-handed
solak, the fast-runner peyk) or to companies in charge of campaign tents (mehteran hayme), the military band (mehteran- alem) or the stables (istabl-i amire), or were
appointed gatekeepers, kitchen administrators, etc. Those pages who had exhibited
an inclination to learning, arithmetic and calligraphy were employed in the scribal
service, taking notes, drafting fermans (imperial decrees), keeping account books
of revenue grants or imperial accounts. When leaving the palace or re-entering, on
extended hunting trips or on campaign, the sultan would be surrounded by troops,
officers and secretaries, many of whom he would recognize from their years in enderun
The second step in kma graduation was to an independent office and source of
income in the form of a dirlik revenue grant. A simple Janissary would receive a modest grant somewhere in the provinces. Members of the elite cavalry regiments and
officers of the various imperial troops expected more substantial revenue grants and
even posts in provincial administration. However, the graduates of the sultans palace were not the only people receiving provincial revenue grants; in provincial service
they joined local Muslim-born (and, up to about , even Christian) young men
who had volunteered for campaigns, hoping to change status from subject to imperial official. Beys and pashas, district and province governors, not only had their own
large households and retinues but also the right to place their graduates in revenue
grants. In a frontier district up to half the fighting force could come from the district
governors own retinue.15 Although in theory any sixteenth-century provincial officer
could be promoted to a position of command according to merit, most eventual governors had received their initial training in the palace or in a grandee household. The
most select of the province governors, those who became vezirs, thus returned to their
starting point in the capital. They attended council meetings in the second court of
the palace, separated only by the Gate of Felicity from the setting of their earlier years
of education in the inner sanctum, enderun/harem of the palace.


Before it was the custom for an Ottoman prince to leave his fathers house
on puberty (and circumcision) and be appointed governor in a significant Anatolian
sancak district. This represented a sedentarized version of the inner Asian tradition of
the wider ruling clan sharing in government. In Ottoman practice a prince did not
actually share in rule at all, but he did receive training as a potential ruler under the
guidance of an experienced pasha. The question of dynastic politics and the tradition
of open, competitive succession is best treated elsewhere; in our specific context we
should look at how the prince-governors household was formed. While growing up
in his fathers household he had no opportunity to choose his own men; the princely
household was composed of people from the royal palace. As a small replica of the
imperial establishment but with no musket-bearing infantry, a princely household in
the early sixteenth century initially numbered about men, and possibly , later
in the century. If the prince grew to maturity and even to middle age at his provincial
posts, the household might grow with his own recruits, even with gifts of servitors
from various statesmen.16 Even so, a princes household amounted to no more than a
small fraction of the sultans establishment.17

Royal and other households

Upon succession to the throne, the successful prince brought his own household
and retinue to be (re)incorporated into the royal household. A major graduation
from the palace also occurred, both to reward service through independent revenue
grants and to make room for the new sultans retinue, albeit at lower ranks than in
the former princely household. Additionally, because in the succession struggle there
was only one winner, to ensure strict father-to-son succession the losing prince(s) and
their male offspring lost not just the political fight, but also their lives. Their own,
now headless, households were also considered royal property. Some servitors were
punished if they had been closely involved in the succession struggle, but most were
also given revenue grants and palace positions.
Supplying an initial household from the palace was not confined to royal princes.
Upon leaving for the provinces, a senior palace page or an officer of the outer service was also given junior palace servitors to staff his first official household. In such
cases relationships formed within the palace played a part. Ethnic origin, the sharing of a mother tongue, or devirme recruitment from the same village were all reasons for potential friendship and solidarity in the enderun. Indeed, on occasion palace
pages supplied information on specific families and boys who should be considered
for devirme recruitment.18 Palace service also required enderun pages to establish ties
with those in the outer service. The Pantry Chamber pages, for example, dealt with
kitchen personnel in the course of performing their duties. A senior page leaving the
palace would therefore have formed various ties with a group of palace personnel,
on personal or professional grounds, whom he would take with him to a provincial
posting, as the nucleus of an official retinue to be developed further as he progressed
through the ranks.
More importantly, there were the white eunuchs, the supervisors and tutors of the
enderun, who left the imperial palace for high provincial office as sancak district governors just like royal princes, and some of whom rose to great office, including the
grand vezirate. The chief white eunuch, officially titled Commander of the Gate of
Felicity, maintained a household even while at the palace. On the death of a certain
Cafer Aa while still in palace office, members of his retinue were listed according
to their status and functions and the resulting register presented to the sultan for his
decision. As Cafer Aa was still technically a slave of the sultan, the latter inherited
the eunuchs own slaves, and, as in the case of defeated princes, this inherited retinue
was incorporated into the royal household. Perhaps the most interesting aspect is the
fact that some of the slaves of Cafer Aa were already in the palace. The fact that there
were subsidiary households within the large royal household is significant in showing
that, while the sultan was everyones master, there were also lesser loyalties to various

The previous discussion has focused upon the royal household and the establishments
of the mera commanders. The military-administrative career was undoubtedly the
most important in Ottoman government at its height, but it was certainly not the
only one. The men of learning (ulema), studying in colleges (medrese) and following a career alternating between teaching and serving as magistrates (kad), formed
a second branch of government, fully developed into a bureaucratic hierarchy by the

Metin Kunt

sixteenth century. The highest members of the ulema career had in earlier times also
been given political positions, some rising to the vezirate. From the time of Mehmed
II the vezirate was reserved for the mera, but the ulema continued to be represented
in the imperial council by the two kazaskers who headed the learned hierarchy. Upon
elevation to the post of eyhlislam, the pinnacle of the ulema career, a kazasker ceased
to hold his membership in the council; he was now above routine politics. It was
the kazaskers who prepared their appointment lists for all ulema positions as medrese
lecturers, magistrates and administrators of public endowed foundations. The members of the ulema had personal retinues too, though considerably smaller than those
in mera households. Ambitious, bright medrese students were chosen as assistants by
senior ulema and served their masters just as young pages did in an mera household.
As the master rose through the ranks, so did the chances of the apprentice-assistants
of appointment to independent office. Senior ulema were allocated posts as medrese
lecturers and as magistrates, the number commensurate with their rank, to be distributed to a dozen or so of their protgs. The impetus behind this convention may
have been simply recognition of the natural phenomenon of senior scholars favouring
their most brilliant pupils. Once the practice became regularized, however, it became
absolutely necessary for younger scholars to seek patronage; an unattached man was an
unsupported man in the struggle to find appointment. As mera household membership became a necessary step in attaining a revenue grant, so ulema appointments came
to be reserved for protgs. Patronage became indispensible. In other ways, too, ulema
households functioned like mera retinues; when not forming marriage alliances with
other high-placed scholarly households, the master chose sons-in-law among his most
favoured students.19
In later centuries there was a noticeable shift in the Ottoman style of government,
away from those in military-administrative positions to civilian bureaucrats and magistrates. Men from the scribal service or with a medrese education, rather than those
trained in an mera household, rose to the highest positions of government. Bureaucratic patronage ultimately became the most important element in Ottoman politics,
until it was deliberately though not altogether successfully breached during modernizing reforms.


In the seventeenth century, the practice of devirme recruitment fell into disuse. One
can speculate that there were by then too many volunteers for palace service, especially once Selim II had incorporated large numbers of his Turkish princely servitors
and troops into the royal establishment in .20 He had opened palace service to
Muslim-born subjects, mostly Turks. This was a situation criticized and condemned
by tradition-minded Ottoman commentators like Mustafa Ali, but into that breach
others followed, so that in a few generations Turks became a common sight at Topkap.
During the seventeenth century a second new development occurred in practices of
palace service and graduation to provincial office. Palace pages tended to stay in royal
service for much longer periods and on leaving the palace were no longer young, but
had reached a mature, middle age. Senior members of the enderun, the sultans swordbearer and cloak-bearer in the Privy Chamber, routinely went out as full provincial
governors, and of great provinces such as Egypt. This may have been partly a result of

Royal and other households

the fact that the white eunuchs, supposedly their supervisors, had been downgraded
in the palace hierarchy, losing their influence to the black eunuchs of the womens
harem. In this new situation there were no more white eunuchs going out to the
provinces as governors. Nor were the new power-holders, the black eunuchs, ever
granted provincial office or a seat in the imperial council; they wielded their power
from behind the closed doors of the harem. Earlier there had seemed to be no prejudice against white eunuchs as commander-governors, but were black eunuchs only to
be trusted indoors? Would they have met with too much resistance in the provinces?
It is impossible to know whether there might have been racial prejudice at a provincial
post or whether they themselves preferred to wield power at the centre, within the
palace, where they might dominate the councils. Black eunuchs normally left palace
service only to retire, usually to Egypt, where they might still have some political clout,
but not as office-holders.
A further novelty of the seventeenth century was that important officials of grandee
households came to be appointed to high provincial office. Vezirs and pashas had
promoted their own household members in earlier times, and it was normal for their
men to receive revenue grants. The new situation, as in the case of the palace pages,
was that they moved directly from service in a pasha household to important governorships, having held posts neither in the palace nor in lower-level provincial service.
Alongside the imperial palace, grandee households thus became a fast track to high
provincial office.
As for the royal troops, the tens of thousands of janissary infantrymen and the six
regiments of household cavalry, the sheer weight of their numbers and the consequent
difficulty of paying their wages adequately rendered them dysfunctional. Instead of
adjusting their pay for the inflation and silver devaluation experienced during the time
of troubles at the turn of the seventeenth century, the sultan and his advisors chose to
extend to them certain financial benefits. Household cavalrymen, the senior service,
were routinely employed in the collection of royal revenues, for which they were handsomely remunerated.21 The Janissaries, left out of this arrangement, expressed their
resentment in various violent uprisings. Ultimately the palace turned a blind eye to
janissary dealings in the marketplace. Both in the capital itself and increasingly also in
important provincial centres where they were stationed, in Belgrade, Aleppo, Cairo,
Damascus and on Crete, Janissaries dominated economic life, especially in the market
place. So, while the household cavalry turned into royal tax collectors, Janissaries
became artisans and craftsmen.
Finally, one can talk of a relative Turkicization of the Ottoman establishment
from onwards, but this would be a very partial view. Even when Turks and other
Muslims infiltrated major households, the ideal remained to include non-Muslims
as well. In any case, Muslim-born household members also needed to be made into
cultured Ottomans, much like non-Muslims. There may have been few devirme, but
foreign-born or subject non-Muslims still came to the palace and to other households.
By the mid-seventeenth century, Albanians and Bosnians from the western reaches
of the empire and Georgians and Circassians from the Caucasus formed the largest
groups in great households. The conquest of Crete, completed in after a prolonged struggle, furnished a fresh source of recruits. By there were many Cretans
in the harem, both men and women, some of whom rose to great prominence during
the eighteenth century.22 Rabia Glnu Sultan, the mother of Mustafa II ()

Metin Kunt

and Ahmed III (), was the most prominent Cretan in Mehmed IVs harem.
During the s, the Greek revolt gave rise to a new wave of slaves. Rebels were
suppressed in various localities, most ruthlessly on the islands of Chios and Samos, in
response to the massacre of Morean Muslims; Chiote and Samian women and children, lawfully enslaved because they were rebels, filled great households and furnished
some prominent Ottomans in later decades.23
As the boundaries of the empire continued to shrink after the Karlowitz treaty of
, Hungarians resisting Habsburg rule sought refuge in Ottoman lands. Some
kept to themselves in areas allotted to them, but others entered Ottoman service in the
palace or in other households. In the nineteenth century there were further waves of
Hungarian and Polish immigrants escaping Habsburg or Romanov wrath in the aftermath of the uprisings. Once again, some immigrants kept to themselves, most
famously in the Polish Village (Polonezky) not far from Istanbul. Many Hungarian
and Polish leaders converted to Islam and became prominent Ottomans, now not
necessarily going through Ottoman households but establishing their own upper-class
Ottoman families.24
The household as a socio-political unit came to an end by the mid-nineteenth century with the conscious efforts of tanzimat statesmen to replace the paternalistic old
Ottoman way with a new, modern, functional bureaucracy. The Ottoman system that
had looked to fifteenth- and sixteenth-century European observers wonderfully meritocratic, because it was not aristocratic, now seemed to reforming Ottomans themselves
hopelessly old-fashioned and subject to factionalism based on personal interests. As in
European societies, the slave trade was abolished, though not slave status as such;25 the
bureaucracy came to be organized not around households but in ministries; promotion was to be based on merit and not on patronage relations. Great houses gradually
disappeared, and with them households, though patronage, sometimes based on family and household connections, declined much more slowly.26


I thank the Collegium Budapest Institute for Advanced Study for their fellowship and gracious
hospitality in spring when I worked on this essay.
I have discussed revenue grants and households at greater length in The sultans servants (Kunt
), chs and ; also in Turks in the Ottoman imperial palace (Kunt ).
For a grant of , see Kunt : .
nalck a; on the economy, see also chapter , by Rhoads Murphey, in this volume.
Shaw a.
Crone .
Jackson ; Ayalon .
nalck b.
Mnage ; Demetriades .
Kunt : , n. .
The most comprehensive studies of the palace and imperial household remain the works of .
H. Uzunarl, on household troops () and Ottoman palace organization ().
Kunt .
Fleischer a: .
Ocak : ; Ocak .
Baer .
Dvid : ; Dvid and Fodor .

Royal and other households

Fodor and Sudar .

Kunt .
Kunt .
On the ulema career, see Uzunarl ; Zilfi .
For additional comment on this, see chapter , by Oktay zel, in this volume.
Darling ; Khoury ; Adanr ; Masters .
Aksan .
For a fuller discussion, see Findley . The most illustrious Ottoman family of Chiote
origin is the grand vezir Ibrahim Edhem Paa (d. ) and his descendants, including Osman
Hamdi Bey, painter and founder of the School of Fine Arts and of the Istanbul Archeological
Mustafa Celaleddin Paa, a Polish convert, married into a high Ottoman family; he was the
grandfather of the poet Nazim Hikmet.
Erdem ; Toledano .
Findley ; Erdodu ; Bouquet .



The construction of a topos

Hakan T. Karateke

hose who received their high school education in Turkey learned from the history
textbooks that one of the reasons for the stagnation and decline of the Ottoman
empire was the lack of participation by sultans in military campaigns. The section that
dealt with this decline usually followed one on the glorious Sleymanic age (
), and gave either the reign of Selim II () or that of Murad III ()
as the beginning of the period of stagnation and decline. Although the absence of the
sultan from the battlefront was not necessarily prominent among the many internal
and external factors that were usually listed in this section, it tended to stick in the
collective memory of those who went through the Turkish education system. This
historical detail was perhaps so salient because it was generally mentioned in connection with the fact that sultans now spent much of their time submerged in pleasures
in their palaces. Such information may have been more likely to capture the attention
of a typical teenager, in contrast to the otherwise quite tedious narratives of history.
One could justifiably view this chapter as my attempt to set the record straight
for at least one of the teachings with which I was inculcated during my high school
education. But the situation is actually much more alarming than that. This largely
unquestioned explanation, ubiquitous not only in Turkish history textbooks, shows
to what degree we may have internalized the standard narratives that have become so
much a part of the natural paradigms originally established on the basis of nothing
but historical narratives. While there may be no way around the texts that we have
at hand, reading them uncritically and without context is probably one of the biggest
obstacles to producing a reasonable historiography of the Ottoman empire.
This essay will investigate the circumstances surrounding the genesis of the view that
the increasingly frequent military debacles from the late sixteenth century onwards
were caused, among other things, by the sultans physical absence from campaigns.
While modern historians might dismiss this line of reasoning as insubstantial, they
tend not to consider under what circumstances such an explanation could be advanced
by Ottomans at the time. I will explore how and why the controversy about the sultans participation in military campaigns arose in the late sixteenth century, and how
it made its way into twentieth-century textbooks. A survey of how the issue is represented in certain contemporary historical and political works enables us to explore

On the tranquillity and repose of the sultan

at least some views on the ways in which public opinion was shaped in the Ottoman
capital, how the imperial image was promoted and, at times, challenged, and how the
concept of rulership evolved over time. Although the topic raises many related questions on the mobility or immobility of Ottoman sultans in general, analysis of these
broader questions of how often, how far, where and, most importantly, for what purposes the Ottoman sultans travelled remains for another study.


Mustafa Ali, a bureaucrat of the late sixteenth century, writes rather apologetically in
one of his major historical works, the Knhl-ahbar, about Murad IIIs reluctance
to go on campaign. In an extra section composed after Murads death in , and
introduced by the heading On the tranquility and repose of the sultan, Ali investigates the reasons for the sultans unwillingness to set out on campaign, even though his
vezirs were very much in favour of it. Murad III, relates the author, would not budge
from the Ottoman capital once he had made the journey from Manisa to Istanbul to
be enthroned. Ali dismisses the argument that the sultan had epilepsy and therefore
could not travel. He produces an explanation that the sultans personal substance
(zat) required immobility. None of the letters in his name, Murad, had the quality to
cause him to move. All of them, says Ali, carried characteristics that would increase his
inclination towards the ground.1
At the time Ali devised this hypothesis, the twenty-one years of Murads reign, in
addition to the eight-year reign of his father Selim II (), had passed without
a sovereign having actively participated in any military campaign. Discontent about
Murads inertia had sprung up already during his lifetime. Ali clearly touches upon
the subject in his Nushats-selatin of , although not as straightforwardly as in
his later work. After describing in detail the squabbles between the commanders Lala
Mustafa Paa and Sinan Paa before they left for the Iranian campaign of , Ali
claims that, if Murad had accompanied the army, all the Iranian lands would have
been seized and the situation on the eastern front would have been resolved completely.2 Another writer, Asafi, focuses on the same Iranian campaign in his ecaatname,
written in , a few years after the Nushat. He claims that, had the sultan taken
the trouble to campaign, then not only the Iranian lands, but all the territories as
far as China would have come under Ottoman control.3 Of course, if the military
campaigns of these years had resulted in overwhelming success, the issue of the sultans participation would probably not have arisen. The arguments revolve primarily
around the fact that, although these campaigns did not end in total disaster, they
were far from being clear victories, and were a significant drain on the imperial treasury. No such criticism is known to have arisen during the reign of Murads predecessor, Selim II. The difference here is that a favourable peace treaty was signed with the
Austrian Habsburgs two years after Selims accession and that in , most spectacularly, Cyprus was taken from the Venetians by the commander-in-chief Lala Mustafa
Alis dismissal of the claim that his suffering from epilepsy was an explanation for
Murads reluctance to go on campaign is clear evidence that the issue was controversial and that various explanations were being openly debated. Salomon Schweigger, a

Hakan T. Karateke

Figure . Sleyman in procession through Istanbul, c.. From Pieter Coecke van Aelst,
The Turks in MDXXXIII: a series of drawings made in that year at Constantinople by Peter Coeck
of Aelst, and published from woodblocks, by his widow, at Antwerp in MDLIII; reproduced
in facsimile (London, ) and as Gravrlerle Trkiye I (Ankara, ), plate .
Reproduced here with permission of the Turkish Ministry of Culture.

Protestant preacher who spent four years from to in Istanbul as part of the
entourage of the Habsburg envoy, heard about Murads disease (fallende Krankheit).4
Aside from his failure to attend a single campaign, Murad did not leave the palace
during the last years of his reign. For two consecutive years he failed even to attend
the Friday procession to an imperial mosque, a unique occurrence in the Ottoman
throne city. According to the historian Selaniki (d. ), a vigilant observer of late
sixteenth-century events, whenever Murad wanted to go out for the Friday prayer, he
was frightened into changing his mind by allusions to an alleged plot by the Janissaries
to dethrone him when he left the palace for that purpose.5 Selanikis explanation for
Murads resistance to going on campaign also cites more directly the strained relations
between the sultan and the Janissaries, recounting that the latter several times made
bold demands of the sultan personally at the imperial council. As a result, Murad did
not trust the Janissaries to support him in the tensions that were likely to arise during
a campaign.6
Critical remarks about Murad IIIs military immobility make even more sense when
one considers his grandfathers and great-grandfathers itineraries (see tables . and
.). Sleyman I and Selim I (), in stark contrast to their descendant, were
among the most militarily mobile of all Muslim sultans of any era. Sleyman went on
thirteen campaigns: five times to Hungary, including the Vienna campaign of ,
and three times to Iran and Iraq, going as far as Tabriz and Baghdad. He was away

On the tranquillity and repose of the sultan

from Istanbul once for twenty months and on another occasion for twenty-three, and
altogether was absent from his throne city for almost one-fourth of his reign; at the
time of his death, at the advanced age of seventy-two, he was on campaign in Hungary.
Selim I was remembered as seldom sedentary (kalill-karar) by the imperial annalist
Naima almost two centuries after his death.7 The successful campaigns of Selim and
Sleyman were constantly cited as acclaimed examples in order to foster the ideal of a
combatant sultan.8 Their military successes and their desire to be at the war front in
person became linked in the minds of many, and laid the grounds for the construction
in later literature of a topos about the physical absence of the sultan from the battleground as an explanation for the increasing prevalence of unsuccessful battles, or even
of civil unrest.
Table .

Campaigns of Selim I




Departed for Edirne,
Died in his tent near orlu



Source: Based on Anon. .

Notes: Reigned : , days; approximately , days (. per cent of his reign) were spent away
from Istanbul. Selim made a trip to Edirne only two weeks after he returned from the Egyptian campaign
(August ) and stayed there for about ve months.

Table .

Campaigns of Sleyman




Belgrade, Hungary
Hungary (Mohcs)
Hungary (Kszeg/Gns)
Iran, Iraq (Tabriz, Baghdad)
Hungary (Buda)
Hungary (Esztergom)
Iran (Tabriz)
Stay in Amasya
Hungary (Szigetvr)
Died on campaign, September

June 10
End of April


Source: Based on Anon. and Anon. .

Notes: Reigned : , days; approximately , days (. per cent of his reign) were spent away
from Istanbul. Only his thirteen campaigns are listed here. Occasional trips to Bursa or Edirne for a duration
of up to three months are not listed. Sleyman also stayed in Edirne from November to April i.e.,
between two campaigns.

Hakan T. Karateke


When Mehmed III succeeded Murad III in , the controversy and rumours of
immobility persisted. Mehmeds absence from one of the first Friday processions of his
reign ( February ), on account of severe weather, may have encouraged these
rumours, although he did appear the following week at the Sleymaniye mosque.11
The Janissaries were very much in favour of the sultans leading the campaign which
was likely to set out during the next season. While the army, under the command
of Koca Sinan Paa, was still at the Hungarian war front, the janissary commanders,
meeting in their Istanbul barracks, were critical of the fact that this robust sultan,
twenty-nine years of age, refused to go on campaign with them, whereas Sleyman
had done so even when he was old and sick and had to use a carriage. Janissaries on
duty at the palace staged a protest at the next meeting of the imperial council in their
traditional way by leaving before the soup. At the end of September they took
an oath stating that they would go on another campaign only if the sultan also were
present.12 That same day, several kadis and medrese professors created a small commotion during a sermon at the Sleymaniye mosque by accusing not only the sultan, but
also the ruling elite, of passivity in the face of insults such as the capture of Muslim
families after the fall of Esztergom castle in early September.13 At the beginning of
November, the preacher of the Ayasofya mosque was sent to the palace to deliver a
moving sermon on the Islamic duty of commanding right and forbidding wrong.
Mehmed III burst into tears.14
Discussion on this subject probably continued throughout the city during these
months. On December, Koca Sinan Paa was appointed grand vezir for the fifth
time, and on December he invited the senior religious authorities, including the
chief mufti and the chief military judges of Rumeli and Anatolia, to a lavish banquet to
discuss the issue. This occurred despite the fact that Sinan Paas dislike of the ulema
was apparently no secret.15 The party initially rejected the idea of the sultan himself
going on campaign, citing the practical difficulties of raising an appropriate army, in
terms both of the necessary number of soldiers and of the splendour that a campaign
featuring a sultan would require. They also raised the example of the last two sultans,
who had successfully guided Ottoman armies from the throne, via commanders, and
had even conquered new lands. A group of ulema reasoning about the number of
soldiers might seem strange in another context, but in this case it indicates the extent
to which the issue had become a public concern. What this group of high-ranking
religious figures actually had in mind when arguing against the presence of the sultan
on campaign, and whether there were ulterior motives for their position, cannot be
established from the documentation at hand. However, it is quite likely that this was
not the recommendation that Sinan Paa was aiming for when he arranged the banquet. By some unclear means, towards the end of the meal opinion shifted in the other
direction, and the meeting finally concluded that, if enough supplies and provisions
could be dispatched to the front, it would then be appropriate for the sultan to set out
with the army.16
According to Baron Wratislaw of Mitrowitz, a Habsburg diplomat imprisoned in
Istanbul during this time, Sinan Paa capitalized on the general discontent of the Janissaries and sipahis (palace cavalry) by further inciting them to complain. They reportedly submitted a petition to the sultan during a Friday procession, requesting him to

On the tranquillity and repose of the sultan

go to Hungary against the Christians, and follow in the footsteps of his predecessors.17
While deliberations continued, Sinan Paa entered Mehmed IIIs presence and advised
the sultan that it would be utterly wrong to send the army to the battlefront under a
commander-in-chief who was merely the grand vezir or one of the lower-ranked vezirs.
If the grand vezir was sent as commander, his deputy (kaimmakam) in Istanbul would
purposely withhold further soldiers and provisions from the army in order to cause the
grand vezir to be unsuccessful, in the hope of damaging his reputation and ultimately
replacing him. If another vezir was appointed commander, the grand vezir himself
would not want him to succeed and thereby become a possible contender for his own
position. A higher authority such as the sultan himself would be a solution to all these
problems, suggested Sinan Paa.18 Such a view evidently arose as a consequence of the
behaviour of contending commanders-in-chief in previous campaigns. Sinan Paa is
generally depicted in Ottoman chronicles as trying to persuade the sultan to go on
campaign out of fear of carrying the responsibility for a likely disaster entirely on his
own shoulders. He had barely escaped execution for his failures during the defence of
Esztergom and other castles a few months previously.
Further criticism from other authors indicates that discontent about the commanders was prevalent. Alis views on the sultans military inertia suggest that it was not only
the idea of the sultan remaining in Istanbul during a campaign to which the bureaucrat was opposed. He was also troubled by a related development, namely that the
military commanders were becoming too powerful on account of the awe they created
around themselves, and that they were intruding into spheres to which they otherwise
had no right, such as the appointment of judges. In Alis critical, perhaps over-sensitive view, this had developed into something like a dictate of commanders (serdarlar
saltanat).19 Hasan Kafi, himself present on the Eri campaign, had few good
words to say about the Ottoman military commanders. In his treatise on statecraft,
Usull-hikem, which he rewrote in the euphoria after the campaigns success, he warns
the sultan not to trust anyone other than himself and recommends that the sultan in
person, and not the commanders, who were often negligent, should inspect the army
before a battle. His advice, though offered indirectly, was that the sultan should lead
the army in person on campaigns.20
However, it appears that there was another faction in the palace, led by Mehmed
IIIs mother, Safiye Sultan, which was vehemently opposed to Mehmeds going to
the battlefront. Although Baron Wratislaws assertion that she had claimed, relying
upon the Alcoran, that no new sultan shall be obliged to go to war for the space of
three years is probably incorrect, it nevertheless accurately reflects her position in
this matter.21 Mehmed perhaps hoped to assert his authority over his very influential
mother by personally leading a successful campaign that would build up his charisma
and aura.
The grand vezir Sinan Paa died in April , and Mehmed III himself led the
army to Hungary, as a result of which the fortress of Eri (Eger) was captured and a victory secured over a combined HabsburgTransylvanian army at the battle of Haova
(Mezkeresztes). According to Ottoman narratives, after the fall of the Eri fortress,
when news spread that the Habsburg army was approaching, Mehmed wanted to
dismiss his army and return to Istanbul. He also showed signs of wanting to flee the
Haova battlefield when the Ottoman army initially appeared defeated. However, in
the end, as narrow a victory as it turned out to be, Haova did validate the formula at

Hakan T. Karateke

this particular moment: when the sultan came along, victory was assured. Talikizade,
participating in the campaign as official historiographer, states that the thirty-yearplus hiatus in sultan-led campaigns had resulted in the amassing of munitions by the
enemy, assaults on Muslims, and the vanishing of the awe they felt vis--vis the sultan.22 Now victory was secured.
After returning to Istanbul, Mehmed assured his vezirs during the Friday prayer
that he would campaign again. His words quickly reverberated throughout the city,
and even caused market prices to rise.23 However, although Mehmed spread similar
rumours later in his reign, especially when his rule became insecure, the Egri
expedition remained his only military campaign. His successor, Ahmed I (),
did not campaign at all. While the Zsitvatorok treaty of with the Habsburgs
stabilized the European front, the celali revolts and Safavid wars in the east dominated the next few decades for the Ottomans, but it was not until that another
sultan, Osman II (), again led the army in battle. Subsequently, other seventeenth-century sultans Murad IV (), Mehmed IV () and Mustafa
II () did lead relatively successful campaigns (see table .).
It seems safe to assume that a new kind of rulership was already in the making
by the late sixteenth century, whereby the practice of imperial seclusion now widened to include the sultans military activities. This may have derived partly from the
personalities of Sleymans immediate successors or from special circumstances surrounding their reigns. Yet, resistance to this development from various groups forced
Mehmed III and, later, his successors to be militarily more active. It did not take much
effort for the general public to be aware of the sultans immobility and to turn it into
easily manipulable common knowledge. It was then used as a political argument when
the need arose, and, once the issue became the subject of open debate, it gained more
political weight, hindering or delaying evolution in the idea of rulership.

This debate probably remained vivid in the political and public memory for some
time. It was handled in some detail by certain seventeenth-century chroniclers, who
used narratives contemporary to the events as sources, though others passed over the
subject in silence. An association of military failures with the sultans not leading the
army may thus have remained a conviction for some, especially those close to janissary
circles. Whenever futile military campaigns multiplied, discontent arose in some sections of society. An uneasy desire to quell any anxiety on this topic perhaps ensured
that, when a sultan did participate in military campaigns, it now became the subject of
extraordinary emphasis. According to Silahdar Mehmed, for instance, no other sultan
ever made as grandiose and heroic an appearance as did Mehmed IV at a parade in
Edirne that preceded his departure for the Kamanie campaign in .24
Neither Veysis Habname nor Mustafa Safis Zbdett-tevarih, both of which were
written in the s, contains a discernibly critical view on the issue of the sultan leading the army. In fact, Safi, an imam at the imperial palace, whose book is an extreme
example of panegyric historiography, cites Mehmed III as an ideal sultan with respect
to his decision to go on the Eri campaign. He claims that no other sultan had ever
attained or even come close to attaining the good reputation that Mehmed acquired
by protecting the honour of the sultanate with this campaign.25 Only a decade or so

On the tranquillity and repose of the sultan

Table .

Campaigns after Mehmed III (setting o from Istanbul or Edirne)

Ahmed I ()

Pest (Austria) (+)

East (Safavids) (~)
{ Esztergom (Austria) (+)
{ East (Safavids) (~)
{ Revan (Safavids) ()
{ Ardabil (Safavids) (~)
* Hotin (Poland) (~)
{ Baghdad (Safavids) ()
{ Hemedan and Baghdad (Safavids) ()
{ East, against the Safavids, who laid siege to the castle of
Van; campaign downgraded when siege is abandoned.
* Poland; campaign abandoned at Edirne when agreement is
reached with the Poles.
* Revan (Safavids) (+)
* Baghdad (Safavids) (+)
{ Azov (Russia) (+)
{ Crete (Venice) (~)
{ Transylvania (+)
{ Uyvar/Neuhusel (Austria) (+)
{ Crete (Venice) (+); sultan decides to join the army in
August , sets o from Edirne, but receives the news of the fall of
Candia on the way and proceeds to Salonica.
* Kamanie (Poland) (+)
* Hotin (Poland), Ukraine (~)
*{ Czehry (Russia); (+); sultan leads the army
as far as Hacolupazar and stays there for three months.
* Russia (~); peace agreement signed before the campaign
*{ Vienna (); sultan leads the army as far as Belgrade and
stays there for ve months.
*{ Austria (); sultan leads the army as far as Soa.
{ Belgrade (Austria) (+)
{ Peterwardein (Austria) (~)
* Lippa, Lugos (Austria) (+)
* Timioara (Austria) (+)
* Zenta (Austria) ()

Mustafa I ()
Osman II ()
Mustafa I ()
Murad IV ()

brahim ()
Mehmed IV ()

Sleyman III ()
Ahmed II ()
Mustafa II ()

Notes: * sultan leads the army

campaign in a foreign state; sultan does not lead the army
+ victory
~ drawn battle: no clear outcome, or peace treaty

had passed since the controversy surrounding this topic, and what Safi meant by such
an allusion must have been easily grasped. Veysi, on the other hand, a kadi by profession, who presented his advice treatise to the grand vezir Nasuh Paa, does mention
the recurring futile campaigns on the eastern front as the greatest public concern of
these years, but he does not even implicitly mention the sultans failure to lead the
army as a cause.

Hakan T. Karateke

A direct reference to the issue is found only in the anonymous Kitab- mstetab.
This book, which was probably written around by someone of devirme origin
and in all likelihood sympathetic to the Janissaries, is an advice treatise and uses this
rhetorical format to bring several issues to attention. Seven questions addressed in the
addendum of the book to a fictitious group of members of the ruling elite are placed in
the mouth of the sultan. The first question enquires whether the reason for the rising
Safavid threat and unrest in Anatolia could be that the sultan had not personally been
leading campaigns to Iran.26
One would expect that such concerns, whether widespread or not, would subside
after Osman IIs Hotin campaign in and disappear by the mid-s, with
Murad IVs successful campaigns to Revan and Baghdad (see table .). Indeed, Kou
Bey, writing in the s, does not raise the issue as a problem, but mentions Sleyman Is participation in campaigns only as part of his efforts to gather information
about his subjects.27 Katib elebis Dsturl-amel, also written during the s in
the tradition of advice treatises, does not touch upon the issue at all. However, a
decade later, when Hezarfen Hseyin is gathering laws pertaining to state institutions
in his Telhisl-beyan (), he regards both options, namely either the sultan or a
commander leading a campaign, as equally valid. Telhis may be regarded as a different
genre, since it is a compilation of legal material, which often had a timeless language
repeating previous wording. However, Hezarfen Hseyin made the compilation for
an audience other than lawmakers or jurists, and often rearticulated other laws in his
own words and integrated them with contemporary views.
The laws of campaigns: first, it is clear that the greatest task of the sultan is to conquer lands, to drive away the enemies, and fight those whom we are required to
fight. This task is carried out either by the sultan going on a campaign in person or
by him appointing someone trustworthy from among the vezirs or commanders as
a commander-in-chief over the army.28
The sources quoted above indicate that sensitivity on the subject was high at the end
of the sixteenth century, gradually receded during the course of the seventeenth, and
by around was accepted as one of two alternative practices of campaign leadership. Such evidence suggests that the debate and the sensitive public opinion about
the sultans going on campaign were context-specific. The debate should therefore be
understood in its contemporary context, and not necessarily as a continuing issue.

While a thorough examination of the historiographical literature of the eighteenth
century may help produce a more accurate perspective on the changing perceptions of
this issue, the actual discussion seems to have faded away. However, the topos resurfaced in modern Ottoman and Turkish historiography. Starting in the second half of
the nineteenth century, some historians dealt with the issue within the larger development of a modernist, positivist, disenchanted and anti-monarchical worldview. Historians in this tradition first judged ahistorically the sultans failure to lead military
campaigns as non-compliant with the standards of the modern world, and eventually
came to view this as one of the reasons for the decline of the Ottoman empire. Among

On the tranquillity and repose of the sultan

others, Frederick II of Prussia and Napoleon Bonaparte enjoyed great reputations

in the nineteenth-century Ottoman military imagination as ideal enlightened leaders
who also frequently led their armies personally.
Hayrullah, in his History of the Ottoman state, a much admired work of its time
(published between and ), philosophizes at length about the benefits of the
sovereigns being at the war front. In his view, it was beyond doubt that the Egri
campaign would have been unsuccessful had the sultan not been there. He sees an additional force in the sultans person which infuses firmness and fortitude into the soldiers.
After listing examples from ineffectual campaigns in which Murad III and Mehmed III
did not participate, Hayrullah evaluates the issue from the perspective of civilization:
even the sovereigns of cultured, civilized and well-regulated countries (terbiye, temeddn,
nizam) were now present in person on the battlefield. The crux of his argument is that
the sovereigns being at the war front should not be considered Bedouinism, a concept
frequently used in opposition to civilization in the second half of the nineteenth century.29 By handling the issue within the political and cultural terms and concerns of his
time, Hayrullah clearly brings a totally new perspective on the controversy.
Another historian who touched upon the issue is Abdurrahman eref, whose History of the Ottoman state was written as a course book for schools of higher education
in the early s. eref was also the last person to occupy the post of imperial annalist of the Ottoman empire. However, his reference to the issue is neither explicit nor
presented in the framework of a decline paradigm. The relevant section in the first
edition of his book describes the deliberations of Mehmed III over whether to lead the
Egri campaign:
Sultan Mehmed III prepared himself to go on the campaign in person. The crescent of the imperial standard, which had withdrawn itself into the shadow of
silence since the time of Sultan Sleyman, [rose] again from the horizon of the
Divine guidance and illuminated joyful eyes.30
A curious approach to the issue comes from Mizanc Murad, in his Tarih-i EblFaruk. The book is a popular history written by a journalist. Murad often colours his
narrative by filling in the inevitable blanks quite imaginatively and by freely inserting
several unfounded claims. Although otherwise adhering to a demystified worldview
which no longer tolerated a predestinarian historical approach or astrological explanations for events, Murads depiction of the issue can best be understood as a product of
his dramatization skills. When dealing with the reign of Mehmed III, he depicts a general situation of despair in late sixteenth-century Ottoman society.31 According to his
narrative, after the fall of Esztergom castle in September , which indeed caused
profound public unease, there was mourning and grief everywhere, earthquakes and
floods occurred, many strange signs were observed, such as the sighting of fish that
had not been known before, and water turned black as ink in some places.32 Yet, a
miracle happened among all this confusion. A divine sign which was particular to the
Orient appeared.33 This miracle was Mehmed IIIs decision to lead the army to Eri.
The labelling of this decision as a miracle was the result of Murads conviction that
the order of the army was disbanded since the sultans stopped going on campaigns.34
He presented this as the only miracle in the entirety of his portrayal of a miserable

Hakan T. Karateke

While the issue was treated from diverse perspectives by widely read, late Ottoman
historians, Hayrullah and Mizanc Murad seem to have a common thread in their
approach. A self-orientalizing angle can be detected in their treatment of this historical event. Hayrullah feels compelled to evaluate the event in respect to its compatibility with the requirements of civilization in general and uses such loaded words as
Bedouinism, thus clearly positioning himself within the common Orientalist discourse
of the time. Mizanc Murad also uses Orientalist images, referring almost humorously
to divine signs and miracles in his narrative of Mehmed IIIs participation in the
Egri campaign. Significantly, neither of these historians attempted to present the issue
within a declinist discourse. Such a viewpoint emerged only in the Republican period.
The first history textbook of the Turkish Republic, designed for high schools and published in , defined the parameters of this paradigm for decades to come. Under
the section headed The main reasons for the period of stagnation, the book states:
In the absolutist Ottoman state, the ruler was the executor of the state affairs and
the commander of the army. The sultans since Selim II except for one or two
did not go on a campaign with desire or ardour. Thereafter, the sultans and the
princes withdrew themselves into the palace and spent their lives among women
and eunuchs.35 [emphasis in original]
One of the most influential general history books in popular as well as academic
history writing in Turkey, smail Hakk Uzunarls Ottoman history reproduces this
view in similar words.36 While it is clear that Uzunarl wrote within the Republican discourse in terms of how he presents the issue, it is likely that he consulted the
original sources as well. A manuscript copy of Kitab- mstetab, quoted above as one
of the most explicit criticisms of the sultans general reluctance to lead campaigns, was
found in Uzunarls private library.37 Later, several generations of pupils at Turkish
high schools would be taught history from textbooks written by Emin Oktay, who also
shared this view.38 A quick glance at more recent textbooks shows that the perspective
has persisted.39 This view has even found its way into a general history of the Ottoman
empire in the English language.40

While the sultans presence on the battlefield probably had a limited effect on actual
military tactics, there is no doubt that it was a tremendous boost to the soldiers morale.
This was especially critical at a time when the outcome of battles depended highly on
personal combat skills. Moreover, perhaps partially based on convictions about divine
assistance to a sultan-led campaign, campaigns carried out under the sultans personal
command raised expectations of a victorious outcome. Following Osman IIs unsuccessful Hotin expedition, brahim Peevi, a contemporary historian, concluded the
narrative of this campaign in his History with the thought that it was surely a warning
sign from God that the Ottomans were not victorious despite the fact that they had
fought with two sovereigns, namely Osman II and the Crimean Khan, against the
Poles, whose king was not even on the battleground.41
Furthermore, the material allowance for a sultan-led campaign usually surpassed
the budget of a regular campaign. Going back to the arguments made during the

On the tranquillity and repose of the sultan

banquet at Sinan Paas mansion, the initial course of action eventually agreed by the
high ulema was immediately to send ample ammunition and provisions to the front as
a preparation for the Egri campaign. A sefer-i hmayun (or hnkar seferi, a campaign
in which the sultan personally took part) was always equipped with much more abundant material provisions in addition to the necessary imperial splendour (tecemml-i
saltanat) required.42 No doubt such a well-equipped and better-organized military
expedition had many favourable implications for the army as a whole, which would
also justify the belief that the sultans leading the army on campaign had the potential
to affect its outcome positively.43
The equation was nevertheless far from being so straightforward. There were obviously many failed campaigns in which the sultan himself took part, whereas several
others led by a commander-in-chief other than the sultan turned out to be overwhelmingly victorious. The perception that, from Selim II onwards, the Ottoman sultans
no longer leading military campaigns is one of the reasons for Ottoman decline is,
whether plausible or not, characterized by historiographical defects. A context-specific
sensitivity in public opinion regarding this matter did indeed arise at the end of the
sixteenth century, and a general debate on the causes of the not very successful military
campaigns of that time did take place. The theme remained vivid in the political and
public memory at least until the s and was reflected in some, but not all, contemporary historiography. By the mid-seventeenth century the debate had subsided, and
sensitivity to it gradually disappeared. On the other hand, many seventeenth-century
sultans again led the army in person on actual battlegrounds. A variant also practised
by these sultans was to lead the army to a principal stage (such as Belgrade or Sofia)
on one of the Ottoman viae militares and to quarter there until the return of the army.
Clearly, neither this latter formula nor the leadership of an able commander-in-chief
(such as a member of the Kprl family) was seen as a problem, or at least not one
that was important enough to be invoked in the explanations for failed campaigns of
the seventeenth century.
Yet, that context-specific debate was reproduced first in late Ottoman historiography, and then in republican Turkish historical writing, in different contexts but
with little reference the original circumstances. Republican historiography presented
the subject matter of a genuine discussion teleologically as a factor in loss of military
strength, even a general decline, of the Ottoman empire from the late sixteenth century onwards. Controversial views were attributed to a period later than the one in
which discussion actually occurred, a practice most probably imported into republican
historiography initially by quotation of late sixteenth- and early seventeenth-century
Ottoman works. In effect, incautious use of such views was no different than the
similar reproduction of other factors of decline indeed, the entire decline paradigm
from Ottoman works into Turkish republican historiography.
However, the appearance of this topos in differing contexts results in its being more
than just a repeated stereotype. It also holds clues to how concepts of rule in the Ottoman empire changed. Around the last quarter of the sixteenth century, as the sultan
was becoming a militarily less mobile ruler, the challenges he faced because of this
development delineated the limits of his authority. The political conjucture, agendas
of pressure groups, and resulting public opinion forced some sultans to be militarily
more active than they might have wished to be. Subsequent perspectives on the same
string of incidents offer us a comparative insight into how expectations of sultanic

Hakan T. Karateke

rule evolved over time. While the seventeenth century saw the delayed acceptance of
evolution in the military role, notions of rule among the Ottoman intelligentsia in
the nineteenth century were shaped by modern disenchanted and anti-monarchical
worldviews. Study of Ottoman perceptions of going on campaign and the sultans
military role therefore has much to contribute to the history of political thought in the
Ottoman world.


Der huzur u rahat- hakani, Mustafa li on Murad III (li : II, ).

li : II, .
li : .
Asafi Dal Mehmed : fol. a.
Schweigger : .
Selaniki Mustafa Efendi : II, .
Ibid.: I, .
Naima : II, .
E.g., Peevi : II, ; Hezarfen Hseyin : .
Some dates are disputed, hence durations are approximate.
Army leaves on July.
Selaniki Mustafa Efendi : II, , .
Ibid.: II, , .
Ibid.: II, .
Ibid.: II, .
li : III, .
Selaniki Mustafa Efendi : II, .
Wratislaw : .
Hasan Beyzade Ahmed : II, . This argument found its way into later historiography:
e.g., Peevi : II, ; Katib elebi : I, ; Naima [] : I, ;
dOhsson : .
li : .
Hasan Kfi el-Akhisar : .
Wratislaw : .
Talikizade Mehmed Subhi : ff.; Woodhead : .
Selaniki Mustafa Efendi: II, .
Silahdar Mehmed : : also quoted in Doru : .
Safi : II, .
Anon. : .
Koi Bey risalesi : .
Hezarfen Hseyin : ; Hezarfen Hseyin : .
Hayrullah : XIV, .
Abdurrahman eref : II, . The second, more concise edition omits the section referring to Sleyman and simply says: Sultan Mehmed III decided to go on the campaign in
person (Abdurrahman eref : II, ).
Mehmed Murad : .
Ibid.: .
Ibid.: . Bu kargaalk arasnda bir mucize vaki oldu. arka mahsus bir ruhaniyet numunesi
ba gsterdi.
Ibid.: , .
Tarih III : . Tevakkuf devrinin balca amilleri: . . . Mutlakiyetle idare edilen
Osmanl devletinde, hkmdarlar devlet muameltnn nzm ve ordunun kumandan idiler.
Selim II. den itibaren padiahlar bir ikisi mstesna heves ve gayretle sefere kmamlardr.

On the tranquillity and repose of the sultan

Bundan sora [!] padiahlar ve ehzadeler tamamen saraya kapanmlar ve mrlerini kadnlar
ve haremaalar arasnda geirmilerdir.
Uzunarl : .
Uzunarl : .
Oktay : .
Kara : , mentioned among the internal causes of Ottoman decline.
Shaw : , quoted among the political and military factors of decline.
Peevi : II, . My thanks to Murat Yaar for this reference.
Selaniki Mustafa Efendi: II, .
However, some sources present the financial aspect of campaign preparation as one of the reasons for the Ottoman administrations reluctance to organize a sultan-led campaign: the grand
vezir Boynu Yaral Mehmed Paa persuaded Mehmed IV to abandon the idea of campaigning
in by arguing how much more expensive it would be if the sultan were present (Naima
: IV, , though there may have been other reasons not mentioned by Naima).



Sixteenth-century Ottoman imperial interpreters as
Renaissance go-betweens

Tijana Krsti

ver the last decade scholarship on the early modern era has begun to recognize
the Ottoman contribution to cultural processes such as the Renaissance, humanism and confessionalization, once thought of as uniquely European.1 With this new
interest in integrating the Ottomans into the study of early modern European and
world history also came curiosity about a variety of cultural intermediaries who helped
construct shared conceptual frameworks across geographic, cultural and religious
boundaries. Captives, converts, merchants, sailors, travellers, artists and diplomats
have become central to investigations into the mechanisms of cultural exchange and
the nature of boundaries which were both created and transgressed in the early modern
era.2 Although professional translators, as cultural intermediaries par excellence, have
long attracted the attention of scholars, until recently their activities were considered
mostly from the perspective of political and diplomatic history. Their backgrounds,
social networks and relationship to the centres of power, the translation regimes that
they adhered to and the subjective conditions in which they produced translation are
only now being recognized as important aspects of historical study on translation and
cultural mediation.3
The present essay examines Ottoman attitudes towards translation and cultural mediation in the era of imperial expansion and consolidation in the sixteenth
century, especially during the reign of Sleyman (). It will engage literature on circulation of information, texts, objects and people in the early modern
Mediterranean through the careers and writings of three sixteenth-century Ottoman
imperial interpreters: Yunus Bey (d. ), Mahmud Bey (d. ) and Murad Bey
(d. late s). All three were converts to Islam who engaged not only in translation
between languages and cultures but also in the project of translatio imperii central
to the articulation of Ottoman imperial ideology and the sultans claim to inherit
the universal sovereignty of the Roman emperors. These men were true Renaissance go-betweens, whose written works were informed by intellectual developments in Europe and synthesized a classical cultural and historical heritage of both
Muslim and non-Muslim provenance into unique visions of Ottoman imperial

Of translation and empire


Translation in the early modern Ottoman period has become only recently a legitimate
subject of study, not only for historians but also for literary scholars and philologists.4
Although it has long been recognized that the process of translation from Arabic and
Persian into Turkish was central to the development of Ottoman Turkish language
and literature between the fourteenth and seventeenth centuries, the social mechanisms and strategies that informed this process are only now being investigated. Borrowing methodological and theoretical tools from translation studies, several recent
works on translation in late medieval Anatolia and the early modern Ottoman empire
have focused on the translators from Persian into Ottoman Turkish, their motives
for undertaking a translation project and their approach to and understanding of the
process itself.5
As Gottfried Hagen has demonstrated in his study of Ottoman translations from
Persian in the sixteenth and seventeenth centuries, the term translation (understood
as either word-for-word or sense-for-sense) does not fully capture the nature of the
Ottoman texts labelled tercme. Although in modern Turkish tercme (and eviri)
denotes translation, Ottoman tercmans (i.e., those undertaking the tercme) were
not neutral mediators of meaning who disappeared behind the text, which is an understanding of translators work now discredited by modern translation scholars as utopian. Rather, they made decisive interventions into the text, not only selectively adding to, omitting from, and occasionally changing the meaning of the original but also
often replacing the authors preface with their own.6 This practice led one scholar to
suggest that Ottoman attitudes towards translation were culturally bound.7 However,
were they indeed so unique?
Reflecting on practices of translation in early modern Europe, Peter Burke has
recently argued that they varied considerably more than general theories suggest. . . .
different norms coexisted and competed, so that we may speak of cultures and subcultures of translation. Furthermore, Renaissance culture was characterized by a
remarkably free style of translation that included the liberty to add material, to shift
the action of foreign plays and stories to another locale and thus make the translation
relevant to the local audience, or even to be so creative that the resulting translations
would be more precisely described as tradaptations.8 Since translations from and into
Latin (and Greek) were central to Renaissance learning, translation strategies were
hotly debated among humanists.9
The diversity of translation practices is also striking in the Ottoman case, especially
in the sixteenth century. As in early modern Europe, most Ottoman translators were
not full-time, paid professionals. They were poets, sufi mystics or imperial administrators, often from the emerging scribal, bureaucratic elite, who engaged in translations from Arabic and Persian for a variety of cultural and political agendas that await
systematic investigation.10 Many derived their social status from their knowledge of
the classical languages of the Islamic tradition, but also from their contribution to the
fashioning of Ottoman Turkish as the language of the imperial chancery, prose and
poetry. Recent research suggests that, starting early in Sleymans reign, Ottoman
Turkish supplanted Arabic and Persian as the language of the Ottoman chancery,
prompting key bureaucrats such as Celalzade Mustafa to refine a new imperial lan

Tijana Krsti

guage, through translation, domestication of Persian and Arabic terms in Ottoman

Turkish, and cultural innovation.11 At the same time, the Ottomans began to use only
Ottoman Turkish in their diplomatic correspondence, abandoning the practice of
granting letters of imperial pledge (ahdnames) in Italian, Greek or even variants of Old
Church Slavonic, depending on the addressee.12 This imposition of the new linguistic
boundaries of imperial identity also demanded a new class of linguistic and cultural
Changes in Ottoman chancery practice were accompanied by the rapid development of divan poetry in Ottoman Turkish. By the mid-sixteenth century the ability to
produce verse in Ottoman Turkish, as opposed to Persian, had become the hallmark
of distinction and, often, membership in the ruling class. The new Ottoman poetic
language entailed an interplay between translation of Persian masters and attempts to
innovate and surpass them.13 Similar developments are observable in Ottoman prose
writing: in the s Mehmed III instructed court historians to write histories only in
Ottoman Turkish, the language of common usage.14 The process by which Ottoman
Turkish became the imperial language also paralleled and enhanced the fashioning of
a distinct visual, architectural, legal and religious Ottoman imperial tradition in the
second half of the sixteenth century.15
The Ottoman imperial language did not consist only of new linguistic and stylistic
features. It also included new ideological concepts that needed to be heralded to both
domestic and international audiences. While bureaucratic and literary elites promoted
these concepts on the domestic level, the imperial dragomans (from Arabic tarjuman)
conveyed the Ottoman imperial agenda abroad. Their existence as a professional group
was the outcome both of the process of bureaucratization inaugurated by Sleyman
and of the Ottomans increasingly dynamic diplomatic contacts with European polities in the Mediterranean.16 Although the Ottomans had used translators since the fifteenth century, it is only in Sleymans era that a diplomatic corps albeit selected in
an ad hoc manner with a discernible hierarchy begins to emerge, subject to the office
of the chief secretary (reislkttab). While sixteenth-century imperial dragomans, all
of whom were converts to Islam with non-Turkish-speaking backgrounds, generally
never attained the literary sophistication in Ottoman Turkish of their contemporaries
from the scribal class, they were nonetheless acutely aware of the importance of language and translation in projecting the legitimacy of Ottoman imperial and religious
authority. Moreover, by drawing on their own cultural and linguistic backgrounds,
through their written works they also brought into the Ottoman empire European
humanist attitudes towards texts and translations, as well as Renaissance discourses of
legitimacy. What follows is a closer look at the career and writings of three sixteenthcentury imperial interpreters, who are well known for their diplomatic roles but less so
for their literary production and their understanding of the translation process.

Yunus Bey (d. ) began his work as an Ottoman envoy in as a young
cavalryman. In he became chief dragoman (ba tercman), in which capacity he
served at least until .17 Originally a Venetian subject from the Peloponnese port
town of Modon, the son of a certain Zorzi Taroniti, Yunus Bey was captured by Ottomans as a young man around and subsequently converted to Islam. During his

Of translation and empire

career, which is well documented in both Ottoman and western sources, he undertook
numerous diplomatic missions to European polities, although he is best known for his
six missions to Venice. He spoke Italian and Greek fluently and was also proficient in
In , before the outbreak of the OttomanVenetian war, Yunus Bey carried
to Venice the request for an alliance with the Ottomans and French against Charles
V. On this occasion the Venetians published a short pamphlet summarizing his oration in Italian before the Signoria as well as the Senates response, probably to inform
Europe of their intentions.18 However, in the same year, they also published a pamphlet in Italian supposedly co-authored by Yunus Bey and Alvise Gritti (d. ), the
illegitimate son of the Venetian doge Andrea Gritti, who had had an influential career
in Sleymans service.19 This pamphlet detailed the organization and administration
of the sultans domains, his palace, and his armies in the battlefield. Particularly striking is the long list of administrative offices given in Ottoman Turkish, along with an
explanation of what the holders of these positions did and what stipend they received.
In addition to mapping out the administrative structure of the sultans palace and his
impressively vast domains, the goal of this work seemed to be to introduce the Ottomans on their own terms and in their own language. Audiences in Italy, especially in
diplomatic circles, must have already been used to the names and titles of the principal
Ottoman administrators, such as vezir, bassa, defterdar or beglerbei, which were frequent in diplomatic correspondence from Istanbul. However, the pamphlet also introduced less commonly known Ottoman functionaries, such as the bustagibassi (in Ottoman Turkish, bostancba, head of the imperial guards), vechilargibassi (vekilharcba,
palace steward), asgibassi (acba, head cook), capigilar (kapclar, gate keepers), etc.
By insisting on this new vocabulary in Ottoman Turkish, the pamphlet encouraged
the Venetian and other audiences to learn it or ignore it at their own risk.
This pamphlet is also unique because it is the first work by an Ottoman author (or
authors) explicitly intended for readership in Western Europe. Since the invention
of the printing press in the mid-fifteenth century, one of the most frequent topics in
printed religious and secular literature was the rise of the Turk. Yunus Bey contributed to this Renaissance trend with a piece of Turcica literature that, unlike the rest,
presented a positive image of the Ottoman empire. During his six visits to Venice,
he not only explored the city and its culture but also made friendships with Venetian
dignitaries, one of whom, Giovanni Francesco Mocenigo, even hosted Yunus Bey in
his house.20 His exposure to Renaissance culture was not, however, limited to Venice:
contemporary Istanbul, the Mediterranean hub bustling with merchant, diplomatic
and cultural activity, was also developing into a Renaissance imperial city interested in
its own classical imperial past.21
Although we know little about Yunus Beys understanding of the translation process, this pamphlet, together with details from his biography, suggests that he used
his status as a cultural intermediary to promote the cause of the Ottoman empire and
Islam among the Ottomans rivals. The pamphlets overall message of the empires
strength, majesty and ability to attract and incorporate non-Muslims is reinforced in
the authors signature at the end of the work: the readers learn that Yunus Bey, who
was Greek and now is Turk & the great interpreter of the Signor & Signor Aluise
Gritti, son of the Doge of Venice, vouch that everything written is true.22 Yunus
Bey died in as a pious Muslim, having commissioned a mosque by the imperial

Tijana Krsti

architect Sinan. This was completed after his death thanks to funds given him for this
particular purpose, in return for help with negotiations in Istanbul, by the Archduke
of Austria, Ferdinand I of Habsburg (Holy Roman Emperor ).23

Another Ottoman imperial interpreter whose interest in translation went beyond
diplomacy was Mahmud Bey (d. ). Born Sebold, the son of Jakob von Pibrach,
a shopkeeper from Vienna,24 he was serving as a page of the Hungarian king Lajos II
when he was captured by the Ottomans in the battle of Mohcs in .25 He had
good Latin in addition to his native German, and by the s had become an imperial interpreter. He participated in important diplomatic missions to Transylvania,
Poland, France, Venice and Prague, and between and his death in was chief
Mahmud is today recognized as the author of a work surviving in a single undated
manuscript, entitled Tarih-i Ungurus (History of Hungary).27 Although first published
in , this work has been studied mostly by Hungarian linguists and has been largely
ignored by historians of the Ottoman empire. It is nevertheless a fascinating source for
study of an Ottoman translators practices in the Renaissance era. It reveals the outlook of a man deeply integrated into the Ottoman imperial enterprise and informed
by contemporary Ottoman diplomatic and ideological issues, as well as by the Renaissance culture of translation.
In the introduction to this history, Mahmud says:
When several among the fortresses in the region of Hungary that the sultan had
conquered with the strike of his sword chose to obey the king of Vienna and
ignore the padiahs command, he came upon them with an enormous army and
subjected them with a single strike. In a fortress named Ustulni Belgrad [Szkesfehrvr] a book in Latin was found, and upon inspection it was established that
it was a history of Hungary from the olden days telling how that region was ruled,
why it was called Hungary and why its capital Buda was called Buda and what the
names of the old capitals were, and what kings came and went, when and whom
they fought, how long they lived and ruled. For this reason I, the humble Dragoman Mahmud, decided to translate (tercme) it. Perhaps the time will come when
the sultan of this time and the world-conqueror of this era will bestow his exalted
look upon this poor ones gift and endow his slave with favor and mercy.28
What the reader expects is a history of Hungary, but that is not exactly what follows.
Instead, after a short exposition on the ancestors of the Hungarians, Mahmud segues into the story of Alexander the Great, which explains why the work is subtitled
Iskendername (Epic of Alexander).
Although Alexander the Great was the subject of many medieval romances in a
variety of languages in both Europe and the Muslim world, in the Renaissance era
the Alexander epic experienced an unprecedented global boom in popularity, from
the British Isles to Indonesia.29 Besides being an epic warrior, Alexander embodied
the notion of a universal world emperor revived by Renaissance politics and looming
large in the mind of many a ruler in the early sixteenth century. To early Ottomans,

Of translation and empire

Alexanders exploits were known through Firdawsis Persian epic the ahname and
Nizamis Sikandarnama, as well as through the earliest surviving Ottoman historical
work, Ahmedis Iskendername (c.). It is well known that Mehmed the Conqueror
saw himself in the image of Alexander and that his palace library owned the standard
biography of Alexander, Arrians Anabasis, copied for the sultans collection in the
s. In fact, comparison of the Conqueror with Alexander the Great was the leitmotif of Kritobouloss History of Mehmed the Conqueror, composed in Greek for the
sultan at about the same time.30 In the early sixteenth century the figure of Alexander/
Iskender, or Dhul-Qarnayn (the two-horned one), acquired further connotations
in the Ottoman context. In addition to being a world conqueror, he was a messianic
figure, and with the beginning of the tenth century of the Islamic calendar (
ah/ ad) messianic hopes began to centre upon the Ottoman sultan, especially
Sleyman, who consciously promoted this image of himself.31 Furthermore, in the
early sixteenth century Alexanders exploits against the Persians acquired new meaning
in the Ottoman context with the rise of the Safavid threat.
What was Mahmud Beys source on Alexander the Great and why bring him into
an Ottoman account of Hungarian history? The Hungarian classicist Istvn Borzsk
has identified Mahmuds key source as Marcus Iunianus Iustinuss abstract of the Historiae Philippicae by Pompeius Trogus.32 While staying close to this source, Mahmud
also provides important cultural glosses, such as referring to Darius as Darab Shah, a
name that would be more familiar to the readers of the ahname. He further embellishes the rhetoric of Iskenders letters to Darab Shah, mentioned in both the ahnama
and Iustinus history, in which Iskender proclaims himself the universal ruler of his
era (sahib-kran- zaman), the ruler of the seven climes (padiah- heft iklim) and the
ruler of the Rum (Rum padiah), all titles which Sleyman also claimed for himself.33
Sixteenth-century historians and diplomats were well aware of the exchange of letters
between Sleyman and the Safavid shah Tahmasp preceding and following the Treaty
of Amasya in . In his Tabakatl-memalik ve derecatl-mesalik, Celalzade Mustafa
cites these letters in full. In one letter Tahmasp pays homage to Sleyman as a world
conqueror in the mould of the legendary Persian rulers Darius, Cemid and Feridun,
but most of all as an Alexander the Great of his time.34 Extant analyses of Mahmud
Beys history of Hungary tend to overlook the prominent underlying theme of Alexanders enmity with Darius and the constant shifting of the geographic emphasis in
the work from the border with the Persians to Alexanders fictive conquests in Europe.
This perspective not only captures well the Ottoman predicament of constantly having to monitor borders in both east and west but also suggests that the text was written
in the mid-s, during or after Sleymans campaign against Tahmasp.
The rest of the Tarih-i Ungurus, about two-thirds of the text, blends in the story
of Alexanders conquests with the establishment of Hungary and its topsy-turvy royal
history. Mahmuds reference to the translation of an old medieval source in Latin
becomes relevant here, as he evidently uses at least one medieval Hungarian chronicle
and probably several such as the fourteenth-century Chronicon pictum or Thurczy
Jnoss fifteenth-century Chronica Hungarorum.35 In this context the mission of the
work comes to the fore: Mahmud represents Hungary as the conquest symbolic of
world domination and a necessary stepping stone towards the conquest of Vienna,
which he characterizes as kzl elma (the golden apple), the last Ottoman conquest in
Christian lands before the End of Days. In bringing together the history of Hungary

Tijana Krsti

with the history of Alexander, Mahmud Bey produces the ultimate blueprint of the
Ottoman imperial agenda. He represents the conquest of Pannonia, where Hungary
was to be established, as of equal importance for Alexander the Greats plan to rule
the world as his conquest of Persia. This proposition was by no means foreign to the
contemporary Ottoman ruling elite. Since the fifteenth century, the conquest of Hungary had been a highly coveted prize for the Ottoman sultans and was described as
such in contemporary chronicles.36 Both after Sleyman conquered Buda in and
after he established direct rule in central Hungary in the early s, the conquest was
described by Ottoman historians as one of Alexandrine proportions.37
Through his work as an imperial dragoman, Mahmud Bey conveyed the Ottoman
agenda to European polities; as an author he also brought European Renaissance sensibilities to the Ottoman empire, synthesizing them with Hungarian history and Islamic
traditions. By weaving together the classical themes appropriated by both Ottomans
and Hungarians, he produced a work that views Sleymans conquest not only as the
culmination of Hungarian history but as the point of convergence of imperial ideas,
past and present. It is unlikely that this work ever reached Sleyman or achieved any
significant circulation. Its style and composition leave much to be desired compared
with contemporary Ottoman histories. Nevertheless, Tarih-i Ungurus is a fascinating example of one Ottomans understanding of translation and its relationship to
imperial legitimacy.

Among sixteenth-century dragomans, Murad Bey (d. late s) was the most prolific
writer of independent works of poetry and prose. In his own testimony, Murad was
captured at the battle of Mohcs in when he was seventeen years old. Originally
named Balzs Somlyai, he hailed from Nagybnya (today Baia Mare in northern Transylvania, Romania).38 According to Stefan Gerlach, Lutheran chaplain to the Habsburg
envoy to Constantinople from to , who befriended him while he was a
dragoman to the embassy, Murad had once studied in Vienna. He became an imperial
interpreter around under the patronage of the grand vezir Rstem Paa, having
already taken part in a diplomatic mission, probably together with Mahmud Bey, to
Transylvania, where he was captured and spent thirty months in captivity. Murad
later wrote that this episode prompted him to write several of his polemical works and
to reflect on the importance of translation for religious polemics. The sources suggest
that, in addition to Ottoman Turkish, Murad knew Hungarian, Latin and probably
German, although it is difficult to assesses his facility with Arabic and Persian.
In his long career, from the mid-s to the mid-s, Murad produced several
fascinating works which he then translated into different languages. In , he
wrote a polemical treatise entitled Guide for ones turning towards God (Kitab tesviyetttevecch ilal hak), in which he argued for Islams superiority over previous revelations.
In he translated Ciceros De senectute into Ottoman Turkish; between and
he translated his earlier religious treatise into Latin; and between and
he composed in Ottoman Turkish a number of religious hymns on the unity of God,
with parallel translations in Hungarian and Latin. Finally, as an elderly man dismissed
from imperial service as a result of his immoderate enjoyment of wine, Murad met
Philip Haniwald from Eckersdorf, and in return for a small fee translated into Latin an

Of translation and empire

Ottoman source (or sources), among which was Mehmed Neris Ottoman chronicle.
This translation became central to the Codex Hanivaldanus.39
As with the works of Yunus Bey and Mahmud Bey, Murad Beys opus has been
known to scholars for almost a century, but has been considered mostly by linguists
and literary scholars. I have discussed elsewhere Murads religious treatise and the
narrative of his own conversion appended to it, which represents the earliest known
Ottoman self-narrative of conversion to Islam.40 This and other Ottoman conversion narratives composed by former Christians familiar with religious debates in postReformation Christendom represented more than cultural translation of a Christian
genre to an Islamic context. Rather, they were an expression of Ottoman participation
in the religio-political developments of the age of confessionalization. In this treatise
Murad also makes intriguing observations about the process of translation, echoing
the debates regarding the translation of the scriptures and classical works into vernaculars then raging among humanists in late Renaissance Europe.
In his introduction Murad declares the argument of Jews and Christians, that they
should adhere to their ancestral faith because their books came down from God before
the Koran, to be invalid, because it is possible that their ancestors corrupted and altered
(tag.yir) the scriptures through endless translation from one language to another (drl
drl dillere dndermi).41 Although this argument draws on well-established antiChristian and anti-Jewish polemics, Murad introduces a personal note.
When this poor one engaged in a religious dispute with some learned Christian
men in the Infidels Land [Kafiristan] and discussed the fact that these books were
like that [i.e., altered through translation], the book of Psalms was before us, in
both Latin and Hungarian, which was translated from Latin. Suddenly, I spotted
a place in the text where instead of one word another one was chosen that considerably changed the meaning of the passage. Since I knew what word should come
in that place I said: Should not the meaning of this be this? at which point [they
were] struck dumb and could not respond.42
In addition to criticizing the translation of Christian and Jewish scriptures, Murad is
critical about the attempts of some Christians to translate the Koran upon learning
a few words of Arabic, giving rise to a work full of omissions and blasphemies.43 He
concludes that none of this is surprising, since for every word in one language there
are many synonyms (maani-yi mtereke) so that to understand it when it is translated
into the final language, one cannot help but infer a different meaning.44 Hence one
must read the scriptures in the language in which they came to mankind. He implies
that the Korans untranslatability is the foundation of Islams superiority and, by extension, of the Ottomans supreme religio-political authority over their rivals.45 Murad
argues that, in addition to corrupting the scriptures, the Christian popes and cardinals
became morally corrupt, which caused them to lose the keys to heaven originally given
by Jesus to Peter. With the arrival of Muhammad, these keys were transferred to the
Muslim community and eventually came down to the Ottoman sultan, the caliph of
all Muslims and the messianic world conqueror.46
However, while criticizing the translation of scripture, Murad is acutely aware, as
a professional interpreter, of the importance of translation for communication of his
own religious and political message. Although self-effacing about his knowledge of

Tijana Krsti

both Turkish and Latin, he writes that he decided to translate (tercme) his treatise
into Latin so that the Christians of Hungary, Germany, Poland, Czechia, France,
Portugal and Spain could learn of Islams potential and become softened to its message.47 The resulting work is a fascinating bilingual treatise, with the main Ottoman
Turkish text surrounded by the Latin translation in the margins. A similar effect is
also produced by his hymns, presented in page-by-page parallel Hungarian, Latin and
Ottoman versions.48 Both the hymns and the treatise reveal a man willing to explore
his Christian past through his Muslim present: in the treatise he integrates reformed
Christian sensibilities with a distinct sufi outlook, while in the hymns he seems close
to espousing Unitarianism, whose proximity to Islam did not escape his Christian
contemporaries. He may have translated the hymns into Latin and Hungarian in order
to make Islam more familiar to audiences in Transylvania, where Unitarianism was
becoming one of the state religions in the later sixteenth century.
Murads attitude towards linguistic and cultural mediation and his attempts to
make the Ottomans an integral part of a broader Renaissance culture appear also in
his In praise of old age (Kitab der medh-i piri), a translation or paraphrase in Ottoman Turkish of Ciceros De senectute, commissioned by the Venetian bailo Marino
di Cavalli as a present for Sleyman.49 Murads introduction describes the work as a
translation of a text in Firenk dili (literally, the language of the Franks here Italian)
which di Cavalli had inherited from his late maternal grandfather, Andrea Foscolo.
It records a conversation which supposedly took place in Edirne between Murad II
() and his son, later Mehmed II, about the conditions and merits of old age.
The text shows a doting and respectful son who hopes to succeed his father not only as
a ruler but as a prudent man as well, a potentially painful and risky topic to bring up
with an aging Sleyman, who not only had one of his sons (Mustafa) executed in
but faced the treason of another (Bayezid), who fled to Iran in and was shortly
afterwards also executed.
According to Marino di Cavallis introduction to the text as translated by Murad,
after his appointment as bailo in Byzantine Constantinople, Andrea Foscolo visited
Murad II in Edirne, where he was hosted with great honour. Marino di Cavalli does
not say explicitly whether Andrea Foscolo himself recorded this dialogue, but states
that: upon finding this [he, di Cavalli] did not see it fit that this beautiful gem should
remain concealed, and [he] wanted to have it translated again from Firenk dili into
Turkish. Di Cavalli also deemed it proper that the symbols and precious sayings
demonstrating the eminent perceptiveness of the glorious ancestors of the felicitous
padiah of the world from the House of Osman should become the inheritance of the
felicitous padiah. Furthermore, he finds the text important
because it makes it known to the entire world that those glorious ancestors [of the
padiah] who came and went did not attain fame only by bravery, sword, strength,
might, majesty and justice but rather also by learning, refinement, fairness, excellence and prudence.
Finally, Marino di Cavalli declares his intention of having the text translated and published in Latin, in order to make it available to different Christian communities not
only in the land of the Franks (Firengistan) but also in Germany, Poland, Czechia,
and Hungary.50

Of translation and empire

The following dialogue is indeed modelled upon Ciceros De senectute. In both

texts, the conversation begins with the younger man admiring the older ones vitality
and presence of mind, asking about the secrets of a long and prosperous life and the
merits of old age. Although the text again draws directly on De senectute in the last
section, which explains why people dislike old age, the entire middle section, in which
Murad II reflects on why old age is better than infancy (alem-i tufuliyyet), up to the age
of nine, adolescence (alem-i sabavet), from nine to fifteen, youth (mertebe-yi ebabet),
from fifteen to thirty, and mature age (alem-i ruculiyyet), from thirty to forty-five, is
not from Cicero.
At first glance, this work appears to be one of those free translations, common
in the Renaissance, that relocate the action of a play to another place and introduce
local actors with whom the translators audience can easily identify. Murad II and
Mehmed II, and others mentioned in the text, such as Evrenos Bey, Bayezid I, Alexander/Iskender and Anuirvan, are key to that effect. However, cultural translation goes
well beyond mere substitution of names. On at least two occasions Murad II illustrates his point through stories attributed to Lokman, the Koranic paragon of wisdom.
There are also references to the Koran and God throughout the text, as well as to the
need to submit to ones destiny. As in Kitab tesviyett-tevecch, Murad Bey finds an
easy affinity between the stoic and sufi approaches to life and ethics. The text espouses
the effort to decrease worldly attachments, limit the appetite for food or sex, and bring
dissident elements into harmony. Sufi pietism and stoics shared the view that the ego
can be ones worst enemy if not properly checked, a theme which pervades much of
what Murad II tells his son and gives the dialogue a sufi veneer.
What exactly was Murad Beys function in producing this text? Although the introduction refers to an original text which he translated, the story about Andrea Foscolos manuscript was possibly designed to elevate di Cavallis position by creating a venerable historical bond between his and the sultans ancestors. Such fictional introductions and references to ancient texts newly discovered at secret or unexpected locations
were commonplace in the early modern period.51 Moreover, none of Murads contemporaries indicate that he knew Italian. Instead, it is possible that he was presented with
Ciceros dialogue in Latin and asked to produce an appropriate Ottoman version of it.
The work would then parallel the Renaissance imitatio, namely the attempt to write
Latin poetry and prose using the language, style and topics of classical Roman models,
of which Cicero was the most illustrious. Commenting on the spirited debate regarding the issues of translatio and imitatio, and the question of how slavishly one should
imitate Roman models, Erasmus approved of imitation, not the kind drawn from one
model from whose lines it will not dare to depart, but rather the kind that takes the
best from the writings of several authors and results in a work that does not appear to
be a mosaic of Cicero, but the child born of [ones] brain.52 In light of Murad Beys
other work, especially the polemical treatise, it is quite possible that Kitab der medh-i
piri was the child of his brain.

Few copies of the works by Ottoman imperial interpreters discussed above are extant
today, suggesting that their circulation and audience was limited. While the impact
of their work on Ottoman audiences may have been minimal or non-existent, they

Tijana Krsti

were somewhat more successful in engaging a foreign readership, especially the diplomatic and scholarly community. Yunus Beys pamphlet went through two editions in
Venice, and both Murads and Mahmuds works are found today in libraries outside
Istanbul, purchased by foreign diplomats and travellers for collections in Vienna, London and Budapest. Sixteenth-century sources show that all three men interacted with
contemporary humanists from Padua, Venice, Heidelberg and other academic circles
who hoped to recover classical texts during their sojourn in Ottoman Istanbul, a city
they often viewed as an overlay on Roman and Byzantine Constantinople containing
hidden treasures relating to classical and early Christian history. In other words, the
cultural and linguistic mediation in which these interpreters engaged was not limited
to their diplomatic and written work; they also participated in a larger international
network of diplomats, travellers, merchants and scholars who converged on Istanbul
and exchanged information, texts and objects which were crucial in shaping the discourse on the Ottomans abroad and on the rise of Oriental studies in Europe.
Yunus Bey, Mahmud Bey and Murad Bey wrote when the Ottoman imperial image
was being crafted. Their works are period pieces of Sleymans era and its experimentation with non-Muslim modes of legitimization in creating the sultans image as a
universal sovereign.53 The openness to non-Muslim elements in imperial discourse, in
which all three interpreters participated, subsided towards the end of Sleymans reign
and in its aftermath, as sultans began to fashion themselves more as Sunni caliphs,
universal rulers of the Islamic community. Although the exchange of religious, literary
and scientific knowledge with the west through translation continued throughout the
seventeenth century, there are no similar attempts at seamless synthesis of disparate
cultural, historical and religious elements for the glory of the Ottoman sultanate.
However, these men were not the only early modern Ottoman go-betweens engaged
in cultural translation. Jewish medical doctors from Spain, Greek scholars educated in
Padua and Rome, converts to Islam in various branches of Ottoman bureaucracy but
also in religious circles, as well as scholars from Arabic- and Persian-speaking Ottoman
provinces, all participated in different translation projects within the empire. Some of
these projects were commissioned by the sultan; others were backed privately, by men
of science and scholarship or even foreign missionaries. In the early s, the Ottoman
polymath Katip elebi engaged a French convert to Islam-turned sufi eyh to translate
the Atlas minor from Latin into Ottoman Turkish.54 In the early s, Isaac Besire, the
Anglo-Catholic chaplain of the English community in Istanbul, and the Dutch scholar
Levinus Warner commissioned Ali Ufki Bey, who later became Ottoman chief dragoman, to translate, respectively, the Anglican catechism and the Bible with the Apocrypha into Ottoman Turkish.55 A closer study of the specific international social networks
that converged in Istanbul and the contexts which produced various types of translation promises to reveal previously obscured channels of intellectual exchange that both
bridged and helped produce the notion of difference between East and West.


Casale ; MacLean ; Andrews and Kalpakl ; Necipolu ; Krsti .

E.g., Bennassar and Bennassar ; Garca-Arenal and Wiegers ; Dursteler ; Davis
; Rothman .
Pym ; Rothman .

Of translation and empire

The absence of studies on translation before the extensive literary translations from Western
European languages in the late Ottoman era can be credited to the ideology and methodology
of early Republican Turkish scholars, who tended to study pre-Ottoman and Ottoman Turkish in terms of contamination by Persian and Arabic influences. The Ottoman Turkish of
divan poets was deemed an artificial construct that detracted from the pure Turkish of folk
poets such as Yunus Emre (see Paker : ).
Pistor-Hatam ; Toska ; Hagen .
Hagen : , .
Paker : .
Burke : ; Burke .
Pym : .
Hagen ; Fleischer b; Woodhead ; Fleischer a: .
Ylmaz : .
Ibid.: .
Kim : ; Andrews : .
Woodhead : ; Fleischer a: . Producing a new, imperial literary language
was not an enterprise unique to the Ottomans. In contemporary Spain, Castilian became the
imperial language (Pym : ).
Necipolu .
Veinstein .
Matuz : ; Bacqu-Grammont .
Pedani b: .
The pamphlet was published in Lybyer [] : .
Pedani b: .
For the efforts of Sleyman and his grand vezir Ibrahim Paa to incorporate the Hippodrome,
the old ceremonial core of Byzantine Constantinople, into Ottoman imperial ritual, see Turan
: I, .
Lybyer [] : .
Bacqu-Grammont : ; Necipolu : .
Petritsch : ; cs : .
cs : .
Matuz : .
The manuscript is in the Oriental manuscript collection at the Library of the Hungarian Academy of Sciences. For a transliteration and facsimile in Ottoman Turkish, see Hazai ; for
a translation into Hungarian with a critical introduction, see Hazai .
Hazai : I, .
Ng .
Raby : .
Fleischer : .
Borzsk : ; Hazai : .
Hazai : I, ; on Rum, see Kafadar .
Kappert : ba.
Hazai : .
Fodor b: ; Turan : .
Fodor b: .
cs : ; Babinger ; Matuz : .
Mnage : .
Krsti .
Murad b. Abdullah, Kitb tesviyett-tevecch ill-hak, British Library, Add. MS , b.
Ibid., a.
Ibid., ba.
Ibid., b: Ve bir dilde olan lafzda nice mani-yi mtereke anlansa hr dile dnderildkde
bir manadan g.ayrya istidllna re yokdur.

Tijana Krsti
According to Islamic tradition, since Gods words, as recorded in the Koran, were revealed to
Muhammad in Arabic, the Koran cannot be translated into another language without altering
Gods words.
Murad b. Abdullah, Kitb tesviyett-tevecch ill-hak, British Library, Add. MS ,
Ibid., b.
Babinger : , for a facsimile of Murads hymns, and , for transcription of the
Ottoman text.
Ettore Rossi published a short study on this unique manuscript together with a critical edition
of the text in Ottoman Turkish and its translation into Italian (: ). The original
manuscript is in the Nurosmaniye Library, Istanbul.
Rossi : .
Wiegers .
Endres : .
Necipolu .
Gnergun : .
Neudecker .



Christine Woodhead

he Ottoman Turkish language was a product of empire, a consciously developed

political and cultural tool. By around formal, written Ottoman had evolved
from its base in the colloquial Turkish of Anatolia into a prestige language dominated
by elements from Persian, the inherited language of early administration and literature, and from Arabic, the first language of religion and scholarship. This amalgam
was considered a natural and appropriate reflection of Ottoman imperial status in
relation to the Islamic cultural heritage, appearing in varying degrees of complexity
in both chancery documents and literary works. By the nineteenth century, however,
there was increasing criticism, particularly among reformers in Istanbul, of this official
language as an unnecessarily complex and artificial hybrid, understandable only with
constant reference to dictionaries. Ottoman Turkish appeared to many as symbolic
of an inward-looking, complacent conservatism responsible for late Ottoman decline,
and as a barrier to political and social reform. In the post-imperial, nationalist era of the
s and s, this language was, by definition, redolent of a failed political entity
and had few supporters. Official use of Ottoman Turkish came to an abrupt end in
November , when the Republic of Turkey adopted a specially devised alphabet
in Latin characters to replace the Arabic script in which Turkish had been written for
almost a millennium. Atatrks language reform movement then proceeded to purge
from the written language most foreign Arabic and Persian elements, aiming to produce a vocabulary and grammar as purely Turkish as possible. By the mid-twentieth
century, less than fifty years after the end of the empire, Ottoman had become effectively a dead language, its literary and historical works rendered largely inaccessible
and alien to subsequent generations of Turkish speakers. So rapidly did the political
and cultural outlook change that even the language of Nutuk, Atatrks definitive
address to the Turkish Grand National Assembly given over five days in October
, soon had to be simplified and modernized for most readers.
Why Ottoman Turkish assumed the form it did in the earlier period, in what ways
and how widely it was used, and to what extent its use promoted or prevented the
spread of a specifically Ottoman literary culture are some of the questions to be raised
in this essay. There will necessarily be more questions than answers. The definitive

Christine Woodhead

break between Ottoman and modern Turkish, and negative views of the later historical language and its written output, long discouraged serious study of it and exacerbated the historians usual problem of acquiring appropriate cultural literacy.1 The
tendency still lingers to consider Ottoman implicitly as an artificial idiom in contrast
to ordinary Turkish, a view subconsciously reinforced for most western scholars by
the usual practice of learning modern Turkish first and then moving backwards into
an Ottoman language which initially appears to be more akin to Arabic and Persian
than to Turkish. Among the points to be raised below is the potential insight to be
gained by a different approach, that of comparing the style and use of Ottoman not
with Turkish modern or otherwise but with other prestige languages in the early
modern world.
Among many other factors rendering the study of Ottoman language and literary
culture difficult is the mere range and intimidating number of manuscripts and documents which survive from at least the period after , the majority of which remain
unstudied. This is despite the efforts of Turkish scholars over the past half century,
particularly in producing critical editions of the divans (anthologies) of individual
poets.2 Equally, over the past twenty years Mehmet pirli and other students of Bekir
Ktkolu in Istanbul have led the way in producing reliable editions of major sixteenth- and seventeenth-century Ottoman histories.3 Nevertheless, much groundwork
remains to be done. One particular problem is that, in some historiography, chancery
documentation and other types of prose writing, the Ottoman Turkish language can
indeed be a barrier to understanding, particularly in its more elevated registers. Sweeping judgements about tired idioms, overblown rhetorical prose and meaningless verbiage are easy to make, and sometimes justified. Not all writers wrote well; bureaucratic
language often was jargon.
For non-Turkish scholars there have also been other discouragements. From the fifteenth century onwards, constant hostile references in western literature to the Turk
imposing an alien and barbaric culture upon Christian peoples established a lurking,
unappealing stereotype. Nineteenth- and early twentieth-century western Orientalists
studied principally Arabic and Persian languages and literatures, while philologists
classified Turkish as a Ural-Altaic language not native to the Middle East and, by
implication, of less significance. In the trio of major Islamic civilizations, Ottoman
came a poor third, a balance still reflected today in most departments of Middle or
Near Eastern studies in western universities. Added to this bias is the fact that, for
political, nationalist reasons, much twentieth-century historical writing in the empires
Arab and Balkan successor states inevitably took an anti-Ottoman stance.4
Assessing the role played by the Ottoman Turkish language and written culture in
the empire as a whole is therefore not easy, and few attempts have been made to do so.
Yet, if language was a tool of empire, how successfully did it function as such? What
was the balance between practical communication and awe-inspiring propaganda?
Was the latter dominant, and were the Ottomans seen as other because of their
mode of expression? It is easy to assume that this was so. Yet, the modern desire for
straightforward clarity in government communications was not necessarily matched
in the pre-modern era in any culture, particularly in public announcements of military
victory or of monarchical largesse, which required an appropriately grand tone. In
return, for writers seeking court patronage, striking metaphors and inspiring allusions
were usually more effective than unadorned prose. Andreas Tietzes studies of works

Ottoman languages

by the sixteenth-century historian Mustafa Ali (d. ) offer the best guide to the
consciously artistic Ottoman Turkish prose style known as ina (literally, creating,
construction), with its parallel, rhymed and sonorous phrasing.5 Yet, not all Ottoman Turkish prose was deliberately complex or rhetorical; the extreme has perhaps
too often been taken as the norm. Major histories such as those by Selaniki (d. )
or Hasan Beyzade (d. c.) were composed in an educated but relatively clear register of written Ottoman, and were not court-centred commissions. Koi Bey (d. after
) and other seventeenth-century writers of advice literature generally chose clarity over style, even when addressing the sultan. As will be shown below, other writers
naturally adjusted their style to suit their intended readership and purpose.
Equally, while composition of the most demanding forms of written Ottoman prose
was necessarily limited to very learned stylists and highly trained chancery officials
the literary and bureaucratic elites it is worth considering how far down and across
the social and educational scales such texts and documents might have been read or
heard, and thereby in varying degrees understood. Ability to compose in a learned and
elegant style is one thing; ability to appreciate it is another, and casts the net much
wider. If modern historians working in a cultural vacuum can attempt to master complex Ottoman, a significant proportion of literate Ottoman subjects native speakers
of Turkish or otherwise could surely have done so too, if inclined, and mostly with
much greater success. Indeed, while western sovereigns might have been impressed by
the imagery, cadences and self-confidence of the Latin or Italian translations of Ottoman imperial letters, the true force and message of the Ottoman language would have
been even more apparent within the empire, among those who could read, half-read or
hear the originals for themselves. Such domestic recipients included Ottoman vassals
such as the khans of the Crimea and the voyvodes of Wallachia and Moldavia, provincial governors and high-ranking judges throughout the empire, and leading scholars
and intellectuals to whom the sultans sent gifts and letters of appointment. During
the sixteenth and seventeenth centuries, particular significance would also have been
attached to imperial correspondence with the Ottomans perceived rivals in eastern
Anatolia and Iraq, the Safavid shahs of Iran. Not only the Ottoman literary language,
but also miniature painting and other arts of the book were thoroughly permeated
by Persian influences. Both craftsmen and artefacts were among the war booty taken
from Tabriz by Selim I in and again by Sleyman in , resulting in several
decades of intense cultural oneupmanship on the Ottoman part and culminating in
the monumental illuminated manuscripts produced by court painters and litterateurs
under Murad III (). An imperial language which incorporated dominant elements of both Arabic and Persian and, by implication, of crucial aspects of these classical Islamic traditions carried a clear stamp of political authority and made a cultural
statement recognizable both within and outside the empire. Without such elements,
how seriously would it have been taken?
However, until the nineteenth century, Ottomans referred to their sophisticated written language simply as Turkish, which implies that they did not perceive
unbridgeable gaps between higher and lower language registers, either spoken or written.6 Strictly speaking, it is inaccurate to call the pre- language Ottoman Turkish, although the term remains in use, partly out of habit and partly because it is more
specific in its chronological and political referent than the alternative old Turkish.
Its use is also valid linguistically in discriminating between the imperial language and

Christine Woodhead

those of the millions of other Turkish speakers over the centuries in Azerbayjan, Iran
and Central Asia.
Most studies of what we will continue to call Ottoman Turkish have naturally
focused on its written forms, in particular its literary and chancery use.7 However, it is
worth considering briefly the properties and broader usage of the language itself. For
instance, while it may be impossible to know to what extent the cadences and lexicon
of learned Ottoman were reflected even in educated speech, such an addictive feature
as the Persian izafet (joining) construction must have been difficult to resist and relatively easy to use, particularly for the many Ottoman Muslims who had studied Persian
as a separate language. Izafet constructions produced a word order almost completely
opposite to that of Turkish, but had much simpler grammatical rules. Did this also
make Ottoman Turkish easier for, say, Greek and Serbian speakers to understand
than ordinary Turkish? The ever-increasing use of Arabic words must certainly have
appealed, no doubt partly intentionally, to the empires Arabic speakers. Hence, one
consequence of fostering what to some Turks appeared to be an increasingly foreign
Ottoman language might have been to make it more accessible, one way or another,
to non-Turkish speakers than ordinary Turkish, and therefore more of an asset than a
liability in terms of imperial communication.
This essay considers firstly the diversity of vernacular language use within the
empire, in order to assess the value of spoken Turkish as a medium of communication
and hence as a factor contributing to imperial cohesion, rather than militating against
it. The assumption here is that (Ottoman) Turkish was a language not only of the
cultural and political elite, and that appreciation of its products was not necessarily
confined to a narrow, closed circle. Rather, it should be seen as a practical and flexible
language working in differing registers, spoken and written, to suit the purpose and
the occasion. In view of the once tenacious stereotype of Ottoman rule as an alien
imposition in many non-Turkish parts of the empire, it is worth emphasizing the
obvious that in the pre-modern era language was not considered the divisive marker
of identity and difference that it is today, that Turkish was relatively widely used, and
that Turkish-speaking officials were not necessarily foreign. Secondly, this essay considers the nature of the Ottoman language and its literary culture, not as an unusual,
artificial hybrid but as an imperial idiom comparable to others, particularly in the
Turkish Muslim world. The focus in what follows is on the period up to , before
the combined challenge of print culture, nationalism and conscious modernization.
As with the study of many other aspects of Ottoman history, when seen in comparative perspective, linguistic and literary developments turn out not to be unique to the
Ottoman case but to derive from the essential nature of an extensive, polyglot and
multicultural empire.8 While Turkish will be used here generally for the more informal spoken and written language, and Ottoman Turkish for the formal registers, this
terminology is not clear cut and should be taken within the Ottoman understanding
of a single language spectrum, rather than the post-imperial tendency to perceive divisive differentiation between elite and non-elite language.

It has been estimated that there were around one hundred languages and dialects
spoken within the Ottoman empire, a situation probably comparable relative to size

Ottoman languages

to that within the Habsburg and Romanov empires, and reflected to a lesser extent
in most western states before .9 Until the nineteenth century, only a handful of
Ottoman languages primarily Ottoman Turkish, Arabic, Greek, Armenian, Hebrew
(together with the Judeo-Spanish Ladino) and Church Slavonic were also written
languages. Significantly, all of the latter, except Ottoman Turkish and Ladino, were
long-established liturgical and scholarly languages with more or less fixed forms.
Within each such language a major variation between written and spoken forms was
commonplace, giving rise to a form of diglossia, where the learned language was virtually unintelligible to speakers of its own vernacular.10 Dialectical variation within
a spoken language could also be quite extensive. The principal and most studied
example here is Arabic, which from the s onwards was the mother tongue of perhaps one-third of Ottoman subjects.11 Turkish was therefore not the only language in
the empire, or elsewhere, with a significant degree of internal variation between registers. Nor was the Ottoman government alone in facing communication difficulties
across a variety of languages. To what extent the Ottomans positively encouraged the
use of Ottoman and Turkish as tools of integration, rather than simply of communication, is unknown, probably not the same everywhere, and probably as complex a subject as their attitude to conversion to Islam, to which language use is obviously related.
Native speakers of Turkish were not confined to Anatolia and the empires northeastern provinces. In northern Syria and northern Iraq, Turkish was widely spoken
due to the migratory presence of Trkmen tribes. In Egypt, a Turkish dialect was a
common, if not the first, language in Mamluk military households, becoming a lingua
franca for the eighteenth-century, multi-ethnic Mamluk ruling class of Cairo.12 In
Algiers, the practice of recruiting new Janissaries from Anatolia helped maintain a
working knowledge of Turkish. Speakers of Turkish were found throughout the Balkans, as a result initially of the policy of settling Trkmen and other Turkish-speaking
communities along the principal military and commercial routes, and of the founding
of new towns, and subsequently of conversion to Islam.13 The seventeenth-century
Ottoman traveller Evliya elebi frequently commented on aspects of language in his
Seyahatname, Book of travels, often including brief word lists of lesser-known languages, such as Abhazian (his mothers native tongue) and Kurdish.14 While recording that in Ohrid in north-eastern Macedonia, Bulgarian and Greek were the local
languages, he also noted that they do speak elegant Turkish, and there are some very
urbane and witty gentlemen.15 In Shkder, in north-eastern Albania, they all speak
Albanian, which is like no other tongue and is a delightful language. However, he
also noted that the inhabitants were all Muslim, and that the town possessed a significant number of mosques and medreses (theological colleges).16 Any urban centre in
which a kadi, sancak beyi and other Ottoman officials were based, and where specifically Muslim buildings such as mosques, medreses, tombs or baths were established,
would also possess many local Muslims and local speakers of Turkish.17 Robert Elsies
study of eighteenth-century Albanian aljamiado literature provides further evidence
that some inhabitants of these distant provinces had a more than passing acquaintance
with the Turkish language and Ottoman culture. Aljamiado literature here constitutes
verse written in Albanian in the Arabic script, pervaded with Turkish, Arabic and Persian vocabulary and with the conceits and style of Ottoman lyric poetry.18 Almost the
reverse of this, and demonstrating another aspect of the complex nature of language
use within the empire, is the case of Karamanl i.e., Turkish written in the Greek

Christine Woodhead

Figure . Gazi Hsrev Bey medrese, Sarajevo, completed c.1537 as part of a major vakf foundation built by Gazi Hsrev Bey, Ottoman governor of Bosnia. Photo Grete Howard.

Ottoman languages

alphabet by Turkophone Greek Orthodox Christians in western Anatolia during the

eighteenth and nineteenth centuries.19
Although Ottoman judicial, military and administrative officials conducted their
business officially in Turkish, there was probably little systematic attempt to impose
this language on local communities in the way that early modern European states
increasingly promoted the use of one particular language over others for the ideological
purpose of ultimate political unity.20 Ottoman communication clearly relied heavily
on bilingual intermediaries, drawn both from the subject populations and from among
government officials. For example, much recent research has shown how Christians
regularly had recourse to the judicial court of the Ottoman kadi, the proceedings and
records of which were held in Turkish. A study of eriat courts in seventeenth-century
Cyprus shows the appointment of Turkish-speaking local Greek Christians as official
interpreters.21 More broadly, if the pace of conversion to Islam in the Balkans continued to increase into the eighteenth century, such that by the census Muslims
might constitute anywhere between and per cent of the population of a given
area,22 this must also have involved a significant level of language acquisition and more
people men in particular becoming bilingual to a degree.
For all state employees, however, Turkish was a compulsory language of business.
The first requirements for devirme recruits from the Balkans and Anatolia were conversion to Islam and learning Turkish; the same applied to slaves acquired for government
service through capture or purchase. However, given that devirme recruits were usually
taken at around twelve years of age or more, the majority probably retained a working
knowledge of their original language which could later be used to advantage.23 Many
fifteenth- and sixteenth-century grand vezirs are known to have kept or revived strong
links with their families and home regions. Among the more striking examples is that of
Mehmed IIs grand vezir Mahmud Paa Angelovi (d. ), originally a captive from
a Serbian noble family, who in the late s and s negotiated regularly with his
brother Michael Angelovi, a high-ranking Serbian official, for Serbias incorporation
into the Ottoman empire.24 Lesser Ottoman kul administrators must also have retained
such local links and languages. It is likely, for instance, that the janissary officers and
their scribes appointed to carry out devirme recruitment in the Balkan provinces were
allocated to their region of origin, where they spoke the local language. In the seventeenth century two rival groups emerged in the Ottoman military-administrative establishment recruits of eastern, Caucasian origin and those of western, Albanian or Bosnian origin each characterized by linguistic and local ties.25
In short, a significant proportion of Ottoman military-administrative officials, particularly those posted to the Balkan provinces in the fifteenth and sixteenth centuries, must
have been effectively bilingual. Thus there was not necessarily a language barrier between
Turkish-speaking officials and others. The situation appears subtly different in the Arab
provinces. Although here too Turkish was officially the language of government and the
political elite, in the predominantly Muslim world it was also just one of several local
colloquial languages, all of which were eclipsed in usage and cultural prestige by Arabic.
Many urban speakers of Arabic administrators, merchants, artisans and scholars who
needed to communicate professionally with Ottoman officials or who sought patronage in Istanbul certainly learned Turkish. On the other hand, before the nineteenth
century, relatively few native Arabic speakers were appointed to significant posts in the
central administration in Istanbul, and no devirme and few kul officials appointed from

Christine Woodhead

Istanbul in the sixteenth and seventeenth centuries had Arabic as their native tongue or
family ties with any areas in Syria, Iraq or Egypt.26 The chief kadis of major cities such as
Damascus, Mecca and Cairo were learned Istanbul Ottomans thoroughly competent in
scholarly Arabic but who did not necessarily speak a vernacular form; most of their deputies and probably all their court staff, including translators (here, for the benefit more of
the kadi than the petitioners?), would have been local Arabic speakers.
Hence the nature and degree of linguistic integration in the Arab provinces differed from that in the Balkan provinces. Although Arabic speakers perhaps had a cultural prestige, they were not drawn into the central government and did not have
the same potential for influence upon Ottoman policy as did speakers of Greek and
Balkan languages. However, as Ehud Toledano has shown, the growth during the
seventeenth and eighteenth centuries of Ottoman-local elites and their households
in Arab towns and cities promoted much stronger social and cultural links between,
on the one hand, Janissaries and centrally appointed officials and, on the other, a
variety of local notables. Among such groups, a significant level of everyday bilingualism resulted,27 although to what extent such Ottomanization extended into literary
culture is another matter.28
Despite these variations, it is likely that Turkish was a workable tool of oral communication throughout the greater part of the empire, operating as an essential lingua
franca. The linguistic classification of Turkish as a language of Altaic origin unrelated,
and implicitly alien, to the Indo-European and Semitic languages spoken by Ottoman
subject peoples is irrelevant when examining its practical use among contemporaries.
Loan words for essential items were common in both directions and in virtually all
languages, as any survey of terms particularly for food, dress and other items of material culture would show.29 The use of more abstract Arabic and Persian vocabulary
in the formal written language simply parallels in a different sphere the borrowing of
such practical, everyday terms.
Differences in spoken language, because they were so common and relatively flexible, were not necessarily an obstacle to understanding. However, oral communication
was one thing; reading documents and texts, and participating in learned culture, was
another. Understanding of written language within the empire was obviously limited
in all areas and all cultures to small literate minorities. Literacy rates in any of the
major languages spoken in the Ottoman empire before were probably as variable
at any given time as they were in the major languages of Europe, higher in urban areas
where schools were founded and government officials, merchants and community
leaders recorded transactions on paper, and barely measurable in rural areas, where
to per cent of the population lived a largely paper-free existence. Oral and aural
capacity to transmit and receive information remained dominant; the content of written documents was routinely amplified by a couriers oral message, useful at the time
but not for the later historian.30


In accordance with long-established Muslim practice, the Ottomans recognized the
religious, legal and cultural autonomy of Christian and Jewish confessional communities within the empire, both in the Balkans, where non-Muslims were initially in the
majority, and in Anatolia and the Arab world, where a myriad of ancient Christian

Ottoman languages

churches and small Jewish communities survived.31 Compulsion in either Islamification or Turkification of whole communities would have been counter to the Islamic
governing ethos, unnecessarily antagonistic and impractical in numerical terms.
At a popular level, Christians and Muslims often shared elements of religious practice, both revering the same saint or holy place and participating together in seasonal rituals.32 However, among the literate, religious and therefore literary cultures
remained distinct, and non-Muslim learned traditions were maintained separately
under Ottoman rule. A recent study of Belgrade, a major Ottoman administrative
centre in the sixteenth and seventeenth centuries, indicates that, despite much social
and occupational interaction in this confessionally mixed and relatively cosmopolitan
city, there was nevertheless no hint of intellectual communication between Muslim
and non-Muslim religious communities.33 On the other hand, a study of literature
printed in Istanbul during the nineteenth century suggests that, in this era at least, it
was possible for Turks, Greeks, Armenians and other groups to participate to some
degree in a common literary culture.34 Printing, translation, the inspiration provided
by French and other western models, plus the sophisticated cultural environment
of Istanbul in the tanzimat reform era, must all have influenced this participation.
However, such a development raises questions about where the distinction between
popular and learned literatures really lay in earlier periods, and whether the cultural
dichotomy which appears so clear in Belgrade tells the whole story. Leaving aside the
liturgical and theological aspects, in terms of storytelling, of heroic and edifying tales,
where did shared popular culture end and differentiated learned culture begin? The
Alexander legend, for example, was familiar to sultans and peasants alike, in a variety
of languages and registers, as were the exploits of the Christian Aya Yorgi (St George)
and his alter ego, the Muslim popular hero Hzr Ilyas.35 What role did literate converts play in the transmission of culture? Cross-cultural interpretations such as those
discussed by Tijana Krsti may not have been unusual, albeit on a different level.36
What might the situation have been had Ottoman printing and, with it, more accessible evidence of reading habits been introduced sooner? Further study of pre-
Ottoman reading habits and cultural appreciation could yield surprising results.37
Leaving aside such broad speculations, the remainder of this essay focuses on Ottoman literary culture, in its usual meaning of the verse and prose forms of initially
court-centred literature produced by and for the Osmanllar, the followers of [the
House of] Osman, the political and learned elites. Although only one of several written cultures within the empire, it was the one consciously developed to project Ottoman values and self-image among those who mattered. The following discussion asks
both how this literary culture might be viewed in a comparative perspective and how
we might assess the possible extent of its appeal. Ideally, in terms of the Ottoman
world in general, we should also consider the significance of Ottoman literary culture
beyond Istanbul and in the reverse influence of Muslim scholarly culture upon the
There is currently no major modern history of Ottoman literature in any western language. The nearest are the nineteenth-century, multi-volume studies of Ottoman poetry by the Austrian diplomat and scholar Joseph von Hammer (d. )
and the British Orientalist E. J. W. Gibb (d. ), published in the s and the
s respectively. In Turkish, the broader, encyclopaedic works of scholars such as
Mehmed Fuad Kprl and Aah Srr Levend, both writing in the mid-twentieth

Christine Woodhead

century, remain essential references, although the volume of publications on specific

elements of Ottoman literature is now increasing rapidly.38
Until relatively recently, writing a general survey of pre-nineteenth-century Ottoman literature was fairly straightforward. It would begin with texts in Old Anatolian
Turkish, the colloquial-cum-written language of thirteenth- and fourteenth-century
Anatolia, which remained the natural and most widespread idiom in ordinary communication, factual reportage and popular culture in Turkish-speaking communities
throughout the Ottoman period. This continuity is illustrated by the poems attributed
to Yunus Emre (d. ), which, in contrast to fourteenth-century texts in English
or French, need relatively little commentary for modern native readers. Our survey
would then chart how, for well over a century, sultans from Mehmed II ()
onwards sought to align Ottoman literature with the prestigious Perso-Islamic culture by attracting Persian poets and writers to their court and by encouraging the
influence of Persian styles through translations and imitative works, particularly of
poetry, metrical romances and historiography.39 This resulted by the late sixteenth
century in Ottoman-Persian verse and prose texts of considerable linguistic and conceptual sophistication. Incorporation of the Arab world into the empire after and
the foundation of more medreses under Ottoman patronage also produced a steady
increase in knowledge of written Arabic and the more widespread use of Arabic words
and phrases within the Ottoman-Persian written language. The pride taken in this aesthetic style is neatly encapsulated in the following declaration by Mehmed Nergisi (d.
), one of the most revered (and later most reviled) Ottoman prose stylists of the
early seventeenth century, all of whose written work is located firmly at the complex
end of the Ottoman language spectrum:
the Turkish [sic] language of pleasing expression [is] distinguished by its gathering
from the surrounding green meadows of various languages the choicest flowers of
meaning approved by men of eloquence and, through collecting thence the fruits
of clarity, admired for its natural qualities of pure and sound measure agreeable
to the palate.40
The Ottoman text of this passage contains around per cent vocabulary of Turkish origin, per cent Persian and per cent Arabic; it is held together, typically,
by the Persian izafet grammatical construction. Such rhyming, rhetorical prose was
closely related to chancery style and bears comparison with other belles lettres traditions, including those of the Byzantine empire and Italian Renaissance states, and
ultimately with its Arabic and Persian models.41
The resulting Ottoman Turkish high style both helped give a cohesive cultural
definition to Ottoman learned and ruling elites of very diverse origins and provided
a suitable vehicle through which to voice imperial cultural and political aspirations:
a facility with Ottoman, as opposed to simpler Turkish, came to be one of the hallmarks of membership in the Ottoman ruling class.42 The Ottoman empire could
thereby be presented as the Muslim civilization which incorporated, superseded and
outshone its predecessors. Proponents of Turki-i basit, the simple Turkish style
closer to the colloquial, always existed. With regard to court poetry, they appear
only as a small minority, with occasional notable figures, such as the poet Nedim
(d. ).43 In prose, however, relatively simple Turkish remained a valid option for

Ottoman languages

writers seeking a wider audience, as seen in some writings by the learned polymath
Katib elebi (d. ).44
Finally, our survey would suggest how, from a high point somewhere in the seventeenth century, from which the empire was thought to be in decline generally,
literature was also considered to be gradually stagnating. A damning verdict by Gibb
encapsulates the late nineteenth-century view: that great race to which the Ottomans
belong . . . has never produced any religion, philosophy or literature which bears the
stamp of its individual genius.45
Such a survey and conclusion could no longer stand scrutiny. Even in Gibbs unequivocal statement which prefaces his immensely detailed six-volume study of classical Ottoman divan poetry, the premier literary genre the inconsistency is evident.
While roundly condemning the Ottomans for slavishly following Persian exemplars,
Gibb yet managed to celebrate several centuries of a highly artistic and meaningful
idiom. His difficulty lay in reconciling the poetry which clearly fascinated him with
the dismissive views during the s of his Ottoman acquaintances in London, men
eager to develop new styles of Ottoman literature under western influences. They considered pre- Ottoman literary culture, including poetry, as largely imitative even
when strikingly expressive and to have been, from around onwards, mumbl[ing]
the dry bones of a long-dead culture.46
However, just as the assumption of a long political decline no longer dominates
Ottomanist studies, the tenacious notion of Ottoman cultural sterility arising out
of statements such as that by Gibb is now being steadily undermined, from several
angles.47 Hatice Aynurs presentation of current Turkish scholarship on seventeenthand eighteenth-century Ottoman poetry emphasizes two significant points. First, it
shows Ottoman poetry to have been a varied and constantly evolving art form, rather
than one which had become fossilized by the early seventeenth century. Second, her
survey suggests that over the centuries thousands of men, and a few women, must have
contributed to the genre.48 Although the most well-known poets are generally those
who attracted court patronage, composing and reciting verse was a highly regarded
activity throughout Ottoman society.49 Men from all walks of life composed verse.
Istanbul, with its concentration of wealthy patrons, remained the centre of poetic
activity, but not the only one. Mustafa Ali noted in more than twenty poets at
the court of Prince Selim (later Selim II, ) in Ktahya in western Anatolia,
and in the mid-s found some thirty poets in residence in Baghdad (not a centre
of princely government).50 The urbane and witty gentlemen of Ohrid admired by
Evliya elebi and the aljamiado poets of Albania were not isolated examples of provincial participation in the so-called elite Ottoman culture. Comparatively little work
has been done on the social context and reception of Ottoman poetry, as opposed
to the technical aspects of composition and the inner world of influence of one poet
upon another. It is nevertheless clear that the close association of divan poetry with
Sufi mysticism, the itinerant tradition of sufi learning, and the popularity of taverns
and coffeehouses among poets and dervishes from various backgrounds ensured that
Ottoman divan poetry was far from an elite, Istanbul-based preserve.51
The names of many poets and writers indicate strong provincial connections,
whether places of origin or of association,52 indicating that echoes of Ottoman culture could both filter down the social scales and radiate through provincial centres,
particularly in Anatolia and the Balkans. The extent to which provincial governors

Christine Woodhead

and their pasha households reflected the cultural as well as the organizational and
clientage aspects of the Ottoman centre deserves closer study, especially the degree
to which they connected with local communities and thereby conveyed elements of
specifically Ottoman culture. Much of Mustafa Alis literary output was dedicated to
the provincial governors and commanding generals whom he served outside Istanbul;
Evliya elebis position of educated companion to his kinsman Melek Ahmed Paa
enabled him to undertake many of his extensive travels during the pashas provincial
appointments.53 As indicated above, both writers appear to have found congenial company in the provinces. In the Arabic-speaking parts of the empire, Ottoman cultural
elements were apparently evident to some degree even in such prestigious centres of
Arab culture as Damascus and Cairo.54 Dina Rizk Khoury, in observing how the eighteenth-century Jalili governors of Mosul sought to create their own court culture, draws
attention to Mosuli intellectuals close interest in political developments in Istanbul,
and to shared cultural values.55 While such values may have been as much Muslim as
Ottoman, and the political dimension cannot be overlooked, this connection is nevertheless indicative of a certain Ottoman cultural influence.
Notable among the innumerable aspirant literary figures drawn to Istanbul from
the provinces is the poet Nabi (d. ), whose output also exemplifies many of the
points already made about Ottoman literature. The son of a learned and religious
family in Urfa, south-eastern Anatolia, in the mid-s Nabi travelled independently
to Istanbul, where he spent some twenty years and through his literary ability gained
the patronage of a close companion of Mehmed IV (). Most of his later life
was spent in Aleppo, writing in semi-retirement. His ten major works range in style
from the complex ina prose of the account of his pilgrimage to Mecca in ,
written for presentation to the sultan, to the relatively simple verse of the book of
advice written in for his seven-year-old son. Other works include two anthologies of divan poetry, one Turkish and one Persian, a verse account of the circumcision
festival of given for the sons of Mehmed IV, a translation into simple Turkish
of a popular collection of forty sayings concerning the Prophet Muhammed, and a
collection of Nabis own letters to Ottoman friends and statesmen.56 In other words,
Nabi composed in a variety of genres, in two languages, and in differing registers of
written Ottoman for a relatively diverse range of readers, adapting his style to suit the
purpose of each work.
An alternative and intriguing insight into the compilation of at least some Ottoman
texts and into the use of language is offered in the recent publication of the autobiographies of the sixteenth-century chief architect Sinan (d. ).57 Five texts are
presented, all by one author. The introductions to each of the first four are sequential
elaborations of the same basic account, compiled originally from Sinans dictation.
They present a rare opportunity to study the process of composition, such as how
simple statements were gradually elaborated, in what form, where and why prayers
and poetry were introduced, the relationship between prose and verse, and the varying vocabulary and imagery used as the text was developed. The fifth text is a work
of consciously literary ina prose, different in format to the other four, but obviously
dependent upon them for content. Together, they show an Ottoman text literally
under construction.
Finally, it may also be useful to re-evaluate the significance of the Ottoman relationship to Arabic and Persian cultural models. Given that major Ottoman engagement

Ottoman languages

with this cultural heritage occurred at roughly the same time and in similar ways to the
early modern engagement with the Greek and Latin heritage in the west, why is only
one of these movements considered a renaissance with positive outcomes?58 Equally,
there are many parallels between the Ottoman relationship to the Perso-Islamic heritage and that of other Turkic ruling dynasties, which undermine the notion of an
imitative, relatively unimaginative Ottoman attachment.
Not only did the Persian influence upon Turkish courtly culture in Anatolia begin
with the Seljuks in the twelfth and thirteenth centuries and thus pre-date the establishment of the Ottoman emirate, it was also dominant in the other major Turkic
empires Uzbek, Mughal and Safavid of the early modern era. More generally,
some degree of cross-influence between the Turkish and Persian languages had existed
for centuries in northern Iran and Central Asia. One particular cultural model for sixteenth-century Ottomans was the court of the last Timurid sultan of Herat, Hseyin
Baykara (d. ), patron of the Persian poet Cami and of the vezir Mir Ali ir Nevai
(d. ). The latters Evaluation of the two languages held that his own literary language, aatay Turkish, was richer than Persian, which it subsumed. aatay Turkish
remained the written and spoken idiom of the successor Uzbek state until the early
twentieth century.59 In northern India, the Mughal empire founded by the aatayspeaking Timurid prince Babur (d. ) took adherence to Persian much further.
By around Persian had been adopted both as the written language of literature
and administration and as the preferred spoken idiom of the elite. As Mehmed II had
tried to attract Persian scholars and poets to Istanbul in the mid-fifteenth century, so
too did the Mughal emperor Akbar () a century later.60 In Iran, the Turkic
dynasty founded by Ismail Safavi in grafted a Turkish-speaking political-military
elite onto indigenous Persian-speaking literary and administrative groups, resulting in
a third variant on the Turkish-Persian cultural relationship.61 Seen in these broader
perspectives, the Persian influence upon Ottoman appears much less dominant and
prescriptive. Persian did not become the first language spoken at the sultans court, as
it did in Muslim India; nor from the late sixteenth-century did it remain a language
of choice for much literary composition beyond certain types of poetry. Ottoman was
a form of Turkish considerably enriched by Persian elements, but arguably not
overtaken by them.
The influence of Arabic upon written Ottoman was also profound, although in
comparison with Persian few literary works were composed entirely in this language.
However, until at least the late seventeenth century, most works of religious and judicial scholarship were composed in Arabic, and facility with the written language was
compulsory for medrese students, from among whom the learned elite were drawn.62
Rudiments of Arabic might also be gleaned by many from Koran classes and elementary education in boyhood. In the fifteenth and early sixteenth centuries, an ambitious Ottoman student from Anatolia or the Balkans might follow the well-established
tradition of travelling for his education, spending several years with highly regarded
teachers in Damascus and Cairo and often combining study in the Hijaz cities of
Mecca and Medina with performance of the pilgrimage. With the establishment in the
s of the Sleymaniye colleges in Istanbul as the highest level of Ottoman religious
education, there was less incentive to travel and more reason to focus on the centre of
patronage and employment in the capital. Although after it was easier to visit the
major Arab cities, paradoxically a scholars need to do so was largely diminished.

Christine Woodhead

As the study of the history of Ottoman language use gathers pace, our understanding of it can only become more nuanced. That there was a natural need to develop,
on paper at least, an imperial idiom to serve political purposes and to provide a badge
of cultural identity is generally accepted as a feature of the fifteenth- and, especially,
the sixteenth-century Ottoman world. Thereafter, as membership of the Ottoman
group began to expand in the seventeenth century, it drew into its cultural orbit a
much greater range and number of people of varying levels of education, outlook
and political commitment. What effects this greater participation had upon the use
of language and the production of literary work is a major topic for study, which will
usefully move away from the traditional focus on official texts and the contents of the
sultans library.


My thanks to Nelly Hanna, Metin Kunt and Rhoads Murphey for comments on an earlier
version of this essay.
Faroqhi : , in a guide for new researchers on the problems, pitfalls and pleasures of
studying Ottoman topics.
For an indication of the progress and potential of literary studies, see Aynur .
For modern editions, and for certain edited collections of Ottoman documents, see the list of
published Ottoman sources in the Reference section of this volume.
On the Ottoman legacy and perceptions of it, see Brown , especially the essays by Barbir
and Todorova.
li , ; also Tietze .
Iz ; Kerslake : .
Interest is increasing in other types of text, e.g., Kafadar .
For a broader perspective on Turkish, see Kamusella : .
Strauss a: ; Houston : on linguistic variety within early modern European states.
Strauss a; Versteegh : .
Versteegh : .
My thanks to Nelly Hanna for this point.
On resettlement policies, see chapter , by Rhoads Murphey, in this volume.
Dankoff .
Evliya elebi : .
Ibid.: , ; on early Ottoman buildings in the Balkans, see Kiel ; on historiographical
interpretations of Islamization in the Balkans, see Zhelyazkova .
For Bosnian merchants from Sarajevo and Mostar petitioning the Doge of Venice in Turkish
in , see Murphey : , .
Elsie : .
Balta ; Strauss : . This example supports the argument that, because Turkish
was not a liturgical language too closely identified with another religion, there would have been
less reluctance among non-Muslims to using it.
Houston : , especially re France; Anderson : on administrative
iek ; see also the essays by Rossitsa Gradeva and Eugenia Kermeli in this volume, chapters and respectively.
Minkov : .
On educated captives employed as Ottoman interpreters, see chapter , by Tijana Krsti, in
this volume.
Stavrides , which also assesses Mahmud Pasas reputation as the exemplary sultan of

Ottoman languages

vezirs on account of his cultural patronage and Muslim piety. See also Murphey : ,
for a sixteenth-century account of Sokollu Mustafa Pasas visit to his home region in Bosnia.
Kunt . On language use in Ottoman Hungary, see chapter , by Gbor goston, in this
But see Zeevi : for the devsirme origin of the Ridwan dynasty of governors of Gaza
in the sixteenth and seventeenth centuries.
See chapter , by Ehud Toledano, in this volume.
For eighteenth-century views, see Barbir : ; Rafeq .
E.g., Lewis . Cf. Kahane et al. , the classic study of Greek and Italian borrowings in
Turkish nautical vocabulary from the early Ottoman period onwards. Zack shows how
far Turkish vocabulary had penetrated spoken Arabic by the seventeenth century (my thanks
to Nelly Hanna for this reference).
Houston on pre-modern European literacy generally, on oral and aural culture. See also Nelly Hannas discussion of literacy among Egyptian artisans, chapter in this
On the so-called millet system, see Masters : .
Cf. Hasluck ; Ocak : .
Fotic : .
Strauss b, ; see also essays on earlier non-Muslim book production within the empire
in Hitzel .
On the Alexander legend, see Stoneman ; on St George/Hzr, see Ocak : ; Brotton .
See her essay in this volume (chapter ), including comments on Ottoman use of the Alexander
On the emergence of a non-religious dimension to culture in eighteenth-century Cairo, see
Hanna : ; on books listed in Damascus inventories c., see Establet and Pascual
As indicated in Halman et al. , a four-volume multi-contributor history of Turkish
Sohrweide ; Woodhead .
Woodhead : .
On insa style and letter-writing in Arabic, see Gully . On the Arabic and Persian content
in literary Ottoman, see Buday ; cf. also Tietze .
Fleischer a: ; Necipolu : , on the curriculum of the palace school for the
future political elite.
Slay ; Murphey : .
Cf. Katib elebi .
Gibb : I, .
Ibid.: I, .
Cf. Bauer for the deplorable state of study into Mamluk literature, resulting partly from
the similarly deleterious influence of nineteenth-century western prejudices.
Aynur : , notes , poets recorded just in eighteenth-century Ottoman biographical
dictionaries of poets; Andrews and Kalpakl : , discuss sixteenth-century Ottoman women poets.
Aynur : .
Fleischer a: , , and passim, for the late sixteenth-century cultural milieu generally.
Andrews and Kalpakl : .
E.g., Aynur : , .
Fleischer a; Dankoff .
Winter .
Khoury : , especially .
Aynur : ff.; Coskun .
Sinan .

Christine Woodhead
Andrews and Kalpakl strongly advocate further comparative study of Ottoman and
European literatures.
Eraslan .
Alam ; see also Dale .
See, e.g., Perry .
Cf. Katib elebi : , and Chambers on courses of study for high-flying
students in the seventeenth and nineteenth centuries respectively.



Ottoman markers of difference

Baki Tezcan

ore than thirty five modern nations, from Slovakia and Ukraine in the north to
Eritrea and Yemen in the south, and from Algeria in the west to Azerbayjan in
the east, have at some point in their history been partially or wholly part of the Ottoman empire. During the last two centuries many of these modern nation-states have
become independent and have developed national historiographies that have with
varying degrees of success constructed nationalities and projected them back to the
Middle Ages or the ancient world. Accordingly, it has become increasingly difficult to
reconstruct the experience of Ottoman subjects in a multi-ethnic and multi-religious
empire in which loyalty to the ultimate political authority had little to do with ones
national belonging, and the term nation meant quite different things.
The difficulty is not just an intellectual one brought about by the rise of nationalist
historiographies. The last couple of centuries have also witnessed an endless shuffling
of people at the whim of political authorities, who did not hesitate to engage in a range
of activities, from genocidal massacres to population transfers, with a view to creating
nations which would be as ethnically and religiously homogeneous as possible. Thus,
in many parts of the former Ottoman territories, the physical remains of the diverse
Ottoman past consist of empty churches, mosques, or synagogues which have lost
their congregations. In a nationalist present that hails homogeneity, it has become
harder to imagine a diverse society in the past, all the more so because the children of
the people who constituted that past are no longer there.
In short, as a result of the history and historiography of the last two centuries,
attempting to reconstruct the Ottoman past as it was experienced by the peoples of
the empire is a daunting task. Such reconstructions used to oscillate between images
of a merciless Tourkokratia (Turkish rule, Greek) and of a tolerant Muslim empire
that created the Pax Ottomanica both in the region and for its subjects of different
ethnicities and religions.1 While neither of these two diametrically opposed images
could adequately reflect the Ottoman past, studies on the subject of Ottoman ethnic and religious groups were, until recently, written generally in defence of one version against the other. However, as the assumptions of nationalism came to be questioned by scholars,2 a new generation of historians has chosen to focus on the actual

Baki Tezcan

experiences of Ottoman subjects as reflected in various contemporary sources, especially the court records, and to revise the more simplistic views of the past. The purpose of this essay is not to provide a summary of these more recent studies, on some of
which more can be found elsewhere in this volume.3 Instead, taking these studies as a
point of departure, I will attempt to construct a historical model which could accommodate the multiplicity of Ottoman communities at any point in time during the
history of the empire as well as the shifting relationships between them.
My inspiration comes from the work of Brian Catlos, who suggests that inter
communal relations in the pre-modern Mediterranean were based on the principle of
convenience, that is to say, conveniencia as opposed to convivencia.4 Catloss work is
grounded in medieval Iberia. He suggests that the diversity of the medieval Mediterranean (c.) was
the consequence of a series of relationships born of a perceived mutual benefit
among majority and minority ethno-religious groups, reinforced by formal agreements set in law, and grounded in a mutually-conceded, if limited, legitimacy.
These relations were subject to on-going renegotiation, either formally or informally, and the standing of minority groups tended to improve or decline according
to their role and influence in the political and economic system of the principality
which they inhabited.5
According to Catlos, a minority groups capacity to dominate socio-economic and
administrative niches, which were not served by the majority, and yet which were
crucial to the larger functioning of the economy, and in particular to the ruling elite,
was essential for that groups survival. Their displacement from such niches as a result
of socio-economic change contributed to their decline and persecution.6
In this essay, I use Catlos model in order to conceptualize the historical experience
of Ottoman communities. This provides a productive starting point from which to
approach Ottoman communities, especially if one wishes to avoid the moralizing narratives of tolerance vs. intolerance, or such grand notions as the clash of civilizations.
However, in order to apply Catlos model to the Ottoman empire, it must be modified somewhat to allow for ethnic diversity within religious communities and to allow
more room for politics, both internal and international.
To start with similarities between Catlos model and the Ottoman historical experience, there is no doubt that Ottoman authorities thought about crucial socioeconomic and administrative niches when they considered their relations with the
representatives of Ottoman communities. The long-standing urban professional experiences of Armenians, Greeks and Jews provided Ottoman cities with fine craftsmen,
such as goldsmiths, excellent professionals, such as architects and medical doctors,
and major international intermediaries, such as multilingual merchants. The role that
many representatives of these communities played in public finance as tax collectors
and creditors is also very well known. It is not surprising that one of the first acts of
Mehmed II after the conquest of Constantinople was to order the deportation of
Christian and Jewish, as well as Muslim, families from various parts of Anatolia and
the Balkans to be resettled in his new capital.7 The existence of Muslim families in
this example should be noted. As will be discussed below, in this early period Muslims
were regarded as just one of several Ottoman communities.

Ethnicity, race, religion and social class

Some professional niches were available for Ottoman Christian communities in the
countryside as well. For instance, Muslim peasants would most likely not engage in
swine husbandry or wine production, leaving to Christians those activities shunned
by their own religion. However, such niches were not as significant for the Muslim
majority in the empire as were the urban professions mentioned above. Once settled, there was not much of a difference between a Turkish farmer and, for example,
an Armenian one. Their ethnic or religious differences did not matter much; their
most important distinction was the one they shared: being a peasant that is, a taxproducer for the ruling class. Therefore the conveniencia model, which foregrounds the
professional niches filled by minority communities, might need certain modifications
when applied outside the cities. On the other hand, this model works well when considering certain political needs of the royal authority, especially if we apply the model
to different ethnic-geographic groups within the same religious community. The way
in which the Ottoman centre approached its Kurdish periphery is a fitting example.
Ten years after Baghdad had fallen to the Safavids in the early s, a bureaucrat in
the Ottoman capital tried to persuade Murad IV () to treat the Kurdish lords
well by making the latters illustrious ancestor Sleyman the Magnificent ()
Just as God, be He praised and exalted, vouchsafed to Alexander the two horned
(zl-qarnayn) to build the wall of Gog, so God made Kurdistan act in the protection of my imperial kingdom like a strong barrier and an iron fortress against the
sedition of the demon Gog of Persia. A thousand thanks and praises to the presence of the Almighty, creator of the races of mankind. It is hoped that, through
neglect and carelessness, our descendants will never let slip the rope of obedience
[binding] the Kurdish commanders (mera) [to the Ottoman state] and never be
lacking in their attentions to this group.8
In the early sixteenth century, Istanbuls alliance with Kurdish lords had secured the
political allegiance and military support of large parts of Kurdistan to the Ottomans
rather than the Safavids.9 Now was the time to remember this in order to reconquer
Baghdad, which Murad IV did in , approximately five years after the above passage was written.
In short, the principle of convenience is indeed helpful in approaching the relations
of the Ottoman centre with various Ottoman communities, both Muslim and nonMuslim. This principle is also instructive when we leave the domain of the political
authority and focus on individual members of Ottoman communities.

When explaining the multi-layered and multi-faceted nature of individual Ottoman
identities to American undergraduate students, I remind them of the primary meaning of identity as stated in the Oxford English Dictionary: the quality or condition of
being the same in substance, composition, nature, properties, or in particular qualities under consideration. In the process of identifying ourselves, we use the qualities
that we share with others, such as being American, Christian or anarchist. Yet at the
same time each one of us shares different qualities with different groups, hence our

Baki Tezcan

multiple identities: Muslim like Ahmad, Shiite like Hasan, Iranian-American like
Fatima, middle class like Michael, female like Grace, cool like Joy.
Ottomans were not very different from modern-day Americans with their multifaceted identities. Take Eremya elebi Kmrjian, for instance. He was an Orthodox
Armenian layman, some of whose writings suggest that he was not very fond of living
under Muslim rule. But then he had good relations with many high-ranking Ottoman
officials and made a lot of money thanks to his connections with the state. Moreover
he had no problem with writing in Armeno-Turkish, a language which offers some
surprising phrases to readers of Turkish, who mostly expect that users of Turkish must
have been Turkish and Muslim in the early modern period. Take such phrases as bizim
ermen tayfasy (our Armenian nation) or dinsiz turk (the infidel Turks), for instance.10
So Eremya elebi was Christian like Yorgo, Armenian like Garo, a state contractor
like Ali, and a writer in Turkish like Hasan.
Depending on the specific interaction in which Eremya elebi was engaged on
any given day, he could evoke one of the facets of his complex identity conveniently
for instance, by underlining his membership in the Ottoman financial elite when he
bid or struck partnerships for tax-farm contracts. The Ottoman political authorities,
in their turn, conveniently accepted Eremya elebis intermediacy in public finance.
He was also clearly a member of the Ottoman literati. His title, elebi, one used by
the learned Ottoman establishment, is one clear indication of this. His awareness of
Muslim Ottoman authors and the latters awareness of him in their respective works
is another.11
Continuing from Eremya elebis example, diversity seems to work best when individuals have a wide network of affiliations, allowing them to relate to multiple collectives by evoking different facets of their multi-faceted identities. Not surprisingly, if
one of these facets receives more attention than the others, the individual might lose
touch with some of the collectives he or she has been affiliating with and come to
essentialize that particular facet of his or her identity. The key to the healthy functioning of the multi-ethnic and multi-religious Ottoman society was, then, perhaps not
so much the absence of national identities as a lack of need to essentialize them. An
Ottoman author of Kurdish origin, ukri, who wrote a history of the reign of Selim I
() in Turkish verse and presented it to Sleyman, probably meant this when
he wrote: Trk ilen Trk ve Krd ilen Krdem (I am a Turk [while being] with a Turk
and a Kurd [while being] with a Kurd).12
That Ottoman individuals were well aware of their national identities, based on
their ethno-geographical and racial ties, and used them for their own advantage when
they needed to could easily be confirmed by looking at the Ottoman ruling class itself.
Metin Kunt demonstrated the significance of cins as ethnic-regional solidarity by
studying the ties among the devirme that is, the personal slaves of the sultan who
were collected from among the Christian population of Anatolia, the Balkans and
the Caucasus. He noted the existence of alliances based on cins connections, such as
Abkhazian, Albanian or Bosnian factions.13 A case also occurs of the term cinsiyet being
used in a racial sense when the subject of reference was the ties between an Ottoman
scholar of African slave origin and the black eunuchs of the palace.14 Yet these ties,
however important they may have been, were only one set of ties among many others,
the most significant of which for the ruling class was their loyalty to the sultan. Thus
when Melek Ahmed Paa was asked by the grand vezir in whether he would fight

Ethnicity, race, religion and social class

Hasan Paa, an Abkhazian like himself, Melek Ahmed quickly replied that he would
fight against any enemy of the sultan.15
Loyalty to the person of the sultan or the empire was not just a phenomenon found
among the members of the Muslim ruling class. As late as , Anthimos, the Patriarch of Jerusalem, stated that
Our Lord . . . raised out of nothing this powerful Empire of the Ottomans in the
place of our Roman [Byzantine] Empire which had begun, in certain ways, to
deviate from the beliefs of the Orthodox faith, and He raised up the Empire of
the Ottomans higher than any other Kingdom so as to show without doubt that
it came about by Divine Will.16
Eremya elebi, in his History of Istanbul, refers to the sultan as our emperor and
boasts about the many ambassadors who come to his court from India, Central Asia,
Persia, Russia, Ethiopia, Georgia, Morocco, Genoa, Poland, Austria, France, England
and the Netherlands in a way which suggests that he is as proud of the power of his
sultan as a Muslim Ottoman would be.17 A seventeenth-century Orthodox priest from
Serres in modern Greece, Synadinos, who wrote a local chronicle in the Byzantine
tradition, refers to the Ottoman sultan in the Byzantine fashion as the basileus, never
calling into question his legitimacy of rule.18 Thus it was quite possible to accommodate multiple national (ethno-geographical and/or racial) as well as religious loyalties
under the umbrella of political loyalty to the person of the sultan or the empire.
The principles that govern the relations between different communities in Catloss
model are reinforced by formal agreements set in law, and grounded in a mutuallyconceded, if limited, legitimacy. In the case of the Ottoman empire, these principles
came to have a little more than limited legitimacy. Ottoman jurists law a term preferred here instead of eriat, as it underlines the role of jurists in the articulation of the
law did not discriminate among different nationalities as long as they were Muslim.
When it came to non-Muslims, the law had a built-in space that protected the rights
of Christians and Jews without, however, granting them full-scale equality with Muslims. While such a legal system should indeed have had only limited legitimacy in the
eyes of its Christian and Jewish subjects, the archival record suggests that Christians
and Jews used the Ottoman courts much more frequently than previously believed.
Convenience seems to have played a role in this development as well.
Marriage and divorce fees at the kad courts were lower than those collected by
the bishops. In addition, simpler procedures for example, no investigation into
the motives for divorcing, no conciliation terms, and no bans on subsequent marriages and the wish to evade some canonical restrictions might have induced
Christians to use the services of the kad.19
When such methods to circumvent the canonical authorities were more difficult
to follow for instance, in Istanbul, where one was within physical proximity of the
patriarchate Christian (and Jewish) women resorted to Ottoman courts to convert
to Islam in order to secure a divorce from their husbands, as a Muslim woman could
not be married to a non-Muslim man.20 This suggests that Christians and Jews not
only used Ottoman courts for their own business but were also aware of some of its

Baki Tezcan

finer points which they could exploit to their own advantage. Recent research on
the jurisdiction of ecclesiastical courts in the Ottoman empire suggests that even in
such legal fields as inheritance the Ottoman courts had the ultimate authority.21 Even
though there were other courts in the daily life of Ottoman subjects, such as the communal and ecclesiastical courts discussed elsewhere in this volume,22 the courts presided over by kads seem to have been the ultimate arbiters of justice. Therefore these
courts should simply be called Ottoman courts rather than kad courts, a term that
is still prevalent in the historiography even though it implies a legal system in which
there were multiple networks of courts with equal jurisdictional authorities, with kad
courts being one of them.
The imposition of a particular legal system, which is historically connected with a
given religious tradition, on the followers of different religious-legal traditions appears
problematic from a pluralistic perspective. Nevertheless, it could integrate different
communities of the empire into one legal system and thus increase the legitimacy of
the system as a whole. The frequent recourse by Christian and Jewish Ottoman subjects to Ottoman courts supports this assumption. In this particular sense, the legal
arrangements that secured the Ottoman version of Catlos conveniencia had more than
limited legitimacy. They had become the law of the land in a way not unlike that by
which the Anglo-Saxon legal tradition became the law of the land in North America
as a result of conquest.
All of this should not be taken to mean, however, that the communities of Ottoman
society were well integrated in all aspects of their lives. As Catlos asserts, diversity in
pre-modern societies could only function if clear communal boundaries were maintained.23 Maintaining such boundaries was important for the survival of minority
groups and the continuing dominance of the majority group. Not surprisingly, one
of the areas in which these dynamics were most visible was the patriarchal hierarchies
involved in miscegenation. That is why, for instance, Muslim men could marry Christian and Jewish women while a Christian or Jewish man could only marry a Muslim
woman if he agreed to convert. While the boundaries between communities were
porous, they were so only in one direction, towards Islam. A Muslim who left his or
her religion would be punished by death.
Also, the principle of convenience was not as convenient to the minority group
as it was to the majority. This was especially so when socio-economic tensions were
carried over to intercommunal relations. Catlos draws attention to the fact that, even
though the Christian majority in thirteenth-century Aragon was much wealthier than
both the Muslim and Jewish minorities, there were still many middle-class Jews and
Muslims who were wealthier than many Christians.24 To make up for a state of affairs
that seemed inappropriate to poorer Christians, minority groups were occasionally
subjected to actions that would confirm their secondary status. Populist movements
articulating economic inequalities expressed their sense of deprivation by targeting
minority communities.25
The channelling of socio-economic grievances towards minorities is clearly observable in the Ottoman empire. In the late sixteenth and early seventeenth centuries,
when the absolutist political posture of the court drew much criticism, a Jewish
woman became the target of rebels. In the last days of March , Esperanza Malchi, the Jewish kiera (lady in Greek), known to have been very close to the sultans
mother, was lynched by imperial cavalrymen who believed that the debased coins with

Ethnicity, race, religion and social class

which they had been paid had come through her tax-farm payments. The soldiers had
also demanded the heads of the chief gardener and the sultans chief white eunuch,
but were appeased after they had lynched Malchi and her son.26 It was also in this era
of extreme social turmoil both in the capital and the provinces that orders were issued
banning the wearing of silk by Christians and Jews. Soon after, the use of alcohol was
strictly prohibited, all taverns were closed, and the office of the finance ministry which
collected taxes from the sale of alcohol was abolished though all of these restrictions were lifted and the alcohol tax was reinstated eighteen months later, in autumn
In short, however accommodating the principle of convenience was for minority
groups, it was ultimately based on acceptance of the political dominance of the majority, or, at least, of the symbols thereof. What, then, is the advantage of the use of
this model over the discourse of tolerance vs. intolerance? Most obviously, it removes
religious labels from the historical analysis and provides us with a model that works
in multiple contexts with different majority and minority groups. However, a much
more significant advantage offered by this model is the possibility of integrating socioeconomic issues into intercommunal relations, and even of assigning primacy to them
in explaining change.


I will now focus on the question of long-term change in the relations between Ottoman communities, by suggesting that it was social mobility which was the destabilizing
factor. It was pointed out above that, once settled, there was little difference between a
Turkish Muslim peasant and, for example, a Greek Christian one. Other than raising
pigs and producing wine, the Greek Christian peasant would not occupy a particular
niche in rural society. Yet, he or she would definitely be paying taxes. Perhaps, then,
the members of the minority group did not necessarily have to dominate a niche: it
was enough for them to be good taxpayers, which was convenient for the ruling class.
For the minority peasant, the convenience was that he or she continued, just as before,
to be a taxpayer after the Ottoman conquest. If anything, the Ottomans were known
to have decreased the tax ratios in certain areas to ease the shift in political loyalties
from former rulers to themselves. There was not much else that affected a peasants
life directly.
I would argue further that the minority peasants tax-paying subject position was
in a sense key to his or her potential affiliation with other peasants belonging to the
Muslim majority. Unlike in the city, where professions and trade brought people from
different communal backgrounds together in all kinds of interactions, villages were
much more monolithic. Thus, again, the way in which a Greek peasant from a Greek
village and a Turkish peasant from a Turkish village could relate to each other would
be on the basis of their shared tax-paying subject status as far as their relationship to
the imperial centre or to their immediate tax collector was concerned.
The central political authority was well aware of this fact. The relevant political
literature that touches upon the relations between the ruling class and the subjects
had only one term for the latter: reaya, the flock. It did not matter whether a peasant

Baki Tezcan

was Muslim or Christian both would be a part of the flock. Despite the fact that
Christian peasants paid additional taxes in the form of cizye (the non-Muslim poll tax),
in terms of being disenfranchised vis--vis the ruling class, they were as far away from
political power as their Muslim neighbours in the next village. In fact, a Christian
peasant might even have had a better chance of joining the ruling class in the fifteenth
or the sixteenth centuries than a Muslim one because of the practice of devirme, which
supplied the sultan with loyal soldiers and administrators of Christian origin.
Thus, while in urban environments communal differences were mediated through
economic and professional interactions, in rural areas there was little ground for communal differences to play a discriminatory role, as all peasants ultimately shared the
common denominator of being subjects in the flock. The political authorities addressed
all communal groups in the same language and treated them similarly. Differences of
class were the most significant differences in the political literature of the time. Authors
warned their readers not to let members of the reaya join the ranks of the ruling class.
Those people who did manage to find a way in were called ecnebi, or foreigner, even if
they were of Muslim origin or perhaps because of their Muslim subject origin.28
The Ottoman ruling class originally consisted of conquerors who were predominantly of Turkish origin and yet included many former members of the Byzantine
and Balkan aristocracies. After the ascendancy of the devirme in the second half of the
fifteenth century, the ruling class came to consist of the descendants of the conquerors
as well as the slaves of the sultan, who were of Christian origin. To this ruling class,
subjects of Muslim origin were indeed foreigners. In short, the ruling class discriminated against all subjects in the same way, treating them all as foreigners. A simple
Christian subject (unless he was selected as a devirme) and a Muslim one were equally
excluded from entry into the ruling class.
As far as the Ottoman authorities were concerned, Muslims constituted one of the
communities that they needed to manage and exploit by taxation. They happened to
belong to the same religion as that followed by the conquerors, but in practice they
were subjects, just like the Christians and Jews, whose taxes supported the ruling class.
When around Knalzade Ali considered the relationship between the masters
of the state (devlet) and the reaya, it never occurred to him to separate Muslims from
[As long as] the union that is concluded among the masters of the devlet lasts, the
supremacy and domination, which are inseparable from that union, continue. It is
an established fact that each devlet begins with a group of men entering an alliance
and acting like the members of one body . . . in helping and supporting each other.
Each individual has a certain degree of power, yet once their power is assembled
together in one place, it becomes more than [the sum of] each ones power. Thus
a small group of allied men prevails over many groups that are disunited. Is it not
obvious that the number of men who are in possession of a devlet does not even
amount to one hundredth of the number of that devlets subjects (reaya)? And yet
since these men are allied and the subjects are not, the former become victors and
rulers while the latter are defeated and ruled.29
Thus, as far as the Ottoman authorities were concerned, all subjects, whether Muslim,
Christian or Jewish, were reaya, or their flock.

Ethnicity, race, religion and social class

The reason for dwelling so long on this particular point the common subjecthood
of Christians, Jews and Muslims alike is that it is central for us to understand how
the principle of convenience stopped working in the later centuries of Ottoman history. According to Catlos,30 it was the inability of minorities . . . to adapt to changes
in the socio-economic environment that might displace them from their niches that
tended to contribute to their decline and persecution. While agreeing that socioeconomic change is the key to explaining the changing status of minorities, we should
also note the political dynamics through which the effects of socio-economic mobility
within the majority population are transformed into communal rifts in a multi-ethnic
and multi-religious society. The first of these is the formation of a collective identity
that connects the ruling class with the majority population, which is at the heart of the
development of modern nationalism.
Even though not every Muslim joined the Ottoman ruling class in the seventeenth century, sizeable groups from among merchants and financiers did find ways
of entering it, despite the protesters who called them foreigners. The late sixteenth
and seventeenth centuries were times of remarkable socio-political mobility in the
Ottoman empire. The practice of devirme came gradually to an end by the beginning of the eighteenth century as members of the Muslim upper middle classes came
to man the administrative and military posts of the empire.31 This was followed
by a gradual shift in the meaning of the term reaya. From the eighteenth century
onwards, the flock came to mean Christian taxpayers only.32 It is not that Muslims
did not have to pay taxes any more. They continued to do so, but as Muslims, for
this became the collective identity that connected the ruling class with the majority
of the masses.
This process is comparable to the rise of nationalism in Europe. The displacement
of a former aristocratic ruling class by a larger bourgeoisie was followed by the invention of the nation, the name given to itself by the people as it formed itself into
the quasi-caste which it had never been in the past.33 Obviously, not every Frenchman could become part of the ruling class, but each one of them could be proud to
be a Frenchman and, with that pride, be induced to pay his taxes happily, while being
conscripted and sent to wars in distant lands. Being French was the collective identity
that tied the ruling class and the masses together and made the latter feel as if they had
a share in the glories of their nation. Yet even a poor Frenchman had to feel superior
to someone. Thus did racism in the modern sense first arise in a democratic society, a
mass society whose expressed ideals were fraternal and egalitarian.34 Around the same
time white commoners in colonial America found a shared identity as white men by
asserting their superiority defined against Indians and Africans.35
Not unlike the poor Frenchman or the white American commoner, the common
Ottoman Muslim felt that he belonged to a political nation in which class differences were mediated by the collective identity that tied him or her to the ruling
class. While it was race that became the marker of this collective identity in colonial
America, in the eighteenth-century Ottoman empire religion was being assigned
that function. The ruling class continued to include many non-Muslims in provincial as well as imperial administration, from kocabas, or non-Muslim community
leaders, to Phanariot princes who ruled Moldavia and Wallachia.36 However, Muslims no longer constituted part of the reaya, the flock, and had moved up to become
citizens of sorts.

Baki Tezcan

It is time to return to another point mentioned earlier. Medieval and early modern
identities were multi-faceted, as individuals were able to relate to different groups by
evoking a particular facet of their identity. Yet if one of the facets of an individuals
identity were to receive more attention than the others, that person might lose touch
with some of the collectives he or she had been affiliating with and come to essentialize
that one particular facet of his or her identity. The Ottoman empire in the nineteenth
century provided ample opportunity for essentializing identities. As the empire could
not compete with Russia militarily, and became increasingly subject to European
imperialist designs politically, Russians became protectors of Orthodox Christians by
the end of the eighteenth century and the French of Catholics later on. As these imperial powers started offering economic advantages to their Ottoman Christian intermediaries whose protectors they had become, many an Ottoman Christian merchant
found himself interacting more and more with European Christians and finding new
meaning in his Christian identity as a key to economic and political privilege (which
could even include citizenship papers). It seemed more convenient now to emphasize
communal identities to the exclusion of others, as there were socio-economic and
political advantages to be gained.
It should not be surprising that the crystallization of a collective identity around the
category of Muslim contributed to the alienation of Christian subjects of the Ottoman
empire, who were left by themselves in the status of flock. Had the Ottoman empire
been located on a continent more or less isolated, like the United States, this alienation
could probably have been addressed successfully in the long term, as majorityminority relations are usually subject to continuing negotiations. Yet the Ottomans faced
several powerful neighbours, whose interest in competing for the political loyalties of
their Christian subjects complicated the course of such negotiations between Muslim
rulers and these subjects. It seemed more convenient to ally with Russians for independence in Serbia or to secure British help for Greek independence rather than negotiating with Muslim overlords who were having difficulty adjusting themselves to the
idea of full-scale equality with their Christian subjects. To put things in perspective,
one might consider how long it took white Americans to get used to full-scale equality
with Native or African Americans. Such international alliances between some Ottoman Christians and foreign powers, in turn, transformed the loyal Christian subjects
of the empire into members of a fifth column in the eyes of many Muslims, who, in
turn, started essentializing their Muslim identity. Thus between the beginning of the
nineteenth century and that of the twentieth, the diversity of the Ottoman empire was
replaced by several nation-states or colonies that became increasingly less diverse during the twentieth century.
In retrospect, it seems that, as long as the most significant marker of difference in
Ottoman society was the one between the rulers and the subjects (reaya), members
of Ottoman communities were able to mediate their communal differences by, on
the one hand, forming a wide range of affiliations with members of other communities and, on the other, being reminded continually that, no matter which communal
group they belonged to, they were all members of the rulers flock. Increasing social
mobility, coupled with the development of a collective identity that brought the new
ruling class together with the Muslim masses out of which they had emerged, destabilized the relations between Ottoman communities. The exploitation of this destabilization by European imperialism led to essentialization of communal identities and the

Ethnicity, race, religion and social class

eventual transformation of a multi-ethnic and multi-religious society into much more

monolithic nation-states.

For examples of such treatments, see the works cited in Gradeva : , n. .
Anderson [] .
E.g., the essays by Gbor goston (Hungary, chapter ), Nelida Fuccaro (Kurdistan, chapter
), Rossitsa Gradeva (Ottoman courts in the Balkans, chapter ), Eugenia Kermeli (Greece,
chapter ), Tal Shuval (Algiers, chapter ), Ehud Toledano (the Ottoman Arab provinces,
chapter ), Stefan Winter (Shiites in Syria, chapter ), and Christine Woodhead (language
use, chapter ).
Catlos , .
Catlos : . This paper is part of Catloss work-in-progress, tentatively entitled Paradoxes
of pluralism: ethno-religious diversity and the medieval Mediterranean.
Catlos .
nalck : , .
Murphey : (Turkish, p. ).
Tezcan .
For a detailed biography of Eremya elebi in English, see Sanjian and Tietze : .
Armeno-Turkish was Turkish in Armenian letters. A similar mixture of languages had earlier
occurred in Crimea, where the Armenian diaspora had adopted Kipchak. This they took with
them to Podolia, where in they printed arguably the first book in a Turkic language; see
Chirli . The exemplary phrases are from Schtz : , .
Bardakjian : ; Sanjian and Tietze : , n. .
kr, Selmnme, Topkap Saray Ktphanesi, Hazine , f. b.
Kunt .
Tezcan : .
Kunt : .
Clogg : .
Eremya elebi Kmrcyan : .
Strauss : ; Synadinos .
Ivanova : ; see also Rossitsa Gradevas essay in this volume, chapter .
Baer .
Kenanolu : .
See Eugenia Kermelis essay in this volume, chapter .
Catlos : .
Catlos : , fig. .
Nirenberg .
Her other son saved himself by converting to Islam and promising to pay their debts to the
treasury. He apparently became known as Aksak Mustafa avu and died during the reign of
brahim (); Selaniki Mustafa Efendi : II, ; Naima : I, , ; the
English ambassador Lellos dispatch of March , National Archives, London, SP /,
f. ; on the term kiera, see Mordtmann .
Tezcan : .
See, for instance, such late sixteenth- and early seventeenth-century sources as the anonymous
Hrzl-mlk, in Ycel : (facsimile, f. b); the anonymous Kitb- Mstetb,
ibid.: (Ottoman text, p. ); and Koi Beg risalesi : , .
Knalzade Ali, , book III, ; cf. a manuscript of the same text in Sleymaniye Ktphanesi, Hamidiye , f. a.
Catlos : .
Tezcan .
Even in the late seventeenth century, the term had no denominational connotations; see

Baki Tezcan

Meninski : II, c. . Cf. Faroqhi c: : From the th/th century onwards,

the term is increasingly used for the Christian taxpayers only; th/th century population
counts distinguish between recy and Islm.
Guillaumin : .
Ibid.: .
Taylor : xiii.
Philliou .




Stefan Winter

he Kzlba were essentially the antithesis of Ottoman din devlet, religion and
order. The term, which refers to the red, twelve-pleated turban emblazoned with
the names of the Shiite imams that was worn by the eastern Anatolian tribal followers
of the Safavid sufi order of Ardabil in the late fifteenth century, seems already to have
been used by Safavid leaders of the time. In Ottoman chancery sources it is first used
in a derogatory sense when a number of these tribes began to revolt and helped shah
Ismail conquer Tabriz in , laying the groundwork for the establishment of a Shiite rival state in Iran.1 In the mid-sixteenth century, judicial opinions by the Ottoman
eyhlislam (chief jurisprudent) Ebussuud Efendi (d. ) and others defined the
Kzlba as illegal heretics whose elimination was a religious duty. This permitted the
state to pursue a veritable inquisition against the heterodox tribesmen whose frequent
revolts continued to shake Ottoman rule in Anatolia, while also providing a legal and
ideological framework for further warfare against the Safavids. Although the Anatolian
Kzlba gradually ceased to be a major concern for the Ottoman state the
celali rebellions, though alluding to an earlier Kzlba revolt, in fact had nothing
to do with them the Iranians and their supporters in Iraq are characterized as Kzlba
throughout the wars of the seventeenth and eighteenth centuries. The conflict with
the Kzlba is thus often seen as a basic fact of Ottoman and indeed Middle Eastern
history in general, as the start of the empires uncompromisingly Sunni identity, as
the paradigm for its treatment of minorities, and as the validation of an unbridgeable
political split between Sunnism and Shiism.
Given the importance of the challenge posed by the Kzlba to the Ottomans, it
is hardly surprising that they have been the subject of sustained scholarly interest in
modern times. At least five collections of Ottoman archival documents dealing specifically with the Kzlba, the Bektais and other unorthodox groups have been published, re-edited, or translated in the past two decades.2 Numerous popular histories as
well as a growing number of scholarly studies have also been devoted to the subject, the
most recent of which have tended to emphasize the role which centreperiphery relations, the formulation of authoritative categories of orthodoxy and heterodoxy and the
establishment of an early modern state bureaucracy played in the persecution of Shiites
in the empires formative period.3 Despite all this attention it is somewhat noteworthy

Stefan Winter

that Ottomanists have not brought more interest to bear on one of the empires largest
and best-documented heterodox populations, whose indigenous leaderships were consistently referred to as Kzlba in certain texts throughout the post-classical age and
indeed well into the nineteenth century. The Twelver (Imami) Shiite communities in
the mountain hinterlands of Tripoli, Sidon and Damascus i.e., in Mount Lebanon,
Jabal Amil and the Bekaa Valley, all in todays Lebanon were integrated into the
structure of Ottoman provincial government in the sixteenth century, with leading
families such as the Hamadas and Harfushes acquiring tax farms and even formal recognition as sancakbeyis (district governors). In the later seventeenth century, however,
Ottoman officials increasingly began to define them in sectarian terms whenever pursuing tax collection or punitive campaigns against them, while at the same time turning effective sovereignty in the area over to the Druze emirs of Sidon. Although successive foreign occupations between and each seemed to bring the promise
of greater local enfranchisement, in effect the reaffirmation of central authority in the
nineteenth-century reforms, the growing involvement of European powers in Lebanese communal affairs, and finally the exile of the Harfush emirs to Edirne in to
all intents and purposes marked the end of the Kzlba problem in Syria.
This essay traces the evolution of the rural Imami community of Syria in its rapport
with the Ottoman state. After a survey of the regions incorporation and the co-optation of its tribal leaderships, it argues that the Shiites were liable to be categorized as
Kzlba not on account of a hypothetical connection to Iran, as is sometimes claimed
in contemporary literature, but as a legal device in order to justify official violence over
more prosaic matters such as brigandage and tax evasion. The most intense punitive
campaign against the Syrian Kzlba coincided with an unprecedented thaw in relations with Safavid Iran in the late seventeenth century, while renewed hostilities with
Iran coincided with the quasi-recognition of a Kzlba mukataa (fiscal district) in
Mount Lebanon in the early eighteenth. We conclude by suggesting that the monetarization and then the decentralization of provincial rule in the age of the ayan
(local notables) ultimately provides a better paradigm for understanding the Ottomans position vis--vis the Syrian Shiites than the question of religious persecution
or tolerance.


The Ottoman conquest of Syria in was facilitated not only by the defection of
a number of Mamluk officers but also by the quick submission of the regions most
powerful Bedouin chiefs and other local notables to Selim I. The claim by later historians of Lebanon that an emir Fakhr al-Din Man the First likewise paid tribute
and was therefore entrusted with authority to rule over the Druze and other tribes of
the coastal highlands on the states behalf has been shown to be a myth endowing the
later Man emirate with historical legitimacy. Rather, a letter preserved in the Topkap
Palace archives shows that among the first tribal leaders to offer their fealty to Selim
in and entreat him to implement Islamic order and establish justice in the area
was in fact a scion of the Shiite Harfush family of Baalbek.4
Owing to the difficulties of asserting direct control in the mountainous hinterland,
the Ottomans were quick to adopt the practice of tax farming by local notables. Tax
cadastres and timar appointment registers indicate that the Harfushes and other Shiite

The Kzlba of Syria and Ottoman Shiism

notable families, such as the Sabs of nearby Jabal Amil, held several tax concessions
throughout the sixteenth century. At the same time, the mhimme defterleri (registers
of imperial chancery decrees) also make clear to what extent the Ottomans struggled
to maintain order in the region and repeatedly had to send out military expeditions
against unruly Druze and Shiites. These expeditions were usually presented in terms of
a struggle against rebel heretics, when in fact they had nothing to do with religion or
politics per se. The Harfushes themselves were denounced as Revafz the standard
pejorative term for Twelver Shiite heretics after they attacked and plundered several
villages in the Bekaa Valley in , but the authorities were concerned above all with
the resulting loss in taxes and indeed continued to rely on the Harfushes to govern
Baalbek. Around the same time, one of the first chancery orders specifically against the
Druze and Rafzis who are not of the four [orthodox Sunni] sects in the province of
Safad (northern Palestine and Jabal Amil) was sparked by their brigandage and illegal
possession of firearms, rather than by their confessional identity, which at any rate was
not news to the Ottomans.5
Whatever policy against Shiism Ebussuud and others might have framed seems
to have had little to no influence on actual day-to-day administration in the Syrian
provinces. The degree of Ottoman reliance on local intermediaries regardless of their
confessional affiliation is illustrated by the Harfushes recognition not only as suba
(sergeant) or emin (commissioner) of the city of Baalbek but as local district governors
as well. As early as , Musa ibn Harfush was directed to lead a force of , archers and join the Ottoman campaign to Yemen, against the Zaydi Shiite imam no less,
for which he would receive the sancakbeyilik of Sidon in return. It is not clear whether
this in fact occurred, but in his son Ali was being referred to as emir or bey in
official Ottoman correspondence and ranked in the provincial military hierarchy as
governor of the desert sancak of Tadmur (Palmyra).6
By the following year Ali had also been appointed sancakbeyi of Homs, and seems,
somewhat unbelievably, to have offered to take over the tax farms of his main competitors, the Druze Mans of Sidon, if the whole region were reorganized and given
to him as a full-blown beylerbeyilik (provincial commandership).7 Ottoman records
do not indicate that he was actually made governor of Sidon, but according to the
eighteenth-century traveller and historian Giovanni Mariti, the Sublime Porte did
name him pasha of the Druze and had several local chieftains deported to Istanbul
specifically to prevent them from rallying to the Mans in . In reality, however,
Ali Harfush himself remained vulnerable. The Ottomans had taken care to disarm his
supporters as well, and after being forced to retreat into the mountain stronghold of
Ghazir he was easily defeated by his Druze rivals after their return from exile.8 Thus,
despite its limited consequences, Ali Harfushs stint as pasha nonetheless underlines
the extent to which the Shiite emirs were entrenched in the Ottoman administration
of hinterland Syria in the later sixteenth century.
In his major biographical dictionary, the Damascene scholar al-Muhibbi (d. )
later qualified the Harfushes as extremist Shiites, with the exception of Musa Harfush, who succeeded his father after the latters execution in and was supposedly
the closest to Sunnism.9 This has led Abdul-Rahim Abu-Husayn, whose detailed
study on Ottoman provincial leaderships in Syria remains the standard in the field, to
posit that the Ottomans must have suspected the Harfushes, whose extremist commitment to Shiism was well known, of being in contact with the Safawids of Persia,

Stefan Winter

with whom the Ottomans were soon to be at war again, and that when Musa became
governor of Homs in he probably dissimulated Sunnism to gain favour with the
Ottoman authorities, and to escape the fate of his father.10
In fact, the chancery orders did not concern the Harfushes confessional identity, let
alone their purported ties to Iran, but only the new ways in which they were to extract
several years worth of back taxes from the district of Homs (though these in turn were
later denounced as bidatlar, illegal innovations).11 A unique, recently published iltizam
(tax farm) register from the province of Damascus in the early seventeenth century
further illustrates the basis of the Ottomans rapport with the Shiites. Covering the
years to , the register details a growing competition between the Harfush
emirs and the Druze Man emirs of Sidon for tax collection concessions in the Bekaa
Valley, where numerous sources of revenue were earmarked for the organization of
the yearly pilgrimage to Mecca. Both the Shiites and the Druze routinely attacked,
denounced and outbid each other in their contest to serve, as the register puts it, this
primary objective of Ottoman din devlet. The Harfushes were temporarily shut out
of the race in , when their erstwhile partners, the janissary ocak (regiment) of
Damascus, decided that the better-connected Druze might be able to produce more


How did the Ottomans conceive of the Syrian Shiites potential link to Safavid Iran?
Modern historiography on Shiism in Syria under Ottoman rule has concentrated overwhelmingly on the migration of scholars in the sixteenth and seventeenth centuries
from Jabal Amil to Iran, where they helped establish Shiism as the state religion and
often attained high office. Their success in Iran, as documented through their own
writings and biographical dictionaries, has in turn served to highlight their marginality
and supposed mistreatment back in the Ottoman empire. Salient and well-publicized
events such as the execution of the legal scholar Zayn al-Din ibn Ali or the dissimulation by the Safavid mctehid Baha al-Din al-Amili of his Shiite identity during a visit
to his native Syria appear to provide ample evidence for this. Historians working solely
within the Arabic and Persian narrative tradition have thus conditioned a very pessimistic image of Shiism under Ottoman rule, theorizing that the persecution of Shiism
was an official Ottoman priority and expressing confidence that this perception will
yet be borne out by undiscovered Ottoman chronicles and mhimme documents.13
This would indeed be an interesting discovery given not only that none of the
collections of Ottoman documents cited above, nor any of the Ottoman chronicles
known thus far, nor the many modern studies of heterodoxy in the Ottoman empire
have turned up any references to official, state-coordinated repressive measures against
Shiite scholars, but also that the many references to known Syrian Shiites that do
occur in the sources of this period consistently fail to connect them with the Kzlba
question. In , for example, the Sunni muftis of Damascus accused a colleague
of corruption (fesad) and heresy (rafz), but were warned to guard against prejudice
and zealotry if an investigation proved this to be untrue. In the Sublime Porte
issued orders to monitor closely some Rafzis from the east who might try to travel
to Damascus and Aleppo as pilgrims or merchants and mislead and corrupt some
of the people of those areas who share their outlook, but significantly did not name

The Kzlba of Syria and Ottoman Shiism

or inculpate the latter.14 During the celali rebellions, Yunus Harfush broke with his
family and supported Canbulad Ali Paa who did actually have Safavid backing
in his attempt to seize control of Syria. However, this not only did not compromise the Harfushes standing as tax farmers of Baalbek, but even enabled Yunus to
demand compensation for damages incurred during the revolt and subsequently to
become the Ottomans point man in the region when the Druze emirs began to pose
a greater threat. The Harfushes only appear to be mentioned in an Ottoman chronicle
after being deprived of the governorship of Baalbek, in the context of the empires
increasingly vicious fight to dislodge the Safavids from Iraq in . Following the
heroic, dramatically narrated death of the vali of Damascus with his troops against the
Kzlba swine near ehrizor that summer, his successor returned to Syria to at least
rout the Harfushes and a thousand of their followers in the Bekaa yet here again
there is no linkage made between their rebelliousness and the empires wider struggle
against a Shiite foe.15
Murad IVs reconquest of Baghdad in eventually gave way to the longest
period of peace in the history of OttomanSafavid relations. For commercial purposes
the shahs found that there was no practical way of circumventing Aleppo and Izmir in
order to reach the European market, and the trade in Iranian silk through the Ottoman empire peaked around . Politically, the two sides began a regular exchange of
embassies, and the Iranians pointedly refused to support a European coalition against
the Ottomans in , even helping to restore Ottoman control over Basra after a
tribal rebellion there the following decade.16 It is thus noteworthy that the Ottomans
began to label and prosecute Syrian Shiites as Kzlba in precisely this period. In
the Sublime Porte received a complaint from the religious notables of Tripoli
that a family of local tax farmers, the Hamadas, was no longer content with dominating the regions Christian- and Druze-inhabited districts but had now also begun to
usurp tax farms whose population was Sunni Muslim. The authorities were evidently
aware that the Hamadas were Shiites, and indeed suspected them of collaborating with
the Harfushes of the Bekaa Valley, who were also being targeted by a new punitive
campaign at this time. Late the next year the government denounced the Hamadas
for the first time as Rafzi Kzlba brigands and ordered the governors of Tripoli,
Damascus and Sidon to capture and punish the Shiite heretics according to law.17
When this did not happen, the government organized a more substantive campaign in
early against the Kzlba sect that had appeared in the Tripoli region and settled
in the highlands and that was swallowing the income legally due to the treasury of
In fact, the Hamadas had held some of these farms, including those inhabited mainly
by Sunnis, for over forty years. Like the Harfushes, their roots appear to go back into
Mamluk times, and they had come to the attention of the Ottoman government as early
as . The punitive campaign, which ended with the Hamadas defeat and dispersal a year later, is among the best-documented events in the history of Ottoman Shiism, recorded not only in several contemporary Arabic chronicles but also in dozens of
mhimme orders sent to provincial authorities throughout Syria and Anatolia, instructing
them to mobilize against the Kzlba or to secure nearby port towns and desert roads
against those attempting to flee. Perhaps somewhat more surprisingly, it is also described
in the chronicle of Defterdar Sar Mehmed (d. ) and, on the basis of his account,
in the official history of the Ottoman vakanvis (official historiographer) Raid Efendi

Stefan Winter

(d. ).19 (Neither version provides any detail beyond that conveyed in the mhimme
registers, raising the question of whether Sar Mehmed may not have relied simply on the
then available chancery documents for his story.) Ironically, at exactly the same moment
that the Sublime Porte was preparing its last great war against a Kzlba rebellion, in
February , an Iranian ambassador arrived in Edirne, where the Ottoman court was
in residence, with a message of congratulations on Ahmed IIs accession and numerous
lavish gifts from Shah Sulayman I, marking a historic high point in diplomatic relations
between the two empires.20
Why are the Hamadas referred to as Kzlba in these sources? Although Lebanese
historiography has often sought to depict Shiites in general as ethnically Iranian,
there is little reason to doubt the familys nomenclature and claims of descent from
an Iraqi Arab tribe, and the Ottoman orders do not identify them with Anatolian
Trkmen or accuse them of rebelling out of complicity with the Safavids.21 Rather,
an answer is suggested in some of the orders sent to the provincial authorities in
Syria in the spring of , in which the campaign to root out and extirpate the
Shiites is characterized as being in accordance with a noble fetva promulgated last
year i.e., around the time the Hamadas were first characterized as Kzlba.22
No legal ruling specifically regarding the Hamadas has turned up in the chancery
archives or in the court registers of Tripoli, and it is likely that the reference here is
simply to a rehash of Ebussuuds famous fetva, which had formed the basis for the
empires persecution of Shiites as Kzlba heretics since the sixteenth century.23 In
any event, the fact that the term was first applied ad hoc in the context of a government punitive campaign suggests that it should be seen as an attempt to classify
the Hamadas topically as illegal rebels rather than to fundamentally question their
ethnic origins or religious sympathies.
That the categorization of the Hamadas as Kzlba was above all a legal device is
further suggested by the fact that this term was now consistently used also for other
Shiites in the region whenever they fell foul of the state authorities. Mushrif Ali alSaghir, for example, the leading Shiite lord of Jabal Amil in the province of Sidon, was
characterized as the refuge of villainy for the Kzlba bandits after he hosted some
of the Hamadas who had fled Mount Lebanon; in the government in turn gave
orders to block the trails leading north so that local Shiites, the Kzlba in the SidonBeirut mountains, could not go to help the Hamadas in a renewed conflict with the
state in Tripoli.24 In , the Hamadas were once again in trouble with the authorities and fled to take refuge on land belonging to the Kzlba known as Ismail Harfush
in Baalbek, although the Harfush emir had just been on pilgrimage with the Syrian
caravan the previous year and was not himself explicitly targeted by the order. His sons
Haydar and Husayn, whose battle for control over the Baalbek district, and especially
its vakf incomes, repeatedly attracted the attention of the central government over the
next decades, were also characterized as Kzlba. However, owing to their protection
by the Shihabi emirs, as well as by the chief black eunuch (kzlar aas) in Istanbul,
the government was essentially powerless to remove them.25 And, finally, the death of
the famous shaykh Nasif Nassar and the annihilation of his forces by Cezzar Ahmed
Paa in southern Lebanon in , which for all intents and purposes marked the end
of Jabal Amils autonomy under Ottoman rule, was also celebrated as a defeat of the

The Kzlba of Syria and Ottoman Shiism


















Western Syria: eighteenth-century eyalet divisions

Stefan Winter


If the term Kzlba was first applied to the Hamadas circumstantially in order to
legitimize an unprecedented campaign of state violence, it soon began to be used as
a general and increasingly meaningless pejorative label. In the eighteenth century the
empire simply did not have the means or the will to follow through the political course
that the official rhetoric against illegal Shiite heretics would have dictated, preferring
to relinquish effective authority in the Syrian hinterland and other provinces to local
or localized Ottoman ayan dynasties. In the Hamadas were reinstated in Tripoli
through the good offices of the Shihabi emirs, who had just recently replaced the
Mans as the premier tax farmers and overlords of the Druze community (though the
Shihabis themselves were actually Sunnis and later converted to Maronitism). Ottoman fiscal records from this period quite matter of factly list the mukataas in the
mountains above Tripoli as being in the care of shaykh Ismail the Kzlba or as being
farmed, since the days of their forefathers, to someone called shaykh Ismail of the
Kzlba under the guarantee of emir Haydar Shihabi ibn Man. In the government complained that the Kzlba faction owed three years worth of back taxes and
was constantly oppressing the highland districts, but was always being reappointed
nonetheless on account of the Shihabis intervention.27 The following year, just before
the Ottomans took the opportunity presented by the Safavid dynastys collapse to
embark on a campaign to capture Iranian territory, which would again be justified in
terms of a war against irreligion, the government explicitly noted the existence of the
Kzlba mukataa annexed to the province of Tripoli and constituting a tax farm in its
entirety.28 While the Ottoman state could not countenance ideologically the existence
of a heretical fiscal territory or institution, the increasingly common, pragmatic use
of the term to describe local administrative realities suggests that the authorities no
longer had any advantage in denouncing the Syrian Shiites per se.
Administrative documents from the region indicate that local Ottoman officials were
also keenly aware of the limits of their authority vis--vis the Shiite enclave. According
to the Islamic court records of Hama, the mustahfzan troops stationed in Musayliha castle on the road between Tripoli and Beirut repeatedly petitioned the imperial
council to increase the stipends due to them from provincial revenues, pointing out
that the said castle lies on the border [hudud] with the Kzlba.29 A separate entry in
the court records of Tripoli concerns the appointment of a new blkba (company
commander) to Musayliha, whose principal responsibility lay in protecting travellers
passing that spot on the road leading to Sidon and Egypt from the Kzlba brigands.30
In the government forwarded a warning to the governor of Sidon that the tax
farmer of Jubayl, Ismail Hamada the Kzlba, was likely to revolt and cause harm to
the tax farms and oppress the righteous as soon as the governor of Tripoli had left on
the annual pilgrimage relief column. A few years later both governors were apprised
that the Kzlba shaykh was constantly providing bandits from the entire region with
a safe haven from the law in his district.31 None of these denunciations, however, gave
rise to military action against the Shiites, nor are they linked with the statutory fetva
against the Kzlba that would implicitly have enjoined such action.
Ironically, the relative autarchy the Kzlba mukataa enjoyed under Ottoman rule
proved in the long run also to be its undoing. By abandoning direct control over the
rural hinterland, the imperial government was also paving the way for more powerful,

The Kzlba of Syria and Ottoman Shiism

better connected ayan such as the Shihabis of Sidon to invest in an increasing number
of tax farms beyond their home province, to rally the local Maronite Christian community under their protection, and finally to extend their emirate throughout the
highland region that would retrospectively be defined as Lebanon.32 In , the
Maronite population began a series of uprisings against the Hamadas that ended with
the intervention of Yusuf Shihabi and the formal transfer of all their former landholdings by the court of Tripoli four years later. Interestingly, this entire episode seems to
have escaped the notice of the Sublime Porte, with the one exception of an attack on
a Christian village that was signalled to the governor of Tripoli in and in which
the Hamada thugs were of one mind and in accord with the Harfush emirs of Baalbek and had gathered a large group of Kzlba riff-raff under their command.33 In
the years following, however, the Hamadas and almost all the Shiite tribes of Mount
Lebanon were driven into exile in the northern Bekaa Valley, where they are still established today. It is perhaps telling of the Ottoman imperial authorities general loss of
control over the provincial periphery in the eighteenth century that the final elimination of the Kzlba mukataa in Syria occurred virtually without their knowledge or


The extension of the Shihabi emirs dominance over Mount Lebanon, the Bekaa and
Jabal Amil and the subjection or violent displacement of the local Shiite communities in the late eighteenth and early nineteenth centuries are covered in detail in the
conventional narrative accounts of Lebanons history.34 In Ottoman sources, on the
other hand, the Shiites are really only mentioned in relation to the occasional governor rebellions and military irruptions that were beginning to undermine Ottoman
sovereignty in the region. But since their only consistent aim in this time of profound
change was to continue and try to secure some measure of autonomy vis--vis more
powerful actors, the Kzlba now begin to appear, depending on the context, as either
enemies or allies of the state. The Shiites occupation of Sidon in in support
of the major wartime rebellion of Egypts Muhammad Abul-Dhahab and Zahir alUmar of Acre occurred at perhaps the nadir of Ottoman imperial power and does not
seem to be noted in any chancery sources; the description of their victory over the governor of Damascus and the Shihabi emirs given in Ahmed Cevdets history appears,
much like his overall summary of politics in Mount Lebanon, to be based entirely on
local Christian sources.35
The situation was different in , when another invasion from Egypt that
of Napoleon Bonapartes forces was halted at Acre by Cezzar Ahmed Paa. The
government in Istanbul early received intelligence reports from the area that the
Druze have until now not leaned toward or followed the infidels. Though some of the
Kzlba were inclined toward the wretched unbelievers, having witnessed their defeat
and weakness they have also regretted this and given up on them.36 However, orders
given later in the summer are unequivocal in their condemnation of the Shiites and
offer insight into Ottoman attitudes towards the non-orthodox subject population
at this juncture. Noting that, initially, the Revafz were established on the lands of

Stefan Winter

my noble state, ploughing and planting and earning their livelihoods in various ways,
and were thus beneficial and sided with the exalted state, the document then accuses
them of having provided aid and supplies to the French; for their treason and base
ingratitude they were now to be reckoned among the enemy. In an echo of Ebussuuds still valid ruling, their blood is declared to be licit and their goods and livestock
to be plundered according to law and their women and children enslaved. Yet it is
noteworthy that the orders in this case did not use the religious-legally charged term
Kzlba, possibly because the equally heretical Druze were again being held up as
models of obedience and loyalty toward the state.37
Over the next years the Shiites and other heterodox groups were often invoked
in chancery correspondence only in terms of their perceived political reliability. In
Cezzar Ahmed, who had temporarily fallen from official favour, undertook to
tarnish his rival Abdallah Paa of the Azm family as an ally of the Druze, Kzlba
and Nusayris and was eventually rewarded by being reappointed to the governorship of Damascus.38 Cezzars successor, Sleyman Paa the Just, made a point of
reconciling with the Shiites in order to demarcate himself from his predecessor, the
Butcher, and in another Abdallah Paa of Sidon reinstated Nasif Nassars son
Faris and other shaykhs in their traditional tax fiefs in Jabal Amil to win their support in his bid to seize the governorship of Damascus by force. The defending vali of
Damascus, Dervi Paa, as well as other local officials, reported with alarm that the
Druze, Nusayri and Kzlba factions had joined the rebellion, with the Kzlba
army even threatening the hajj route near Homs, but since subduing them through
warfare is futile, and as some Druze clans had already been won over, the vali suggested sparing neither money nor presents to conciliate the other groups and pull
them over little by little to the Damascene side.39 Just like the Druze, however, the
local Kzlba did not have a single, clear-cut position during the rebellion anyway.
Only the previous year the Shiites of Mount Lebanon (Tripoli) had backed their
Maronite neighbours in an ammiye (popular uprising) against Abdallah Paa and
his ally the Druze emir; in late June, meanwhile, Dervi Paa could note with satisfaction that emirs Husayn and Salman Harfush, certainly the most powerful Shiite
figures in the region (though they are tellingly not identified as such), had agreed to
join the Ottoman cause.40
These documents appear to be the last that refer to the Shiites in Syria (or anywhere else in the empire) as Kzlba. The final Egyptian invasion by Ibrahim Paa
ten years later, in , heralded a new epoch, especially for non-Muslim and heterodox populations, by extending them at least a theoretical legal equality and generally
transforming the relationship between the modernizing state and the individual social
and economic subject. However, whereas the Shiites had welcomed previous governor
rebellions precisely as a means to reinforce their own autonomy, Ibrahim Paas intrusive, centralizing reforms and his decision to ally with Bashir Shihabi now only engendered opposition on their part. As early as the leading Harfush emirs, as well
as members of the Hamada family, were in contact with the Ottoman expeditionary
commander Mehmed Reid Paa to coordinate the resistance against the Egyptians.41
Despite efforts by some members of the Harfush family to collaborate with Ibrahim
Paa in order to retain control of Baalbek, by the Shiites as a whole were playing
a leading part in the uprising that would break his grip on the coastal highlands and
help return the region to Ottoman rule.42

The Kzlba of Syria and Ottoman Shiism

In the following years, Ottoman rule consisted for the most part in the futile search
for a power-sharing agreement between the Maronite and Druze communities of
Lebanon that would be agreeable to their respective French and British backers. In
, foreign minister ekib Efendi appointed a twelve-member meclis (council) to
serve each of the two semi-autonomous kaim-makamiyes (district deputyships) that
had been established under European auspices (but which were in fact similar to the
consultative councils being introduced in other provinces). These summarize well the
Shiites ambiguous status in the reform era: while each of the six locally dominant sects
could name both a judge (kadi) and a counsellor to the meclis, the Shiites had to be
represented by a Sunni judge as the state did not recognize any Shiite judicial authority.43 The Harfushes, meanwhile, returned to dominate the Bekaa after the Shihabi
emirates demise but were now confronted by Frances ever more ardent defence of the
Christian communitys interests. Ottoman foreign ministry papers, for instance, detail
the komisyon called in to investigate Salman Harfushs various transgressions
against the French protg Yusuf Juday and others, some of which already dated back
almost a decade.44
Banished initially to Crete, the Harfush emirs were permitted to return to Syria in
after humbly petitioning the sultanate, in a letter harking back to earlier notions
of personal loyalty and patronage, to consider their separation from their homeland
and their longing for their sons and families after so many years.45 While the Shiites
then largely stayed aloof from the sectarian violence between Christians and Druze
later that year, Salman Harfush was declared an outlaw when he sided with the Yusuf
Be Karam insurrection and died captive in Damascus a few years later.46 The rest of
the family was banished to Edirne in ; the last trace of the Harfushes in Ottoman
documentation concerns their appointment to various sinecures in the civil bureaucracy in Istanbul towards the end of the century.47

The Kzlba of Syria add to our understanding of Ottoman Shiism in that they give
a name and a face to a phenomenon which is almost totally anonymous elsewhere
in the empire. First used in Ottoman chronicles and chancery documents of the sixteenth century to denote an ill-defined but clearly problematic tribal-cum-religious
movement spreading out from Anatolia, the application of the term to the Hamadas,
Harfushes and other already well-known Twelver Shiite notables of Lebanon casts
a different light on the Ottomans perception of heterodoxy. After the conquest in
, these families were often co-opted into the Ottoman provincial administration,
as mukataacs or even as governors of secondary sancaks such as Tadmur and Homs,
with fiscal and police responsibilities over a vast section of the Syrian coastal highlands.
Between and the Sublime Porte occasionally chose to invoke Ebussuuds
fetva and castigate them as Kzlba heretics over more mundane offences such as tax
evasion and brigandage, but even these documents make clear that the authorities saw
them neither as connected with the Anatolian tribes nor as supporters of Iran. In the
eighteenth century the central state acquiesced in the near autonomy of the coastal
hinterland under the rule of local notable dynasties, still regardless of their confessional affiliation, and began to use the term Kzlba less as a call to arms or a legal
justification for killing Shiites than as a mildly derogatory ethnographic identifier.

Stefan Winter

The changing paradigms of Ottoman provincial administration the monetarization and privatization of government office beginning in the sixteenth century, the
devolution of power to local or localized ayan, and finally the attempt at centralizing
reforms in the nineteenth century thus probably provide a better framework for
understanding the fate of Syrias Shiites than the question of religious orientation or
persecution. For much of the Ottoman period, these Kzlba, the inevitable occasions
of violence between tribalists and government agents notwithstanding, were fully integrated in the state administrative system in the region. Their marginalization in the
modern era at the hands of local competitors acting in concert with European imperial
interests, far from being peculiar to their situation as Shiites, might in turn be seen as
expressive of the empires development as a whole.


I am grateful to Sylvain Cornac for his help in procuring documents and to the participants
of the Great Lakes Ottomanist Workshop (GLOW) in Montral for their feedback and
Hinz : ; Mlikoff : ; Veinstein b.
Ahmet Refik ; z ; Sava ; ener ; ener and Hezarfen .
See Imber ; Zarinebaf-Shahr .
Akgndz : IV, ; on the Druze emirs, see Salibi , .
BOA, Mhimme Defterleri [MD], : (#); MD : (#); MD : (#).
BOA, MD : (#); MD : (#); MD : (#).
BOA, MD : (#); MD : (#).
Cited in Worbs : .
al-Muhibbi (n.d.): IV, .
Abu-Husayn : .
BOA, MD : (#); MD : (#); MD : (#); MD : (#); MD :
(#); MD : (#).
Nagata et al. : esp. , , , ; analysed in more detail in Winter :
Stewart .
BOA, MD : (#); MD : (#); MD : (#).
Naima : II, .
Faroqhi : ; Longrigg : , , .
BOA, MD : (#, ); MD : (#); MD : (#); MD : (#);
MD : (#, ).
BOA, MD : (#, ); see also BOA, Ali Emiri, II., Ahmed .
Sar Mehmed Paa : , ; Raid Mehmed Efendi : II, , .
Winter : .
al-Amin : II, ; cf. Abu-Husayn : ; Abu-Husayn : , .
BOA, MD : (#); MD : (#) and MD : (#) (identical); MD :
Ebussuud Efendi : ; Imber a: , , .
BOA, MD : (#); MD : (#); MD : (#); MD :
BOA, MD : (#); MD : (#); BOA, Cevdet Evkaf .
BOA, MD : (#); Cevdet Paa : III, .
BOA, Maliyeden Mdevver :; :; MD : (#).
BOA, MD : (#); on the Azerbayjan campaign, see Longrigg :

The Kzlba of Syria and Ottoman Shiism

Damascus, Markaz al-Wathaiq al-Tarikhiyya, Hama Islamic Court Records : (#),
: (#).
Tripoli, Qasr Nawfal, Islamic Court Register :.
BOA, MD : (#); MD : (#); Cevdet Dahiliye .
Winter : , .
BOA, MD : (#); Tripoli, Islamic Court Register :.
al-Zayn ; Abi Haidar : , ; Hamada : I, , ; II, .
Cevdet Paa : I, , , ; cf. Abu-Husayn .
BOA, Hatt- Hmayun (HH) .
BOA, Cevdet Dahiliye .
BOA, HH i; see also HH .
BOA, HH a-b, HH , HH .
Rustum : II, (#), (#), (#).
Ibid.: IV, (#), (#), (#), (#), (#),
Havemann : , , ; Farah : .
BOA, Hariciye Mektubi Kalemi /; Hariciye Nezareti Tercman Odas /.
BOA, Divan (Beyliki) Kalemi /.
Kerr : , ; BOA, rade Dahiliye .
BOA, Cevdet Dahiliye ; Dahiliye Mektubi Kalemi /, /, /.



The celalis c.

Oktay zel

odern historiography depicts Ottoman history in the late sixteenth and early
seventeenth centuries as a period of crisis, generally associated with the celali
rebellions which occurred primarily in Anatolia and, to a lesser extent, northern Syria.1
Discussion of celali causes has centred largely around sixteenth-century politicalfiscal and economic-demographic developments. Recently added to the argument has
been the agency problem of the predatory state, the nature of the overall transformation that the empire and society experienced in parallel with the celali movements
throughout the seventeenth century, and climatic changes. The present essay reassesses
the issue in the light of new findings based on recently explored sources. It examines
internal demographic, economic and political dynamics and conditions, as well as
conjunctural factors in the s which led marginalized groups within rural society,
primarily the peasant masses, into a violent reaction in the form of banditry and rebellion. A reconstruction of the historical context and causes is followed by discussion of
the destructive character of the celali movements and its consequences in Anatolia, and
of the nature of this violence as a phenomenon within Ottoman social history. The
central argument of the essay is that the political-level analysis of the celalis falls far
short in understanding both the peculiarities of the historical process which prepared
the ground for the celali movements and the extent of destructive violence throughout
the seventeenth century. Banditry and occasional rebellions continued, transforming
rural society, the economy and the ecological environment as well as leaving a legacy
of institutionalized violence which became an inherent characteristic of later Ottoman
Anatolian society.


Financial crisis and the military institutions
The late s and early s were critical years for the Ottoman fiscal administration. Rising prices, increasing war expenditure, failure in the flow of tax revenue to
the central treasury, and consequent difficulties in payment of salaries to thousands

The reign of violence

of kapkulus, both the imperial guards and the standing army, infantry and cavalry,
placed great pressure on the Ottoman treasury. The pragmatic solution to this crisis
was devaluation and debasement of the currency: between and the Ottoman ake was devalued by per cent and its silver content reduced by per cent.2
Another attempted devaluation in and inability to pay full salaries in the new
currency provoked revolt among the imperial cavalry in Istanbul, appeased only by the
execution of the chief treasurer and the vezir held responsible. During a similar revolt
in , the grand vezir was sacked and the treasurer saved only with difficulty.3 Fiscal
concerns had become the governments chief priority.4
Debasement had been common practice from the time of Mehmed II ().
Such a fierce reaction in was due partly to the large increase in the number of
salaried kuls during the third quarter of the sixteenth century, and partly to a growing
sense of marginalization among Istanbul-based troops.5 While regular recruitment of
Christian boys through the devirme continued, numbers had increased also through
recruitment of local Muslim peasants, particularly during periods of border warfare
and princely succession struggles. Peasant recruits who had subsequently been given
status and payment as kapkulu soldiers were considered outsiders (ecnebi) by the Istanbul-based kapkulu troops and their incorporation as violation of an established rule.6
Moreover, during the long wars against Iran ( and early s) and Habsburg
Hungary (), the position of the kapkulu soldiers deteriorated as the financial crisis deepened. Salaries, even if paid on time, had drastically lost value and purchasing power as a result of debasements and rising prices.7 Contemporary sources
also mention increasing distrust on the part of these soldiers towards the pashas and
vezirs under whom they fought, due to unpaid salaries and unkept promises or to their
incompetence, insults and abuse of power.8
Kapkulu troops may also have considered themselves the losers within a climate
of selective promotion which had begun to affect all categories of the imperial elite.9
Stationed in Istanbul, they were among the first to observe how high positions and
lucrative revenue holdings were sold through advance payments (often seen as bribery by contemporary observers) rather than given on merit. By the s, sale of office
had become established practice as all governorships were given to senior military
administrators in return for advance payment in gold to the treasury and to grand
vezirs.10 Rank-and-file kuls witnessed their hopes for higher positions gradually fade
away. Increasingly vulnerable to fluctuations in fiscal and military administration,
they began to react more violently to deteriorating conditions.
Istanbul kapkulu troops were not the only ones affected by the crisis. From the midsixteenth century, thousands of middle- and lower-ranking provincial cavalry timarl
sipahis financed by revenue-holdings (dirlik) in return for military service also began
to react to similar financial and administrative changes in the Ottoman redistributive
system.11 The sipahi class had originally been composed largely of local elements, most
of whom had connections with landed and/or military aristocracy from pre-Ottoman
times. From the mid-fifteenth century onwards, they were incorporated more closely
into the Ottoman timar system as fief-holding soldiers, gradually losing previous privileges as their dirliks shrank in size and lost value.12 Over time they were increasingly
replaced by timariots of devirme-kul origin,13 and in both cases, during the sixteenth
century, there was a tendency towards smaller, less valuable fiefs.14 By mid-century
there was a growing pool of disaffected sipahis, particularly in Anatolia.

Oktay zel

Inflation and debasement of the currency probably had negative effects on the sipahis too, in addition to natural fluctuations in revenue income on account of climate
changes and unpredictable harvests. The more their fixed-income dirlik lost value,
the more difficult it was to equip themselves and their mounted soldiers (cebeli) for
military campaigns, particularly during the thirty-year period of regular campaigns
from onwards.15 By that year, many had already begun to squeeze their revenue
sources, the taxpaying peasantry, by extracting illegal payments and food, and were
drifting into lawlessness and de facto banditry. By a significant proportion of
remaining fief-holders simply could not afford to attend military expeditions on their
diminishing resources; many sought to evade military duty.16
The most serious manifestations of reactionary discontent among provincial sipahis
occurred from the last quarter of the fifteenth century to the s, through their
direct involvement in the political crises caused by struggles for the throne among
competing princes: between Selim [I] and Ahmed (and his son Murad, ),
and between Bayezid and Selim [II] ().17 The former coincided with the most
violent phase of OttomanSafavid rivalry and its offspring, the kzlba revolts, which
lasted until and created havoc in Anatolia. A certain proportion of the sipahi
class clearly participated actively in these revolts, either for religious-spiritual motives
through sympathy for the Safavid cause or due to deterioration in their financial and
military/political status.18 In addition, from the s onwards, the imperial administration gradually increased the number of soldiers with firearms, which further undermined the position of the sipahi cavalry as a whole.19 Hence, during the BayezidSelim
rivalry in the s there was a significant number of disaffected sipahis in the countryside who hoped to restore their fiefs or to better their condition, and to whom the
competing princes appealed for support.20
As argued originally by Akda, the celali rebellions of the late sixteenth and early
seventeenth centuries were closely related to the general financial crisis and the transformation of the kapkulu and sipahi institutions. By the s, not only sipahis in
Anatolia but also kapkulus everywhere were participating in unlawful activities and
violent reactions. The same soldiers increasingly undisciplined behaviour and uncontrollable desire to gather as much war booty and spoils as possible during campaigns
and regular frontier raids fits into this picture.21

Oppression and demo-economic factors

The ordinary subjects of the empire (reaya), particularly the peasantry, were the primary victims of maladministration, financial crisis and over-taxation, and of extortion
by agents of the imperial administration, mostly the kuls and sipahis.22 Economic
and demographic factors also contributed greatly to the deteriorating conditions of
the peasantry. The empire-wide survey registers (tahrir) indicate an almost linear
trend of growth in the sixteenth-century taxpaying population and agrarian economy.
The population growth rate varied between regions, but often exceeded per
cent. Both rural and urban settlements expanded, as did agricultural production.23
Arable lands were extended to the limits, particularly in areas with fertile soil; new
villages emerged, some developing rapidly into towns. There was also a clear trend
towards sedentary life among semi-nomadic elements in regions such as Bozok and

The reign of violence

However, economic and demographic expansion brought its own problems.

Repeated fragmentation of peasant farmsteads (iftliks) produced a surplus population with minimal plots of land to cultivate, or none at all. By the s the number
of landless peasants exceeded the number of households who still held some plots of
land, however small in size.25 Another critical and related development was a sharp
increase in the number of unmarried adult males. Although the gross agricultural product expanded in parallel with the expansion of agricultural land, this was ultimately
limited, and the growth of population far exceeded the increase in arable land. Two
significant problems arose: demographic pressure, particularly in densely settled areas
of fertile land in Anatolia, and a decrease in per capita production, which produced
a serious subsistence crisis during the third quarter of the sixteenth century.26 These
factors combined to push a large proportion of the most dynamic peasant groups out
of their villages, to migrate and seek livelihoods elsewhere.27
However, signs of demographic pressure and subsistence crisis were present
throughout the empires rural areas. The crucial question is why the celali movements
as such, with chronic waves of big rebellions of large armies, took place only in Anatolia and northern Syria, and not elsewhere. The Balkan provinces were plagued by lowlevel banditry and highway robbery.28 In the Arab provinces, including North Africa,
tribal banditry was chronic and there were frequent regional rebellions with their own
characteristics. However, neither the Balkans nor the Arab provinces experienced a
celali-type phenomenon.
Probable answers lie in three principal semi-structural characteristics of Anatolia
under Ottoman rule. First is the existence of a large, uncontrollable Turkish/Trkmen population, particularly in the Karaman, -il and Zulkardiye provinces, as well
as western Taurus regions of Anadolu province, whose socio-political characteristics
were shaped by the tension between the lifestyle of dynamic, mobile semi-nomadic
tribes and the centralizing policies of the state.29 Likewise, in south-eastern Anatolia
and northern Syria, the chief problem for the central administration was the Kurdish
and Arab tribes who were routinely active in highway robbery, raiding and rampage.30
Second is a culture of popular dissent associated with messianic ideology and a tradition of extreme self-sacrificing zeal, evident from the Babai revolts of the s,
through the popular-messianic rebellion of eyh Bedreddin in the s, to the early
sixteenth-century kzlba era.31 This helped fuel resistance to the centralizing state.32
Third is the unique source of chronic socio-political instability created by the practice
of prince-governorships, in Anatolian provinces only, and the Ottoman system of
competitive succession to the throne. Dating from the long period of civil strife after
Bayezid Is defeat at Ankara in , the periodic attempts of rival princes to gain the
support of notables and the ordinary population in their respective localities destroyed
the precarious social balance, undermined class barriers and further exacerbated popular discontent. Resistance to imperial centralizing policies among the most mobile and
economically deprived segments of the reaya, peasants as well as the semi-nomadic
Trkmen and Kurdish elements, increased during the dynastic struggles of
and especially .33 In the latter period such men were recruited en masse as
soldiers in the warring camps and offered either sipahi fiefs or kapkulu rank. As the
mainstay of the rival princely armies, they became further politicized. In this context,
Arcanl and Kafadar suggest that Sleymans sons may be considered the first celali

Oktay zel

From another angle, Yunus Ko emphasizes the diminishing capacity of the state
to absorb the most mobile, dynamic elements of the lower echelons of society by
offering them opportunities for upward mobility and status change.35 This had particular relevance in Anatolia, where ever-increasing numbers of young peasants who
filled the medrese lodges as students graduated with no prospect of employment in the
Ottoman religious and bureaucratic establishments and resorted to low-profile almstaking by wandering in their localities. This soon deteriorated into chronic banditry,
with student (suhte) bands active in both rural and urban areas until the early seventeenth century.36 Further, those levend mercenaries recruited by rival princes in the
s who subsequently failed to obtain fiefs or kapkulu rank formed brigand bands,
often equipped with firearms.37 Early in the reign of Selim II (), there already
existed a sizeable, dangerously mobile rural population in the Anatolian countryside,
active in low-level but chronic brigandage.

The explosion c.
The number of soldier-brigands with firearms increased during the s and s
as provincial governors formed their own retinues by employing these same people,
the uprooted peasants, as sekban and sarca mercenaries. Contemporary sources indicate that such a trend then accelerated, encouraging more and more peasants to leave
their homes to join such retinues.38 This fatal combination the sultans provincial
agents and their sekban retinues furthered acts of oppression targeting primarily the
remaining taxpaying villagers. Akda estimates that in the s there were some hundreds of thousands of levendat and sekbans active in celali bands, either independently
or in the retinues of provincial governors.39
The general situation of provincial governors themselves also deteriorated rapidly,
due to the high cost of obtaining and maintaining office.40 The resentment of increasingly marginalized members of the askeri resulted in acts of rebellion in the capital and
increasing lawlessness in the provinces.41 Special sultanic decrees (termed by historians
adaletnames, justice decrees) reminded governors of their primary duty to observe
justice and avoid illegal acts of all kinds.42 In practice, however, the breakdown in
authority was a much more complex phenomenon that cut across all sectors of society,
from the palace, through all levels of administrative officials, down to sekbans of reaya
The long wars of the late sixteenth century had imposed extra burdens on the rural
population and economy. That of against Iran proved especially detrimental on account of the war-time obligations upon both peasants and nomads of extra
taxes, labour and animals in provisioning large armies both on campaign and in winter
quarters. Exactions continued for the Hungarian war of , and in the early
s for the celali campaigns against the rebels in Anatolia.43 The imperial treasury attempted to overcome the cash shortage not only by debasing the currency, but
also by regularizing extraordinary levies, the avarz and tekalif,44 and by multiplying
the rate of these taxes and services in kind as well as the poll tax (cizye) paid by nonMuslim subjects. In response, the taxpaying reaya began to resist these exactions by
arming themselves against the tax-collecting agents of provincial governors and other
officials. Meanwhile, the kapkulus of stanbul as well as provincial sipahis also began
to evade military campaigns in ever-increasing numbers.45 The result was dangerous

The reign of violence

militarization of the countryside, with highly mobile and discontented elements at

all social levels. Furthermore, Sam White draws attention to extraordinary climatic
fluctuations which coincided with these developments as a serious external pressure
on the peasantry. To him, the years of drought and famine played a crucial role in the
rise of violence and banditry during the last decade of the century.46
And and And see this development as the consequence of a system of predatory
finance which aimed to maximize revenue for the imperial treasury and the individual
wealth of grandees by creating highly politicized kinship networks that made arbitrary use of the opportunities amply provided for them by the imperial system.47 The
result was the exclusion of certain sections of the agents i.e., the askeri class from
the redistribution mechanism of privileges or political favours, and of the taxpaying
subjects from the imperial system of welfare and prosperity, eventually pushing them
all into the margins of the law. Cook refines the agency problem further by drawing
attention to the same institutional elasticity of the Ottoman ruling elite that left its
agents considerable room for manoeuvre in order to maintain themselves in the zone
between the sultanate and its subjects, without any overwhelming interest in the preservation of the prevailing social structure either in Ottoman Anatolia or elsewhere.48
Theoretically, they functioned as the instrument of Ottoman centralization, but in
practice they acted quite independently with the means (both military and economic)
to reproduce themselves. They did this not necessarily by preserving such basic institutions as the timar, but rather by adapting themselves to changing conditions pragmatically and often at high cost even for themselves. Here, one should note that lowerranking members of the Ottoman ruling elite such as timariots and kapkulu soldiers
were not part of the emerging elites, or of the empires sub-state structure, as And
and And define them. On the contrary, along with the reaya, they were at the losing
end of the gradual institutional dissolution and deeper imperial transformation beginning in the later sixteenth century.


Contemporary sources point unanimously to the crucial impact of the Hungarian
campaign on the escalation of violence and devastation in the Ottoman countryside.
Although victorious at Haova, the Ottoman army struggled on the battlefield and
lost substantial numbers of men. Many more fled and disappeared, leaving the sultan
and the commanding vezirs in a state of shock. A summary roll-call revealed that thousands of Janissaries had deserted and that, according to contemporary sources, some
, sipahi provincial cavalry had evaded the campaign altogether, while others had
sent substitutes. Punishment was severe: both deserters and those who had failed to
appear were to be persecuted wherever caught, and all their ranks, fiefs and income
sources revoked. Thousands of Janissaries and sipahis are said to have fled to the other
side (te yaka) i.e., Anatolia. Most became involved in large-scale banditry, often
forming their own bands from the levends and sekbans already active in the countryside.49 This finally brought home to the imperial administration the seriousness of the
celali movements in Anatolia.50
In , two years after the Haova campaign, and coinciding with a major earthquake in Anatolia, came the first great celali rebellion, which set the precedent for
others throughout the seventeenth century. Contemporary sources mention for the

Oktay zel

first time a rebellious pasha, Hseyin Paa, the governor of Karaman, in association
with Karayazc Abdlhalim, a former police chief and head of a sekban division in
Sivas province, who rebelled in the same year with a local force of some , men
and declared himself ruler of the Urfa region. Karayazcs army defeated the Ottoman
forces sent against him, and roamed the countryside of north-central Anatolia for four
years with such success that the imperial administration sought to buy him off with the
governorate of orum. When he died in , Karayazc had an army of , men
under his command.51 Paradoxically, the Ottoman forces sent against him were themselves composed partially of former celalis and levends, and also behaved like celalis in
the Anatolian countryside, ruining villages and causing great distress to the peasantry.
Karayazc had complained to Istanbul about the oppressive and celali-like behaviour
of the Ottoman commander, a point raised again by Nasuh Paa in .52
After Karayazcs death, his brother Deli Hasan continued to devastate Anatolia
to such a degree that the government attempted to conciliate him with the governorship of Temevar in the western Balkans. Moving with his men to Rumeli, he
subsequently fought with the Ottoman army in Hungary. The most serious rebellion
led by an Ottoman pasha was that of Canbulatolu Ali Paa, governor of Aleppo.
Canbulatolus rebellion had a significant local character closely linked to his familys
near hereditary rule north of Aleppo and to a struggle for regional autonomy; it was
therefore particularly important for the Ottoman administration to win him over,
again by conciliation.53 Although Ali Paa was executed in , various members
of the Canbulatolu family continued to hold high positions in Ottoman service.54
Kalenderolu and other rebel pashas in this period in Sivas, Kayseri, Ayntab and
Aydn commanded armies of , to , men, mostly levends and sekbans.55
The greatest celali armies were led by leaders of sekban origin, with between ,
and , men under their command. Most of these rebellions took place simultaneously in different parts of Anatolia between and and easily defeated the
government forces (also composed largely of Anatolian levends) sent against them.56
At the peak of this devastation and anarchy came the great flight of peasants from
the land (byk kagun, ).57 Some tried to defend themselves by building fortified towns (palankas); most sought refuge in already fortified local cities or fled to
Istanbul, Rumeli, Syria or even Crimea.58 The government response, between
and , was relatively successful in asserting control; according to one contemporary source, around , celalis had been killed by .59 However, this was only a
temporary and partial success. Brigandage by celali bands, in small or large numbers,
and the unlawful activities of provincial officials continued.
The second phase of the celali rebellions broke out in , with the revolt of Abaza
Mehmed Paa, governor of Erzurum and formerly the treasurer of Canbulatolu Ali
Paa.60 Mehmet Paas declared aim was revenge for the deposition and assassination
by Janissaries of Osman II (). With , of his own men, mostly sekbans
recruited from local Turks, Kurds and Trkmen, as well as sipahis, and with support
from many Anatolian provincial governors, he laid siege to Ankara with a force of
,. He was pardoned twice, in and , and, by analogy with the treatment of previous celali rebels, was appointed to governorships in the Balkans before
eventually being executed in .61 Abaza Mehmed Paas revolt was followed by the
local rebellions of Cennetolu around Manisa (s), lyas Paa (governor of Anatolia) around Balkesir to Manisa and Midilli, the Trkmen clan chief Boynuinceli

The reign of violence

Hac Ahmedolu mer around Kayseri (all s), and Grc Abdnnebi,
Karahaydarolu and Katrcolu around Ktahya and Isparta (s). After the shortlived counter terror of Murad IV () came another period of disorder in the
imperial administration and provinces, and in a further phase of rebellion led by
Abaza Hasan Paa, a former member of the imperial cavalry, whose eight-year revolt
destabilized the Anatolian countryside from Kayseri to Aleppo.62 In this rebellion also,
the rebel pasha aimed to intervene in Istanbul factional politics, supporting the vezir
bir Mustafa Paa, who was of Trkmen origin and a nephew of Abaza Mehmed
Paa.63 These later rebellions partially coincided with and were exacerbated by the
lengthy war with Venice over Crete from to .
Sekbans again came to prominence in Anatolia, this time providing the manpower
to challenge the sipahis and Janissaries.64 They remained significant in other local
activities of rebellious seventeenth-century pashas, finally and most seriously during the Hungarian war of against Austrian-led coalition armies following
the Ottoman defeat at Vienna. Violence again erupted in Anatolia, comparable to
that at the beginning of the century. The last celali leader of note was Yeen Osman
Blkba (later Paa), who between and led thousands of sekbans and freefloating levends. Similar to earlier rebel pashas, he was first made governor of Rumeli,
then commander-in-chief of the Ottoman armies against Austria; while in power he
appointed his men to governorships in Anatolian provinces but was executed after a
failed attempt to become grand vezir.65 Disorder reigned in Anatolia for the duration
of the Austrian war, military expenditure produced a major financial crisis in Istanbul
and kul revolts, and the taxation system was overhauled to meet the urgent need for
cash. When peace was agreed at Karlowitz in , the general picture of internal
chaos and disorder, as well as the violence and financial crisis, were strikingly similar
to the conditions of the late sixteenth century, and to the situation in when the
treaty of Zsitvatrok concluded the earlier long war in Hungary.


Mustafa Akdas seminal work on the celali movements was carried out from the s
to the s. Recent research, with access to a greater variety of sources, confirms that
it was a much longer-lasting phenomenon, causing far greater damage, than has been
generally known. Violence became the underlying characteristic at all levels of seventeenth-century politics and society, both in the provinces and in Istanbul. It brought
about changes in the imperial structure, either dissolution and breakdown (as in the
timar and social order) or gradual change and transformation (as in the kapkulu and
household/patronage networks, and in the wider application of tax farming). It also
produced a profound transformation in Anatolian economy and society.
The number of those involved in the celali movements in general is unknown,
though scattered archival and chronicle evidence suggests that, in the first decade of the
seventeenth century, it was between , and ,.66 The mass abandonment
of rural settlements, particularly during the great flight of , left half-ruined,
drastically depopulated villages and towns.67 As administrative records of the period

state formation
political conflicts
administrative &
fiscal centralization
and bureaucratization


struggle for
the throne
(civil war)

Figure .

historical legacy
(Babai, Bedreddin,
kzlba rebellions)





population growth
subsistence crisis
surplus population
climatic change
(little ice age)

kuls and sipahis

(leaders of bands)
rebel pashas
(Great Celalis)

sekbans (mercenaries)
(pulled peasants)

(pushed peasants)

demographic crisis
changing settlement patterns
economic crisis & dispossession of
peasantry (towards rural
proletarianization & dependent peasantry)
migration & re-nomadization
environmental catastrophe
diffusion of military into
rural social fabric & economy
from agricultural to pastoral economy

legacy to 18th/19th centuries

(as army corps,
retinues of
elite, ayan)

(ekiya, haram)


deepening financial crisis

dissolution of timar and kul institutions
change in taxation system (from timar
to tax farming/iltizam)
sekbanization of Ottoman army
emergence of provincial elite or rural gentry (ayan)

celali armies (= rebellions)


banditry (= terror)


The reign of the celalis: causes, symptoms, characteristics and consequences

military (= idari &

askeri) & fiscal crisis
long & exhausting
intra-elite conflicts
& corruption

(exclusion &

militarization of the

(over-taxation &
illegal exactions)


The reign of violence

clearly attest, large tracts of arable lands left uncultivated for years were often acquired
by members of the local askeri class. Years of famine often followed, which combined
with attempts to enforce as much tax collection as possible, furthered the desertion
of villages and drove large numbers of people into hunger, death or banditry.68 Rural
violence meant the disappearance, temporarily or permanently, of a great number of
Anatolian rural settlements and caused also major destruction in some cities targeted
by the celalis.
Rhetorical dramatization of the celali phenomenon in early seventeenth-century
advice literature was not simply exaggeration for effect.69 We can now consult detailed
correspondence (telhis) between grand vezirs and sultans from the turn of the seventeenth century; mhimme and ahkam series of outgoing imperial orders, including
fiscal concerns; registers of the kadi courts; and the annual account books of the great
evkaf or pious endowments.70 Narrative sources have been re-evaluated and subjected
to closer textual and contextual readings.71 These all provide further evidence of the
extent and complexity of the movement. Most significant among these new sources
are the seventeenth-century avarz and cizye tax registers,72 which allow comparison
with preceding tahrir surveys and make it possible to observe changes in the taxpaying
population, in settlement patterns and in the composition of both rural and urban
Anatolian society.73
Between the s and s, a period which includes the most violent phase of the
celali terror and the greatest rebellions, the rural taxpaying population in central, northcentral and eastern Anatolian provinces such as Konya, Bozok, Amasya, Canik, Tokat,
Harput and Erzurum fled to safer areas, either high mountains or urban settlements
with better protection.74 Akda gives dramatic examples from the Ankara region, where
official reports show that, in in the district of Bac, thirty-three out of thirty-eight
villages were totally depopulated, while in the two districts of Haymana over eighty
villages, once amounting to two-thirds of the rural population, had no inhabitants left.
Similarly, in the Afyon region south-west of Ankara, officials found peasants in only
ten villages.75 Gnhan Breki has recently drawn attention to the existence of similar
reports for western Anatolian districts.76 In the Ayntab region, immigration into the
city from its rural surroundings early in the century was followed in the s by peasant flight and abandoned villages, due most probably to Abaza Hasan Paas rebellion.77
Katip elebi, who visited Anatolia in the twelve years prior to , emphasizes particularly the large number of refugees flowing into cities, especially to Istanbul.78 Simeon
of Poland, who traversed Anatolia in the aftermath of the great flight, observed a
similar picture, noting especially that half of the western Anatolian city of Bursa had
been destroyed, burned and depopulated by the celalis; the situation was similar for
Kayseri and Ankara.79 Evliya elebi confirms similar destruction elsewhere in Anatolia
in mid-century.80 Even the English ambassador to Istanbul, commenting in the s,
states that, out of , villages formerly in the entire empire (in ), only ,
were left inhabited in .81 This statement, allegedly based on two surveys, though
perhaps grossly exaggerated, conveys a sense of the contemporary perception of the
degree of destruction and desolation of the Anatolian countryside.82
The tax registers of the s also show a dramatic drop in recorded population
in both urban and rural settlements. In Bozok and Harput in the province of Rum,
the number of rural taxpayers was barely per cent of that recorded in pre-celali
registers: some to per cent of the rural population had disappeared from the tax

Oktay zel

surveys.83 In Manyas, the avarz register of refers to the depopulation of some

quarters of the town as a result of the attacks by the celali bandits.84 It seems that no
significant recovery was observed in the depopulated villages during the rest of the
century, despite sporadic references to some peasants returning to their villages after
ten, twenty or even thirty years, usually to find their lands occupied either by askeris
or by peasant refugees from other worse-affected regions.85 Rare examples of similar
registers from the late seventeenth century suggest that the situation was no better in
the s. The population of the city of Lazikyye (Denizli), for example, decreased
even further (by . per cent) between and .86
Systematic comparison of avarz/cizye and tahrir registers also shows a significant
change in settlement patterns. Some to per cent of pre-celali rural settlements
(villages and cultivated fields, mezraas) were by the s totally ruined; some disappeared for good. The majority of deserted settlements were in the open plains, as
was the case in Amasya.87 This confirms Htteroths findings for the inner Anatolian
plateau, namely the province of Karaman: here the rate of abandonment in mountain
villages was to per cent but in the open plains was around per cent.88 A closer
examination of settlements in the Amasya region shows that the abandoned villages
included almost all the newly founded, smaller-size settlements of the sixteenth century. Similarly, villages established by the gradual settling of semi-nomadic groups
had also disappeared. This can justifiably be seen as a sign of re-nomadization.89 One
might expect such a phenomenon to have been particularly the case in the grey area
or zone of transition between Rum (Anatolia) and am (Syria), the area including the
regions of -il, Mara and Aleppo.90 These were populated largely by semi-nomadic
Turkish, Kurdish and Arab tribes;91 here mezraa-type settlements constituted a larger
proportion of settlements compared with other parts of Anatolia.92 Sources also indicate from the s onwards a north-westward mass movement, a nomadic invasion,
of Boz Ulus, Halep and Yeni-il Trkmen tribes from this particular zone into central
and western Anatolia, all the way to the Aegean coastal areas.93 Simultaneously, some
new villages appeared in different locations, including the higher mountain plateaux.
This indicates one of the most drastic ruptures in the historical geography of Anatolia: a great shift in settlement patterns from open plains to mountains. According to
Htteroth, this period lasted for over two hundred years, until reversal began in the
nineteenth century.94 With constant banditry and heavy exploitation by tax collectors,
security became the main determinant of settlement patterns. The plains were increasingly left to either the poorest peasants or the new iftlik-owning askeris, who formed
the nucleus of an emerging landed aristocracy all over Anatolia.
Other factors add to the picture of critical change. With large tracts of arable land
destroyed or left uncultivated, and with a drastically reduced number of peasants
remaining in villages, a gradual expansion of animal husbandry and a shift towards
a more pastoral economy occurred.95 Frequent years of famine went hand in hand
with disease and pestilence.96 Earthquakes (especially those in and ) and
other natural disasters and extreme climatic fluctuations compounded the problem.97
Although specific detail is lacking, it is clear that such catastrophic conditions must
have affected the birthdeath ratio to the detriment of the former. What the avarz
registers of the s show as a drastic decrease of to per cent in the recorded
taxpaying population can be accounted for partially by such a demographic crisis or
Malthusian scissor.98

The reign of violence

On the other hand, there were perhaps as many uprooted peasants mobile in the
countryside as there were recorded in the remaining villages, even in the mid-seventeenth century. The registers of the s suggest that a significant proportion of the
population was either still in hiding to evade tax registration or was active in brigandage; also, tens of thousands of sekbans were either employed in the ever-expanding retinues of provincial governors or wandering around unemployed as celalis. In all cases
they simply went unrecorded in the avarz/cizye registers. The same sources also reveal
a significant level of internal migration.99 In effect, the Ottoman imperial treasury lost
considerable tax revenue simply by erosion of the imperial tax base through the loss
of taxpaying subjects, as well as through destruction of the agricultural structure and
economy.100 This is mirrored in the case of large imperial evkaf, whose rural income
sources suffered from these developments. Being unable to collect village revenues on
time or in full, they often had to close their soup kitchens.101
Inevitably, the cycle of population dispersal, evasion of registration and tax erosion
further deepened the empires financial crisis. Indicative of this was the extreme instability in the post of chief treasurer, the obvious scapegoat for financial failures. Evidence in contemporary chronicles testifies clearly how deadly a task the job of imperial
treasurer became during the seventeenth century, with frequent changes of tenure as
a result of kapkulu rebellions or of the factional conflicts within the Istanbul ruling
elite, and occasional execution.102
However, perhaps the single most important financial consequence of the celali
movements was its crucial role in the dissolution of two basic institutions of the
classical Ottoman imperial regime, the timar and kul systems. As discussed above,
the initial stages of the breakdown of both institutions were key factors in the formative celali period. The first wave of rebellions accelerated the process, as a result both
of peasant flight and the sipahis own activity as celalis. Howards study of inspection
registers for reveals that there were substantial numbers of vacant and unassignable timars in western Anatolia and more than kl timars or reserve fiefs in the
province of Rum without holders in .103 The villages whose revenues belonged
to these timars were ruined and their inhabitants had dispersed; no one was interested
in them.104 With the sipahis also went the bases of both the provincial administration
and the military system. The former was gradually replaced by tax farming (iltizam),
with similar financial, administrative and military functions.105 Through creating revenue units (mukataa), mostly from former timars, and farming them out to the highest bidders under the iltizam system, the state in effect centralized and monetized
the revenue-extraction system.106 The decline in sipahi numbers also created further
avenues of military employment for a significant proportion of the celalis, recruited
as sekban mercenaries either by provincial governors or by the state. There emerged a
self-perpetuating, semi-institutionalized celali-ism which lasted until the early eighteenth century. This crucial change from timar to iltizam, the fine line between sekban
and celali, and the implications of both aspects for the transformation of the imperial
administration as a whole have yet to be adequately studied.
As for the kul institution, both Janissaries and the sipahs had increasingly rooted
themselves in the Ottoman countryside during the sixteenth century, also contributing significantly to the celali movements, often as leaders of small-sized locally active
brigand bands rather than initiators of big rebellions. These members of the kapkulu
class seem to have benefited most from the celali chaos. Avarz registers of the s

Oktay zel

show a significant number of Janissaries and sipahs of varying ranks, along with many
provincial sipahis, established in hundreds of villages across Anatolia, holding iftliks
in their own names.107 These were either kuls who had acquired lands abandoned
by peasants or former peasants who had obtained kapkulu rank. In other words, it
is highly likely that some iftlik-holding kapkulus were native (yerli) peasants, often
referred to angrily as outsiders (ecnebi) in contemporary sources. However, we see the
same groups in the Balkan countryside also referred to in these seventeenth-century
sources as sons of kul (kul olu) or local Janissaries (yerli yenieri).108 A study of the
local and imperial dynamics behind these similar processes in different parts of the
empire would assist in reaching clearer conclusions on the role of the celali movements
in such a development, particularly in Anatolia.
It is nevertheless certain that most Janissaries in Anatolian villages were men of
devirme origin who had acquired reaya iftliks. The maliye ahkam registers and
the kadi sicils contain many legal cases throughout the seventeenth century where
returning peasants had found their lands occupied by the members of the askeri and
demanded them back.109 There are also cases where such askeris claimed exemption
from the compulsory avarz and tekalif taxes paid by all land-holders and houseowners irrespective of their status, military or reaya. All in all, peasant flight, insecurity, and the appearance of askeri iftliks in villages signify a steady change in the position of the peasantry, who gradually lost much of their former freedom and became
dependent sharecroppers or wage labourers.110


Although socio-economic factors have always been prominent in historical discussion
of the celali phenomenon, some historians have recently proposed an analysis emphasizing the political perspective and the states point of view, focusing either on institutional dissolution, on state-making, or on centralization or consolidation of imperial
power.111 In particular, Karen Barkey has demonstrated masterfully how statebandit
(i.e., the celali leaders) relations developed with a certain degree of flexibility on the
part of the central government and a large margin of negotiation and bargaining.112
Such analyses have contributed greatly to general understanding of the phenomenon
and of the changing nature of the state in the post-classical period. However, their primary focus on the politics of the celali movements overshadows the internal dynamics
of socio-economic conditions and the rural misery which initially produced the main
human source of the movement, the push factors which drove large masses in rural
Anatolia first to the margins of their villages and then out of their rural confines, both
physical and economic.113 It is true that peasants were often utilized by rebel pashas or
other celali leaders for their own agendas, which were essentially different from those
of the uprooted reaya. It is also clear that leaders of the major rebellions often sought
personal ends, bargaining with the imperial centre over their own reincorporation into
the askeri class and abandoning the large majority of their followers or allies to join
together again under another celali leader or as smaller independent brigand bands.
The problem here is that the large, explosive pool of uprooted peasants (and
nomads) is usually treated simply as the followers of the major celali leaders, as instruments of intra-elite conflicts over military-administrative positions and large revenue
sources. As marginalized peasants, they had, perhaps, no clearly defined agenda of col

The reign of violence

lective action, nor a persistent unifying goal other than simply freeing themselves from
demographic-economic pressures and financial injustices. However, they were initially
driven into banditry and rebellion for reasons essentially different from those of the
celali leaders of askeri origin. Seeing them totally in the context of a power struggle in
a declining/dissolving or centralizing empire or in a process of state-making at
imperial level using the states terminology of banditry gives only a partial picture of
the celali phenomenon. Rather, the celali movement represents a social phenomenon
with a dangerously self-destructive character. Without a visible ideology to channel
their modest everyday concerns, or a higher cause to die for zealously, seventeenthcentury celalis became the principal actors in a vicious circle of collective violence, if
not a collective action.114
However, the celali movement was not just a response to objective deprivation, but
also a social action which may well be seen as a rising, if not a peasant rebellion in
the narrower sense.115 While serfs in the feudal west rose either against their landlords
or the centralizing state as their lords principal ally, in a sense Ottoman peasants
did the same, though in their own traditional ways. In the absence of an immediate
landlord, they reacted through banditry and occasional rebellion, taking advantage of
their freedom (both legal and de facto) to challenge conditions partially created by the
state.116 Moreover, while European peasant rebels aimed to better their conditions as
peasants under a landlord, with no option of changing either their status or their class,
Ottoman peasants did have alternatives. Becoming a student (suhte), a mercenary
soldier-brigand (sekban) or simply an outright bandit (celali) meant becoming
disconnected from ones place of origin with no intention of return.117 This alienation
from original social confines and peasant life was perhaps the key factor in the development of chronic banditry.118
In this respect, it is imperative not to equate celalis i.e., former peasants with the
peasantry as a whole.119 Celalis neither acted with a group consciousness nor revolted
in the name of the peasantry as a class. Seeing the movement as artificial with no
proclaimed ally or enemy and no significant ideology120 makes it neither less real nor
asocial. This was a social action, but one of marginalized and excluded groups who
spoke only for themselves. It succeeded in that many celalis were partially incorporated
into the newly emerging system during the seventeenth century as a new sekban military force, and became fully institutionalized in the following century as the troops
of ayan-governors. These bandit-soldiers and their violence, both legitimized and to
some extent controlled by their new patrons, became an integral element of a decentralized imperial system, of what Tezcan calls the second empire, restructured on a
much more collaborative basis with provincial leaders.121
The pejorative tone of the imperial term celali rebel (celali ekiyas) for all forms
of resistance and rebellion significantly blurs the full social character of the rural disturbances in Anatolia. In both contemporary imperial parlance and modern historiography this blanket terminology denotes not only the chronic reaya brigandage but
also the big askeri-led rebellions. Officially, the unlawful brigand-like acts of Ottoman officials-turned-bandits (with their sekbans) were termed oppression (zulm ve
taaddi), whereas similar acts of vagabond levendat groups were generally called banditry (ekavet or fesat). However, just as the distinction between sekban as mercenary
and celali as brigand was ambiguous, so too was the line between the violence committed by either side. As Cook emphasizes,

Oktay zel

tax collection and banditry collapse into the same undifferentiated activity of living off the land, so that whether or not a man is a rebel comes to depend less on
what he does than on the more or less fortuitous fact that he has or has not an
official authorization for his maraudings.122
Despite official attempts at differentiation, celali terminology quickly became the
shorthand description for all kinds of rural revolt and violence, appropriately perhaps
given that the nature of the violence and its destructive impact on rural society was
essentially the same.
To add to the confusion, Ottoman bureaucrats also applied the same celali terminology to the banditry, rampage and highway robbery which had been routine among
some Trkmen, Kurdish and Arab tribal, nomadic or semi-nomadic populations long
before the Ottoman period and continued without much change under their rule.123
We have already seen that semi-nomadic elements were also an integral part of the
celali phenomenon, although their involvement had causes significantly different
from those of peasants.124 Leaving aside their rebellious acts as celalis, banditry as a
criminal act and banditry as a way of life become extremely difficult to separate one
from another. Similarly, the imperial administration was quick to label as celali those
peasants whose legitimate attempts at self-defence by organizing themselves militarily
under a leader (say a former sipahi) against the excesses of an administrative official
(say a governor) got out of hand and became violent and destructive.125 Only in such
limited cases could we speak of a certain degree of resistance or rebellion on the part
of the Anatolian peasants by allying themselves with a celali or bandit chief against
official oppression.126 Also, calling these peasants celali reduces to an absurdity the
distinction between victim and victimizer. Finally, given that the most common characteristic of the acts referred to by the blanket term celali is the undifferentiated and
illegitimate use of violence, those whom contemporary sources often refer to as janissary thugs (yenieri zorbalar) and soldier-brigands of the sekban corps (blk ekiyas)
active throughout the period both in Istanbul and in the Asiatic provinces were also
part of the same celali movement.127
Contemporary sources show that the ideals of justice and the legitimacy of state
power also diminished greatly in the seventeenth century, as the state had serious difficulty in upholding these effectively under conditions of collective violence and intraelite conflicts at the centre. Yet they still remained the only hope for the real losers of
the period, namely the desperate peasants holding on in their villages or returning after
some years or decades, who needed to resort to judicial mechanisms either through
the kadi courts in provinces or the imperial divan in Istanbul.128 The imperial centre
was itself often in a state of desperation. It had to allow peasants to arm themselves
for self-defence against the celali terror;129 it was reduced to hiring peasants and celalitype sekbans to augment the imperial troops sent against celali armies in Anatolia.130
The relationships between state, peasants and celalis were constantly changing. How
was a celali leader or pasha to deal with the celalis under his command once he had
been pardoned and had accepted a position for himself and his immediate associates
within the askeri class? The resulting uneasy relationship often led to a former rebel
pasha being prepared to sacrifice his men (whom he himself now termed celali) simply
by driving them into the most dangerous battlegrounds against, say, Iranian armies in
the East.131

The reign of violence

In conclusion, analysis of the celali phenomenon primarily at a political level falls

short in understanding the full picture. Newly explored sources now allow us to reexamine demographic-economic factors, social deprivation and desperation in rural
Anatolia as essential and significant components of the internal dynamics of the celali
movements as a social phenomenon. Only then will we be able to contextualize other
political and military factors, including the management of violence in operation,
where the same celali group or army fights against the state forces today, against Austrian or Iranian armies tomorrow, and against another celali army the following day, in
accordance with constantly changing positions. As ordinary bandits at local level or as
sekban adventurers, recklessly seeking mercenary employment from one end of Anatolia (or even the empire) to the other, swinging between the positions of celali-turnedsoldier and soldier-turned-celali, their mere physical presence and violence became
the only source of survival and, to some degree, a psychological driving force. In the
absence of a comprehensible target or a common political or ideological cause, all celali
activities became potentially self-destructive both for the perpetrators themselves and
for the targeted groups, primarily the peasantry, as well as for the physical and agricultural environment. The resulting atmosphere of general lawlessness eventually created
its own logic, justification and self-perpetuating character, such that banditry came to
represent not the exceptional in Anatolian social history, but an integral part of the
new routine of rural life, surviving in varying forms and contexts until the beginning
of the twentieth century.132

Celali, the most common and pejorative Ottoman imperial term for bandits and rebels,
originated from the kzlba-related rebellion in the name of eyh Celal in in central
Tezcan : ; cf. Tezcan a, where the author interprets the fiscal crisis of these years
by reference to the dictates of market forces rather than the difficulties faced by the Ottoman
imperial treasury.
Selniki Mustafa Efendi : I, , , ; li : III, ; Hasan Beyzade
Ahmed : I, ; II, .
Koca Sinan Paa telhisleri .
Ibid.: , , .
Ibid.: , , .
Akda ; nalck .
li : III, , ; Peevi : II, ; Andreasyan : .
Barkey : .
Koca Sinan Paa telhisleri : passim.
Arcanl and Kafadar .
zel .
In , faced with resentment among Anatolian sipahis following the execution of his son
Mustafa, Sleyman I ordered that fiefs worth , ake and over be given only to imperial Janissaries, leaving provincial sipahis with smaller timars (Turan : ; Cezar :
zel : ; Barkey : .
Faroqhi b: , .
Akda : , ; Akda ; Cezar : .
Turan ; Yldrm .
Akda ; Cezar : , ; Yldrm .

Oktay zel

Cezar : ; Ycel : Kitb- Meslihil Mslimn, .

Turan : , .
li : III, ; cf. Ycel : Kitb- Mstetb, .
nalck .
zel : ; McGowan : ; nalck . Cf. Tabak for a Mediterranean
perspective on such developments.
zel ; cf. Gndz : .
Cook : ; zel : ; z : .
Cook ; z ; slamolu-nan .
Gm ; cf. Htteroth , .
Peevi : II, , ; Hezarfen ; Adanr : .
Kafadar : ; Yldrm .
Soyudoan : .
For a recent re-evaluation of dissent and popular rebellions in the Ottoman empire, see Barkey
: ff.
Cezar : ; Cook : ; Ko .
Cf. Cook : , and, for the s, Tansel : , , .
Arcanl and Kafadar .
Ko ; cf. Kafadar : .
Akda : .
Jennings, ; cf. Cezar : .
Cook : .
Akda : .
Koca Sinan Paa telhisleri ; li , III: .
Peevi : II, ; , .
nalck a; cf. Uluay : .
Topular Ktibi , I: .
Koca Sinan Paa telhisleri : , .
Akda : ; nalck ; cf. Faroqhi b.
White : esp. ff.
And and And .
Cook : .
Celali historiography rarely considers the fact that these siphis were accompanied by their
peasant retainers (cebeli). Around , there were c., dirlik-holders and , retainers in the Asiatic provinces (Ayn-i Ali Efendi ). The campaign could have produced
some , fugitives of sipahi-related origin. On the reliability of figures given in contemporary sources, see Akda : .
Peevi : II, ; Naima : I, ; Topular Ktibi : I, .
Griswold : ff.
Topkap Saray Archive, Istanbul R (Karayazcs letter); TSMA E (Nasuh Paas letter). For access to these documents, I am most grateful to Gnhan Breki. See also Andreasyan
: ; White : .
Griswold : ; Barkey : .
Naima : II, ; Griswold : .
zel : .
Griswold : ff.; Safi : II, .
Akda ; Griswold : esp. .
Peevi : II, ; Topular Ktibi : I, , , . Cf. Akda : ; Darling : .
Safi : II, ; cf. Topular Ktibi : I, . Murphey : .
Naima : II, .
Peevi : II, ; Naima : II, ff., , , . For an analysis of
the historiography of Abaza Mehmed Paas rebellion, see Piterberg .
Seluk .

The reign of violence

Naima : II, , , , , ; Murphey : .

Akda ; Zakaria : ; Andreasyan, .
Akda : ; Silahdar Mehmed (n.d.).
zel : ; Koi Bey risalesi : .
Simeon : , , , ; Akda : .
White ; Htteroth : .
Abou-el-Haj [] ; z .
Faroqhi a, b; Orbay , b, .
Abou-el-Haj [] ; Howard ; Tezcan ; Piterberg .
McGowan ; Darling .
See esp. Kiel ; zel ; z , , ; Gke . For sample published
avarz registers, see nal ; z and Acun .
See e.g. Naima : II, ; Simeon : ; Andreasyan : ; Ycel :
Kitab-i Mstetab, , ; Koi Bey risalesi : , ; zel : .
Akda : .
Breki : .
Canbakal : .
Katip elebi : ; cf. Akda : .
Simeon : , , ; Andreasyan : .
Evliya elebi [] : II, .
Roe (n.d. []): ; cf. Zinkeisen : ; Htteroth : ; Arcanl and Kafadar
: .
Griswold : .
zel , ; z , .
BOA: MAD . I am grateful to Dr zer Kpeli of Ege University for allowing me to consult
his unpublished article on this register.
zel : ; cf. Akda : .
Gke : , , .
zel : ; cf. Ko : .
Htteroth : .
For earlier works referring to re nomadization during this period, see Htteroth :
; Faroqhi : ; zel : , ; z .
Kafadar : .
Survey registers dated suggest that, in Zulkadriye (Mara) and Aleppo provinces, the
tribal nomadic elements constituted more than half of the total population (Murphey :
For expansion of small settlements and gradual sedentarization of tribal elements in some of
these areas during the sixteenth century, see Htteroth : ; nalck : .
White : ff.
Htteroth : ; nalck : .
Htteroth ; Adanr .
Htteroth : ; Ger ; Orbay . The destructive effects of drought and
famine on crop patterns and daily lives of peasants in the sixteenth- and seventeenth-century
empire are often mentioned in local studies.
See Ambraseys and Finkel . For climatic fluctuations, often referred to as part of the little
Ice Age, see Goldstone ; White .
zel .
zel : .
zel .
Faroqhi a; Orbay .
Orhonlu : ; Koca Sinan Paa Telhisleri : passim.
BOA, Mhimme Defteri (Zeyl) : /; cf. zel : , n. .
Howard : ; cf. Ayn-i Ali Efendi ; Koi Bey risalesi .

Oktay zel
nalck a; on absentee-governorship and iltizam in Ayntab, see Kunt ; Canbakal
: .
Darling, : ff.; Adanr : . Cf. Salzmann .
zel ; cf. Akda : .
Raduev et al. .
zel : , .
McGowan : ff.
slamolu-nan ; Faroqhi : ; Faroqhi ; Barkey .
Barkeys somewhat ahistoric and chronologically vague argument on state-making or centralization in the celali context is less convincing but deserves separate study.
Cf. Griswold : , who emphasizes the depth of despair in Anatolia and the extraordinary
spiritual malaise and physical poverty epidemic in the peninsula.
And and And ; cf. Adanr : .
Cook : ; cf. Akda : ; Akda : .
Goldstone ; cf. Faroqhi .
Faroqhi : .
Cf. Barkey : .
Faroqhi : .
Barkey : .
Tezcan b, ; see also Aksan : , ; Yaycolu .
Cook : . For similar developments in the Ottoman Balkans during the seventeenth to
nineteenth centuries, see Adanr : ; Moutaftchieva .
Soyudoan : .
In areas such as the province of Karaman, however, one might expect that the agricultural
crisis, drought and famine of the late sixteenth century did not differentiate much between
nomads and peasants. See White : ff.
E.g. Cennetolu, who is said to have appeared as the defender of peasants in the province of
Aydn in the s (Uluay : ).
Cf. Barkey : .
Peevi : II, .
Faroqhi ; Adanr : .
Naima : II, , .
Akda : .
Peevi : II, , on attitudes to celali troops in the campaign.
Cf. Adanr : ; Reid .




Ottoman frontiers in the early modern period

Dariusz Koodziejczyk

hen studying the borders of any empire that claims to be universal, one must
differentiate between its imaginary territorial ambitions, often encompassing
the whole universe, and the real limit of influence, dictated by geopolitical as well as
ecological concerns. In their best-selling neo-Marxist book Empire, Michael Hardt
and Antonio Negri stress the crucial importance of borders in the making of any
empire: the sovereignty of Empire itself is realized at the margins, where borders are
flexible and identities are hybrid and fluid. It would be difficult to say which is more
important to Empire, the center or the margins.1 The belief that a border is indispensable for any empire or civilization is also shared by a Polish author who can hardly be
suspected of Marxist sympathies. In his essay on the eastern frontiers of Europe, Jan
Kieniewicz writes: a border first reveals the extent of rule, designates the reach of strivings for hegemony, and is a necessary part of an empire. On the other hand, a border
is the consequence of an axiological process of spatial identification.2
Both quotations describe an empire or civilization which, by acknowledging its
limits, has already lost its faith in the universality of its might and values. The other
side of the coin is represented by Denys Hays description of medieval Europe as an
aggressive Christendom uncommitted to any single continent, which only with time
limited itself to the Latin West: so long as th[is] notion was a living notion, Christendom was potentially the whole earth.3 Against this background we can consider Maria
Pia Pedanis explanation of the Ottoman reluctance to conclude formal treaties with
Christian neighbours: [t]o accept the idea of a border with another country meant
also to recognize the right of the other to exist,4 and therefore of the Ottoman empire
to have limits.
With the conquest of Constantinople in and of the three holy cities of Islam
Mecca, Medina and Jerusalem in , the Ottomans appropriated the universalistic claims of both the Roman empire and the Muslim caliphate. Whether or
not the Ottomans really believed that they could extend their civilization and religion to the utter limits of the inhabited world (some certainly did), such a claim was
often expressed in their propaganda and political language. Even in the late seventeenth century, the Ottoman grand vezir Kara Mustafa Paa challenged the Habsburg
hegemony in Central Europe and deliberately humiliated the ambassadors of France,

Dariusz Koodziejczyk

PolandLithuania and Russia by regarding their rulers as infidels and hence inferior to
the Ottoman padiah. In relation to their Muslim neighbours, the Ottomans stressed
their exclusive right to patronage over Mecca and vehemently denied it not only to
the Shiite Persian Safavids, whom they regarded as heretics, but even to their Sunni
Muslim brethren, the Moghuls of India.


The once powerful vision of Paul Wittek, which credited the spirit of Holy War,
shared by Muslim gazi warriors, for the Ottoman rise from a marchland principality to
a three-continent empire, has been seriously challenged in recent decades, though not
overturned entirely. Colin Heywood, while labelling Witteks view as wildly romantic, admits that many Ottoman soldiers probably still believed in the ever-advancing,
ever-victorious frontier (serhadd-i mansure), at least until the war of , and
perhaps even longer.5 The Ottoman term uc, like the earlier Islamic sugur (Arabic
thughur), denoted a marchland or frontier zone bordering on non-Muslim territory.
Its open, dynamic character stood in contrast to the type of precisely demarcated and
recognized boundary known to the Ottomans as hadd (pl. hudud) or snr.6 Arguing
that in the Ottoman Maghreb the infamous Barbary coast the gazi spirit survived
even after the treaty of Karlowitz in , Heywood identifies such common elements
of frontier realities as the abundance of unemployed young males, especially outlaws
and recent converts, sharing the martial spirit and the prominence of the slave trade
and ransom/redemption systems in the local economy.7
Presenting themselves as good Muslims in order to gain religious legitimacy, the
Ottoman rulers soon became the captives of their own policy. Muslim scholars did
not hesitate openly to pronounce the rulers obligations with regard to cihad. Even the
most peacefully minded sultan was expected to widen the domain of Islam (darlIslam) at the cost of the domain of war (darl-harb), inhabited by infidels and predestined for Muslim conquest sooner or later. Consequently, a lasting peace with
non-Muslims, however politically rewarding, was difficult to attain without violating
Muslim religious law. Apart from the domain of Islam and domain of war, Muslim
legal scholars, especially from the Shafii school, discerned two intermediary categories. The first was the domain of jihad (darl-cihad), referring to territories already
in Muslim hands but still insecure. In the sixteenth and seventeenth centuries the
Ottomans constantly referred to (Libyan) Tripoli, Tunis and Algiers as darl-cihad
vel-harb. In the late seventeenth century this term applied also to Belgrade, endangered by Habsburg troops, and to the recently conquered Polish fortress of Kamieniec
(today Kamjanec in Ukraine).8 In effect, the domain of jihad could be identified
with the early Ottoman uc.
The second intermediary category had a more practical value. This was the domain
of covenant (darl-ahd). According to the Koran (: ), Christians and Jews were
permitted to retain their religion and live in peace on the condition of paying a tribute
(cizye) to a Muslim ruler and accepting his superiority. This applied both to non-Muslim subjects of Muslim rulers and to non-Muslims living under non-Muslim rulers,
where the latter recognized a tributary status vis--vis a Muslim sovereign.9 Today,
textbooks of Ottoman history typically mention four Christian tributary states in
south-eastern Europe as falling into this category, namely Wallachia, Moldavia, Tran

Ottoman frontiers in the early modern period

sylvania and Ragusa (Dubrovnik). Yet, this group was much larger, to mention only
several Christian tributary states in present-day Georgia. According to Gbor goston, a common feature of the frontier territories was the condominium that is the
joint rule of the former power elite and the Ottoman authorities.10 In Ottoman eyes,
such European countries as Venice, the part of Hungary controlled by the Habsburgs,
PolandLithuania and Muscovy were seen, if temporarily, as tributary states. Demand
for tribute even if symbolically small therefore figured prominently in diplomatic
correspondence between Istanbul and neighbouring Christian rulers. If a Polish king
or a Muscovite tsar refused to send tribute directly to the sultan, at least they were
expected to send gifts to his vassal, the Crimean khan.11 The question why the famous
FrenchOttoman alliance of was never officially ratified by Sleyman might be
easily explained by the fact that the sultan could not do so formally unless Franois I
agreed to become his vassal.
The sultans claim to be caliph and political leader of all Sunni Muslims takes the
question further. The ideology of cihad could be used not just against Christians,
but also against the Shiite Safavids, regarded as heretics, and even against Hindu and
Chinese rulers. In various periods, the Ottoman sultans were at least nominally
regarded as sovereigns by numerous local Muslim rulers in Sumatra, India, subSaharan Africa and Central Asia. As late as , the Muslim ruler of Kashgar, Yakub
Bey, placed the name of Sultan Abdlaziz on his coins and pronounced it in the Friday
prayer (hutbe).12
Judging by their religious principles and imperial propaganda, the early modern
Ottomans seemed to claim the entire world, recognizing no fixed borders. Yet, conclusions drawn merely from ideological credos and based on ready-made theoretical
models can be misleading. For instance, Rifaat Ali Abou-el-Haj combined Witteks
view of the Ottoman state as an ever-expanding organism invigorated by the religious
zeal of Muslim warriors with the application of the open frontier concept, developed
by the followers of Frederick Jackson Turner.13 Stressing the state of perpetual warfare
on the pre-Karlowitz Ottoman borders, Abou-el-Haj defined these as rough, vague,
indefinite military zones between the belligerent forces, where neither the Ottomans,
nor even the Europeans, were familiar with the idea of a linear boundary until after
.14 Several critics have disagreed.15 Maria Pia Pedani concluded that, since OttomanVenetian demarcations in Dalmatia dated back into the fifteenth century, by
the Ottomans had already had common borders with the Republic of Venice for
two centuries.16 If there was any novelty in the post-Karlowitz demarcation, it was due
to Marsigli (alias Marsili), the Habsburg head commissioner. As well as using modern
cartography to delineate the HabsburgOttoman boundary, he envisioned a sanitary
cordon, furnished with lazarettos, in order to protect Habsburg territories from future
A useful model of border development was proposed in by Lucien Febvre.
He traced the change from a medieval uninhabited zone, separating sparse human
communities, to an early modern fortified military line, to a mental ditch between
nationalities . . . a moral frontier characteristic of the post- nationalist era.18
However, perhaps the most sophisticated model of the historical development of
European boundaries was proposed by Peter Sahlins with regard to the French
Spanish border in the Pyrenees. Sahlins proposes three rough stages in the development
of his Pyrenean border. In the first stage, which prevailed until the late seventeenth

Dariusz Koodziejczyk

century, royal jurisdiction was over subjects and not over a delimited territory. Territorial boundaries remained unimportant compared to the boundaries of jurisdictional
competency in the borderland.19 The second stage was characterized by the creation of
a military border, envisaged to reach natural, geographical limits, preferably mountains or rivers. Foreign enclaves were purged and sanitary cordons created, as during
the War of the Pyrenees between and .20 For security reasons, borderlands
remained undeveloped and depopulated; few paved roads were constructed, as the
respective governments tried to prevent possible invasions of perfidious and duplicitous neighbors.21 The third and final stage meant, first, nationalization of the state
border, and then politicization of national boundaries and national territory. Vaguely
defined, natural military borders gave way to precise demarcated ones; the border
established by commissioners twisted, turned and zig-zagged, leaving everything
French on one hand and Spanish on the other. The expression territorial violation
and the associated formula of rape used in reference to a border properly belongs
to the Age of Nationalism.22
Having in mind the models of Febvre and Sahlins, we can easily identify the first
primitive stage, characterized both by uninhabited zones, separating human communities, and by the state jurisdiction over subjects rather than a delimited territory,
with the Ottoman uc or the Turnerian open frontier. Yet, long before the Ottomans also knew the terms hudud and snr, referring to demarcated linear boundaries.
As well as the demarcations with Venice already mentioned, the first Ottoman demarcation with Moldavia originated in the late fifteenth century.23 At the opposite extremity of the empire, the Ottomans likewise tried to define their borders. In , following the pacification of with Shah Abbas, Osman II ordered the demarcation
of the boundaries of the eastern provinces bordering on Safavid Iran, namely Basra,
Baghdad, ehrizul, Van and Erzurum, in accordance with the border established by
the treaty of Amasya in .24 The Ottomans also recognized spatial limits to their
power in the south, with Sleyman I referring to Yemen as one of the furthermost
frontier areas of my well-protected dominions.25 By analogy, the port of Muskat was
labelled the frontier fortress of the infidels before its capture from the Portuguese in


Ottoman borders in the east and south are rarely discussed by historians.27 Long
scholarly neglect of the large area of OttomanSpanish confrontation in North Africa
provoked Andrew Hess to term it the forgotten frontier.28 It was the Ottoman land
frontiers with their Christian neighbours in Europe, especially the OttomanHabsburg
frontier in Hungary, that mostly influenced scholarly imagination. Numerous narratives tell us that the almost permanent HabsburgOttoman confrontation had turned
Hungarian lands into an area of constant warfare. Specialized paramilitary formations, such as the notorious akncs and hayduts, devastated and depopulated the country, leaving only Abou-el-Hajs rough, vague, indefinite military zones between the
belligerent forces.
Yet, there is clear evidence of another kind of frontier. Josef Blakovis map of the
Ottoman province of Uyvar (today Nov Zmky in Slovakia) in was drawn on
the basis of a contemporary Ottoman survey register. It clearly shows the Habsburg

Ottoman frontiers in the early modern period

enclaves of Nitra and Leva (Levice) surrounded by Ottoman lands, and also many
Ottoman enclaves dispersed within Habsburg territories.29 Also, the Ottoman kanunname of the province of Uyvar lists specific tolls on the export of oxen, hides and salt to
the Habsburg domain of war, and vice versa, on the import of pottery, glass and iron
products as well as English cloth to the domain of Islam. We learn that construction
timber in Ottoman Hungary originated from the mountains in Habsburg Slovakia
and was floated southward along the Vah river, and, conversely, that millstones in
Habsburg territories originated from the Ottoman province of Uyvar.30 It seems that
constant warfare left much space for everyday local trade. Certainly, the Habsburg
Ottoman border in Hungary was neither rough nor vague for the local inhabitants.
A recent study by Mark Stein confirms that, despite the raids by garrison troops and
the campaigns against the Habsburgs in , as well as an official ideology of
gaza, the century saw little change in the size of the territory controlled by the Porte.
In fact, the Ottomans substantially reduced border garrisons and even considered the
liquidation of the post station (menzil) in Uyvar, as urgency of communication was
not deemed particularly important in peacetime.31
A glance further to the Ottoman north-east brings another surprise. A first demarcation between the Ottoman empire and PolandLithuania was ordered in ,
following the treaty of , although the appointed commissioners failed to meet in
the steppe due to mutual distrust. For almost another century, the border remained
open, though both sides tacitly admitted that it ran along the rivers Dniester, Jahorlyk and Kodyma. Strangely, the lack of formal demarcation did not spoil the friendly
relations which lasted for most of that period, with the exception of the Polish
Ottoman war of . The first successful demarcation occurred shortly thereafter,
in , confirming the actual border and even moving its eastern section slightly
southwards in favour of PolandLithuania.32
Most striking and seemingly unusual is the PolishOttoman demarcation of .
In the Ottoman army conquered the Polish fortress of Kamieniec, along with
the province of Podolia. It took another eight years of campaigns and diplomatic
disputes until both sides agreed to demarcate the new border. This demarcation,
effected between August and October , provides us with the richest documentation regarding border-making in the whole history of PolishOttoman relations.
Both sides prepared in advance impressive documentary and historical evidence in
order to further their claims.33 Two official protocols survive, one in Turkish and
one in Polish.34 When compared with Sahlinss theoretical model of border development, this demarcation, though almost two centuries older, surprisingly fits the third,
most advanced stage, embodied by the FrenchSpanish demarcation of . In the
Pyrenees in : The first stone is situated on the northeast side of the road from
Puigcerd to Llvia, at the site called Pontarro de Xirosa, next to the old stone which
had been the boundary of Llvia, Ur, and Cldegues.35 In Podolia in : An earth
mound was raised as a boundary marker on the top of a hill covered with an oakgrove, situated in the kble direction from the road leading from Jazlivec and passing
at the said place.36 The only notable difference is that the nineteenth-century border
stones were numbered and the distance between them was recorded in metres, while
in the seventeenth century the distance between boundary mounds was measured in
hours of walk. A memoir by a Polish participant provides a vivid description of this
























Zapor ozhia







U kr aine


Black Sea

Oz qalesi





Ottoman north-eastern frontiers in Europe, c.



R. Kodyma












(Kamieniec Podolski)

Pod ol ia

(Alba Iulia)









R. Ingul


R. P

R. Kuban

K uba n


R. Mius





R. D


Ottoman frontiers in the early modern period

When it came to raising a mound, the Turks, using spades attached to their saddles, in the twinkling of an eye raised a mound of turf after digging around a big
oak trunk [that was] in the middle. Then, after finishing the job their superiors
climbed on top of it and ululated like dogs with their faces turned up, praising
God that they had conquered so much with the sword.37
The most precise demarcation in the long history of PolishOttoman relations therefore occurred not after the Karlowitz treaty, but before. The extant protocols of the
post-Karlowitz demarcation, restoring Podolia to Poland, are much less precise
than those from . The commissioners agreed that settlement in the border area
should be restricted. As in the Pyrenees around this time, there was a shared predilection towards natural frontiers, in this case the River Dniester.38
This apparent regress observed between the OttomanPolish demarcations of
and is not the only challenge to the neat, positivist model of border development. During the very same demarcation of , while the western, Podolian section of the PolishOttoman border was meticulously delineated, its eastern section,
dividing Poland from Dnieper Ukraine (ruled autonomously by the Cossacks, who
had recently acknowledged Ottoman suzerainty), was left vague and open. This was
not by accident. Writing to his envoy in Rome, the Polish king expressed his hope for
a quick return of the Cossacks under Polish patronage, concluding that to demarcate
Ukraine from Poland would be premature, as whomever the Cossacks choose, [that
ruler] will keep the Ukraine as well.39 The notion that royal jurisdiction over subjects
had primacy over territorial control, typical of the first, primitive stage in Febvres
typology, coincided with elements typical for the second, military, or even the third,
linear stage within the very same demarcation.40
The above statement is also valid in respect of Ottoman demarcations with
Russia. In , when the first formal OttomanRussian treaty was negotiated in the
Crimean capital Bahesaray, neither side undertook any demarcation but deliberately
left the border zone empty, forbidding settlement on either side of the bordering River
Dnieper.41 Only after the OttomanRussian negotiations of was a demarcation successfully effected, in . However, unlike the Ottoman post-Karlowitz
demarcations with Venice, the Habsburgs and PolandLithuania, the demarcation
with Russia did not leave any physical evidence in the terrain. In their final protocol
the commissioners clearly stated: we have confirmed the border by no other signs
but this written testimony.42 And yet, notwithstanding the apparently more primitive character of OttomanRussian demarcations, already in both sides knew to
further their arguments by such modern devices as a printed map of German press,
supplemented with handwritten corrections by a member of the Russian delegation.43
Seventeenth-century Russians were also familiar with quarantine requirements that
even affected the Russian embassy returning from the Crimea after the peace negotiations in . While the envoys were detained for one month in the border town of
Borysov-Gorodok, the contents of their papers were dictated through the wall of fire
and smoke so that they could be immediately known in Moscow.44
The fact that in the very same year the Ottomans demarcated one section of their
border (Podolia in ) and refrained from demarcating another section (Ukraine in
from Poland and in from Russia), or that a meticulous border demarcation
with Poland in was replaced twenty-three years later by a much less detailed one

Dariusz Koodziejczyk

cannot be explained simply by a rise and fall of the spirit of holy war, or by a fluctuating level of maturity of Ottoman state structures. Perhaps an explanation lies in
the ecology of the delimited lands. While more densely populated, agricultural areas
such as Dalmatia, Hungary or Podolia were demarcated already in the sixteenth and
seventeenth centuries, the Ukrainian Black Sea steppes waited for demarcation until
the Russian conquest and colonization in the late eighteenth century.45 The East European steppe, like North African deserts or Caucasian and Yemeni mountains, enjoyed
looser state control which enabled the preservation of so-called tribalism, tolerated by
the Ottomans.46 Even in the nineteenth century, under modern French administration, Algerian frontiers with Morocco were demarcated only as far as Teniet sidi-Sassi,
around km from the Mediterranean. Beyond this extended the barely controllable
pays du fusil.47
The Ottomans ability to effect demarcations did not rule out their attachment to
the gaza spirit. Two Ottoman copies exist of the meticulous protocol of the Polish
Ottoman demarcation of . One, provided with seals and signatures by the Ottoman commissioners, was given to their Polish peers.48 The second was recorded in the
survey register of the Ottoman province of Podolia.49 Contrary to expectations, this
second copy, intended merely for domestic use, is much more rhetorical. Mehmed
IV, for whom the conquest of Podolia in was the first campaign in which he
took part in person, is referred to in the second copy as the father of victories and
gazas. Moreover, the second copy begins with a religious preamble that is missing in
the first copy. This expresses the omnipotence of God and the temporary character of
all human borders, quoting a saying of the Prophet Muhammad which promises that
sooner or later all the lands of unbelievers will be open to Muslim warriors. The implication is that one should not treat borders too seriously, since only God may hand out
kingdoms to the rulers in this world.50 The preamble omitted from the Polish copy
also refers to giaours [who,] having perverse ideas and hostile to the manifest faith,
flee consistently and incessantly from their solid castles, fortresses, and forts. The
ideological preamble could well serve internal Ottoman propaganda, but clearly was
considered inappropriate when dealing with foreign partners.
A parallel example is provided by the document of the post-Karlowitz Ottoman
Venetian demarcation, dated . This demarcation was the result of prolonged
negotiations between the commissioners of the two states. Yet, the preamble of the
Ottoman copy, produced by a chancery anxious to conceal the unpleasant contents of
the Karlowitz treaties, gives the impression that this was a unilateral privilege granted
by a benevolent sultan at the humble request of the Venetian infidels.51 The topos of
defeated barbarians, humbly asking for peace regardless of the actual power relationships, was not an Ottoman invention. The ancient Romans, when forced to pay tribute to the Huns, presented this tribute as a voluntary subsidy and even granted Attila
the honorary title magister utriusque militiae.52
To sum up, while the ideology of holy war could still be alive, at least among some
segments of Ottoman society, and served well internal propaganda, everyday praxis as
well as political, logistical and geographical limitations obliged the Ottomans to adopt
a pragmatic and flexible approach in their dealings with neighbouring states. Interestingly, such flexibility is also discernible in their dealings with subjected territories
already situated within the Ottoman frontiers.53

Ottoman frontiers in the early modern period


One such territory was Yemen. Conquered between and , and reconquered
in , it remained under loose Ottoman control only until the early seventeenth
century, to be conquered again in the nineteenth century.54 The Ottomans were very
conscious of Yemens distant location, and service there was extremely unpopular
among Ottoman soldiers.55 Numerous pious foundations, preserved today in Yemen,
prove that the Ottoman governors tried hard to gain local support and popularity. In
fact, the conquest of lowland Tihama, inhabited by Sunni Shafii Muslims, went easily after the Ottomans eliminated the local Tahirid dynasty. Yet, in the mountainous
north they encountered the Zaydi imams, who claimed their own leadership over the
people of Islam, and sometimes even the caliphate. Robert Serjeant, commenting on
the troubles experienced in Yemen by all foreign invaders, from the Abbasids to the
Egyptians in the s, observed that one broad pattern easily discerned is the entry
of foreign conquerors from the lowlands, their initial success but ultimate inability to
conquer the northern highlands and their eventual retreat.56 An early seventeenthcentury English traveller claimed that the Ottomans kept , soldiers in Yemen
fighting continuously in the field against the Arab king in the mountains.57 To quote
a modern French scholar, it was la rencontre de deux lgitimits.58
Ottoman Turkish as well as pro-Ottoman Arab chroniclers employed a rich vocabulary to prove the legitimacy of Ottoman claims to Yemen and to discredit and dehumanize the enemy. The Zaydi imamate forces were referred to as bastards (haramzadeler), the enemies of religion (ada al-din), heretics (ahl al-ilhad), idle troops of

Figure . The mountain fortress of Thula, seat of the sixteenth-century Yemeni leader Imam
al-Mutahhar. The mountain visible behind the walled town is topped with a fort provided with
cisterns and a subterranean granary. The fortress proved off-limits for Ottoman troops of the
period, even though they were equipped with heavy guns. marcus wilson-smith/Alamy.

Dariusz Koodziejczyk

the devilish army (jaysh al-batil min al-jundil-shaytani) and, perhaps most typically,
insects (hasharat).59 Yet, sooner or later subsequent Ottoman commanders learned
that Yemen was unconquerable. Then another policy and language had to be adopted.
The pro-Ottoman Arab chronicler al-Nahrawali related the negotiations following
the futile seven-month siege of the Yemeni fortress of Kawkaban by the Ottoman
army between September and April . After negotiations began, the Ottoman pasha entrusted his scribe with writing a letter to the local Zaydi commander,
Muhammad bin Shams al-Din, addressing him as my brother (ya akhi) and notifying him that, though the Ottomans are not in need of these lands, since they rule
most of the inhabited world, they should still be obeyed. Therefore, continued the
scribe, the pasha will not reject your request but will stipulate that the sermon and
currency (khutba wa sikka) be in the name of the sultan.60 In the subsequent negotiations, Muhammad bin Shams al-Din was granted attributes of power: a standard
and an imperial diploma (berat), according to whose tenor he admitted his error and
prostrated himself before the Ottoman might, allowing the Ottoman pasha to make
him subject to pacts and agreements and impose on him conditions and limits.61 In
short, he was granted his own lands by the Ottoman sultan on the sole condition that
he would officially accept Ottoman suzerainty. Similar agreement was then made with
Muhammads uncle, Imam al-Mutahhar.62

Figure . Shahara: suspended bridge leading to a mountain village-fortress, from which

Imam al-Qasim successfully defied the Ottoman sultans claim to control northern Yemen
in the early seventeenth century. Jack Jackson, Getty Images.

Ottoman frontiers in the early modern period

The pacification of Yemen did not last long, and soon another Zaydi pretender,
Imam al-Qasim, rebelled. He concluded peace treaties with several Ottoman pashas,
in , and .63 His territories first encompassed only the mountain fortress of Shahara, but then also Saada and al-Hayma. His son, Imam al-Muayyad,
continued the struggle against the Ottomans, and by the last Ottoman garrisons
evacuated Yemen. Paradoxically, at no time did [the Ottomans] abdicate their sovereign rights over the land, for Yemenis still looked to them and their agents for protection against foreign especially European danger: ironically, the imam sought
autonomy only.64
Yemen was not alone. Baki Tezcan quotes a document of Sleyman from around
, granting hereditary rule, land property rights, tax exemptions, and wide autonomy to the Kurdish emirs in return for their loyalty to the Ottomans. The document
is referred to as a muahede (identical with anahdname, granted also to foreign rulers)
and is corroborated by the sultans solemn oath.65 Some Kurdish territories, which
retained their autonomy until the nineteenth century, were situated deep inside the
Ottoman domains and hence cannot be labelled as frontier zones. There were many
more Muslim local rulers whose position was analogous. The Crimean khans, the
sherifs of Mecca, Caucasian lords and some provincial notables, such as the Lebanese
Mans, enjoyed local autonomy long before the eighteenth-century era of the ayans.
These minor rulers were often accused of duplicity as the Ottomans never accepted
their full sovereignty.66 Muslim Yemeni Arabs were labelled similarly to Christian
Montenegrins, as mountain bandits (ekiya-i cebeliye).67
Maps depicting the early modern Ottoman empire also tend to leave a false impression that the Ottomans controlled much larger territories than they actually did. Did
such desert lands as Fezzan, Nubia and Arabia really belong to the Ottoman empire?
Or, rather, should the Ottoman control of these lands be compared with their control of the seas? After the final Ottoman conquest of Crete in , the Ottomans and
Venetians agreed that sea waters belonged to a fortress within cannon range. On the
other hand, the Ottomans claimed full control of the Sea of Marmara, the Black Sea,
the Aegean Sea, the Red Sea, and large sections of the eastern Mediterranean, though
this control was often problematic in practice.68


At first glance, Poland and Yemen do not seem to have much in common within the
Ottoman worldview. Yet, at critical points in the seventeenth century, their history
was interrelated. Haji Ali, the contemporary Ottoman chronicler of the last decades
of Ottoman rule in Yemen, describes how in the Ottoman governor Fazli Paa
waited desperately for reinforcements from Istanbul. The chronicler explains that no
help could arrive because of the janissary rebellion and the murder of Osman II, which
was an indirect result of the latters unsuccessful Hotin campaign against Poland
Lithuania in .69 When assistance finally arrived, it was probably too late, and
the province was evacuated by . In the Porte contemplated a reconquest
of Yemen.70 Again, these plans were frustrated by a more urgent campaign against
There was another element that linked Yemen and Poland: their rulers concluded
peace treaties with the Porte and obtained ahdnames from the Ottoman sultans. One

Dariusz Koodziejczyk

is tempted to ask, somewhat provocatively: what makes a modern historian resolve

that Yemen lay inside the Ottoman empire, while Poland and perhaps Venice as well
lay outside? If judged by Ottoman propaganda, their rulers the Yemeni imam, the
Polish king and the Venetian doge were all Ottoman vassals. If judged by the realities, all these countries preserved sovereignty but could not entirely ignore Ottoman
diplomatic pressure and military threats. Ottoman influence was at best disputable,
often negligible, though at times the rulers concerned knew how to use the Ottoman
umbrella against their other enemies.
Analogously, one could ask why modern historical maps depict the Kurdish principalities, such as Bitlis, as regular Ottoman lands, while the Crimean khanate and
Moldavia are depicted by stripes signifying their autonomy? In the seventeenth century all their rulers received Ottoman berats, confirming their position. It seems that
the more one studies the Ottoman realities, the more stripes one sees. Like any other
early modern rulers, the Ottoman sultans faced constant challenges to their legitimacy
as distributors of justice, providers of security, and last but not least Muslim gazis.
Negotiation with his own subjects, not just with external monarchs, was a part of the
sultans mtier.
We have tried to demonstrate the difference between the imagined Ottoman frontiers, promoted by Ottoman propaganda and reaching to the furthest corners of the
inhabited world, and the real borders of Ottoman control, sometimes meticulously
demarcated, sometimes not. Yet, there is one more aspect. In the models of boundary
development formulated by Febvre and Sahlins, both authors stress the importance
of emotional attitude towards ones own territory and its borders, characteristic of the
third, most advanced stage. Did the Ottomans have any emotional attitude towards
their borders? Where did they feel at home?
Certainly not in Yemen, labelled as the tomb of the Turks.71 Not in Podolia, where
Ottoman soldiers cured their homesickness with opium and vodka, as evidenced by a
poem left by one garrison member.72 Interestingly, Ottoman Hungary was not considered a proper Ottoman land either, though it was directly ruled for almost one and a half
centuries and densely garrisoned. The Ottoman chronicler Selaniki recorded a reported
comment by the grand vezir Mehmed Sokollu upon returning from the Szigetvr campaign in . Having crossed the River Sava, he announced: from now on it is again
the inner land (imden ger i-il dir). In the French translation of Gilles Veinstein,
the phrase sounds even more persuasive: maintenant, nous sommes rentrs chez nous.73
For the seventeenth-century Ottoman traveller Evliya elebi, not only Ottoman
Hungary, but even Bosnia, situated on the hither i.e., southern shore of the Sava,
constituted the borderland, extending between Belgrade and the Habsburg lands.74
Moreover, in an Ottoman fiscal official from Uyvar reported the necessity of
importing iron from another Ottoman fortress, Varad (today Oradea in Romania),
assuring his superiors that transport was easy although the road passed through nonMuslim regions. Stein, who studied this report, concluded that the term non-Muslim
regions must refer to Transylvania.75 Notwithstanding the fact that the ruler of Transylvania was an Ottoman vassal, maps show that the road from Varad to Uyvar did
not run through Transylvania, but through the Ottoman provinces of Eger and Buda.
Referring to them as non-Muslim regions sounds quite odd and must reflect either
the geographic ignorance of the writer or his feeling of alienation in outer Ottoman
provinces, apparently exotic for a Turkish-speaking Muslim newcomer.

Ottoman frontiers in the early modern period

It now seems that the borders of the Ottoman world ran closer to the imperial centre than the furthest extent of Ottoman military conquests, and certainly much closer
than the extent of Ottoman imperial claims. The Ottomans certainly tried to domesticate the surrounding space. Mushrooming foundations of mosques, hammams, or
simply fountains, built in the Istanbuli style from Podolia to Yemen and from Algiers
to Azov, not only served the imperial propaganda or self-promotion of their founders,
but helped them to become accustomed to new surroundings, climate, customs and
language encountered on the street in short, to start feeling at home.


A preliminary version of this essay, entitled Do universal empires have frontiers? An Ottoman
case, was presented in Warsaw in October at the mid-term meeting of the international
research project Tributary empires compared: Romans, Ottomans, Mughals, and beyond,
sponsored by the COST programme (European Cooperation in the Field of Scientific and
Technical Research). The author also wishes to thank the Andrew W. Mellon East and Central
European Fellowship Program for enabling him to conduct short-term research at the American Institute for Yemeni Studies in Sanaa.
Hardt and Negri : .
Kieniewicz : .
Hay : .
Pedani : .
Heywood : .
Ibid.: , .
Heywood : .
Panaite : ; Heywood : ; Koodziejczyk : , , , , , ,
, , (for Maghreb); Koodziejczyk : pt. I, , .
Mawardi : ; French trans. : .
goston : .
Koodziejczyk .
Davison : .
Turner ; Abou-el-Haj .
Abou-el-Haj : , and n. .
Koodziejczyk : ; Veinstein a: ; but see Heywood : .
Pedani : .
Gherardi : ; Marsili ; Stoye : .
Febvre : ; English trans. : .
Sahlins : , , .
Ibid.: .
Ibid.: , .
Ibid.: , .
Gemil .
Feridun Bey [] : ; Kl : .
Blackburn : . Also the province of Lahsa (al-Hasa) in the Persian Gulf was regarded as
a frontier bulwark against the Portuguese and Safavids (Mandaville : ).
Orhonlu : , , facs. . On the still under-researched south-eastern Ottoman provinces, see zbaran and Casale .
But see Peacock , which appeared too late to be taken into account in the writing of the
current essay. Although a number of articles in that collective volume intend to fill the gap,
some by means of archaeological surveys, Peacock nonetheless admits in his introduction that,
although the Ottomans frontiers with the Hapsburgs are quite well known [. . .], the Otto-

Dariusz Koodziejczyk

man frontier with the Safavids and the Russians has been much less studied [. . .], while the
Ottoman frontier with the Funj is almost entirely unknown (Peacock , ). On Ottoman
Hungary, see also the essay by Gabor goston, chapter in this volume.
Hess .
Blakovic : map I.
Kabrda : .
Stein : , .
Koodziejczyk : , and map .
In a letter to his envoy in Rome, the Polish king John III Sobieski mockingly compared Ottoman negotiating methods with the aggressive style of Louis XIV: letter to Micha Kazimierz
Radziwi dated in Jaworw on August ; Warsaw, Archiwum Gwne Akt Dawnych,
Archiwum Radziwiw, dzia III, listy Jana III Sobieskiego, sygn. . I would like to thank
Kirill Kocegarov for drawing my attention to this fragment.
Koodziejczyk : , and map .
Sahlins : .
Koodziejczyk : .
Ibid.: .
The Turkish protocol invokes the saying of a Polish commissioner: God (may He be exalted!)
has separated the lands of Moldavia from our Polish lands by the River Dniester, while in the
Polish protocol, composed in Latin, the same sentence reads: Inter nos et Valachiam ipse Deus
flumine Tyra dislimitavit (Koodziejczyk : , and map ). Strikingly similar
wording was used in by a member of the Aragonese Cortes, who maintained that God
created the Pyrenees to free the Spaniards from the French (Sahlins : ). Also the famous
phrase Il ny a plus de Pyrnes!, reportedly exclaimed by Louis XIV after the death of the last
Spanish Habsburg, fits perfectly within the above context.
Sobieski to Radziwi, dated in Jaworw on August (see note above).
A methodological proposal for solving the apparent contrast between linear and cyclical narratives of history was formulated by Reinhard Koselleck (: ), who stressed the uneven
pace of change in various areas of human experience (political, economic, psychological, etc.)
and the simultaneous presence of different time strata in any given period.
The original Ottoman document, drawn up in January and confirmed in February by
Grand Vezir Kara Mustafa Paa, is held in Moscow, Rossijskij gosudarstvennyj arxiv drevnix
aktov, fond , opis , no. . The sultans ahdname, confirming the peace conditions, was
issued in Edirne in late April or early May ; see ibid., fond , opis , no. ; Feridun Bey
[] : , contains another copy.
Ne inymi znakami, no tokmo [. . .] na six pismax tu granicu [esmy] utverdili; the Russian text is
published in Polnoe sobranie zakonov Rossijskoj imperii, s goda, vol. IV: (St
Petersburg, : , esp. ); the contemporary Italian translation reads: questo stabilimento non lhabbiamo determinato con montoni o qualchaltri segni, ma solamente con questinstrumento di scrittura; cf. Paris, Archives du Ministre des Affaires Etrangres, Mmoires et
documents, Turquie, tome , fol. ab, esp. a.
Spisok s statejnago spiska Velikago Gosudarja Ego Carskago Velicestva poslannikov: stolnika
i polkovnika i namestnika perejaslavskogo Vasilja Mixajlova syna Tjapkina, djaka Nikity
Zotova, ed. N. Murzakevic, in Zapiski Odesskago obcestva istorii i drevnostej, vol. , otdelenie
vtoroe i tretie (): , esp. .
Ibid., .
William McNeills Turnerian approach, crediting the successful taming and colonization of
the Black Sea steppe for the rise of the Russian empire, is very convincing in this respect
(McNeill ). Nevertheless, his treatment of Hungary as a part of the East European steppe
was vehemently and with good reason opposed by Hungarian scholars (Dvid and Fodor :
xxii). It is precisely the difference between Hungary and the Black Sea steppe that resulted in
the different character of Ottoman borders with the Habsburgs and Russia.
In India, dense forests played an analogous role to steppes, deserts or mountains in the Otto-

Ottoman frontiers in the early modern period

man context, preventing the expansion of early Muslim states in Bengal (Eaton : xxiv).
Tayeb : .
Koodziejczyk : and facsimile XXIVae.
Koodziejczyk : pt. , , ; pt. , .
Ibid.: pt. , .
Venice, Archivio di Stato, Documenti turchi, busta , doc. ; cf. Pedani a: ,
no.; on the Ottoman chancery efforts to conceal the effects of military failure in , see
Koodziejczyk .
Whittaker : .
z : .
Farah .
al-Maddah []: ; Soudan : .
Serjeant and Lewcock : .
Bidwell : .
Blukacz : ; Van Donzel : .
The epithet haramzadeler was constantly used by the sixteenth-century Ottoman chronicler
Rumuzi (Hathaway : ). All the remaining epithets are taken from the Arabic chronicle of his Hijazi contemporary al-Nahrawali (al-Nahrawali [] : , , and
; English trans. : , , translated by Smith as devilish troops of the false army
and ); on hasharat, see also Farah : .
al-Nahrawali [] : ; English trans. : .
Ibid.: ; English trans. : .
Ibid.: ; English trans. : .
Soudan : ; Yahya bin al-Husayn, the seventeenth-century Yemeni chronicler and
grandson of Imam al-Qasim, referred to all these treaties as al-sulh, a term equally fitting to the
domestic or international peace; see Yahya bin al-Husayn, Gayat al-amani fi akhbaril-qutrilyamani (Cairo, [] ), pt. , , , .
Farah : , .
Tezcan : , . See also the essay by Nelida Fuccaro, chapter in this volume.
Matthee : . Analogously, the Tatars, Moldavians and Ukrainian Cossacks were
often labelled as traitors by the Ottoman, Polish and Russian authorities, each claiming to be
their sovereign.
Reinkowski : , n. .
Pedani : and ; Koodziejczyk .
Haji Ali, Ahbarl-yamani, Istanbul, Sleymaniye Library, MS Hamidiye , fol. a; cf.
Hathaway : and , n. .
Hathaway : .
From a proverb referring to Shahara, the mountain seat of Imam al-Qasim (d. ) (recorded
in the Lonely Planet guidebook Arabian Peninsula, Hong Kong, , p. ).
The poem by a certain Hasan was published from a British Library manuscript by Orhan aik
Gkyay (); the term horlka, not known to Gkyay, was a Turkish loanword from the
Ukrainian term for vodka. On the everyday life of garrison members in Ottoman Podolia, see
Koodziejczyk : .
Veinstein a: .
Moacanin : .
Stein : .



The case of Ottoman Hungary

Gbor goston



n August , at the battle of Mohcs in south-western Hungary, Sultan

Sleymans army of , to , men annihilated the badly organized and
obsolete Hungarian royal army of , to , men. The battle of Mohcs proved
one of the most important events in European history of the early sixteenth century,
since it led to the direct confrontation of the Ottomans and Habsburgs, the two superpowers of the time in East-Central Europe. King Louis II () of Hungary and
Bohemia, along with most of the magnates and prelates of Hungary, perished in the
battle. Although the sultan