(Cambridge Studies in Islamic Civilization) Milka Levy-Rubin-Non-Muslims in The Early Islamic Empire - From Surrender To Coexistence-Cambridge University Press (2011)

Non-Muslims in the Early Islamic Empire
The Muslim conquest of the East in the seventh century entailed the
subjugation of Christians, Jews, Zoroastrians, and others. Although
much has been written about the status of non-Muslims in the Islamic
empire, no previous works have examined how the rules applying to
minorities were formulated. Milka Levy-Rubins remarkable book
traces the emergence of these regulations from the first surrender
agreements in the immediate aftermath of conquest to the formation of
the canonic document called the Pact of Umar, which was formalized
under the early Abba4sids in the first half of the ninth century. What the
study reveals is that the conquered peoples themselves played a major
role in the creation of these policies, and that these were based on longstanding traditions, customs, and institutions from earlier pre-Islamic
cultures that originated in the worlds of both the conquerors and the
conquered. In its connections to Roman, Byzantine, and Sasanian traditions, the book will appeal to historians of Europe as well as Arabia
and Persia.
Milka Levy-Rubin is a Lecturer in History and Middle Eastern Studies
at The Hebrew University of Jerusalem and Curator of the Humanities
Collection at the National Library of Israel. She is the author of The
Continuatio of the Samaritan Chronicle of Abu4 l-Fath al-Sa4mir alDanafi (2002).
Cambridge Studies in Islamic Civilization

Editorial Board
David O. Morgan, Professor Emeritus, University of Wisconsin-Madison
(general editor)
Shahab Ahmed, Harvard University
Virginia Aksan, McMaster University
Michael Cook, Princeton University
Peter Jackson, Keele University
Chase F. Robinson, The Graduate Center, The City University
of New York
Published titles are listed at the back of the book.

From Surrender to Coexistence
MILKA LEVY-RUBIN
The Hebrew University of Jerusalem
and the National Library of Israel
cambridge university press

Cambridge, New York, Melbourne, Madrid, Cape Town,
Singapore, So Paulo, Delhi, Tokyo, Mexico City
Cambridge University Press
32 Avenue of the Americas, New York, ny 10013-2473, usa
www.cambridge.org
Information on this title:www.cambridge.org/9781107004337
Milka Levy-Rubin 2011
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2011
Printed in the United States of America
A catalog record for this publication is available from the British Library.
Library of Congress Cataloging in Publication Data
Levy-Rubin, Milka, 1955
Non-Muslims in the early Islamic Empire :from surrender to coexistence /
Milka Levy-Rubin.
p. cm. (Cambridge studies in Islamic civilization)
Includes bibliographical references and index.
isbn 978-1-107-00433-7 (hardback)
1. Dhimmis Legal status, laws, etc. Islamic Empire History.
2. Religious minorities Legal status, laws, etc. Islamic Empire History.
3. Islamic Empire Ethnic relations. 4. Minorities (Islamic law)
5. Islam and state Islamic Empire History. 6. Islam and politics Islamic
Empire History. 7. Religious tolerance Islamic Empire History.
8. Islamic Empire Politics and government. I. Title.
ds36.9.d47l49 2011
305.6095609021dc22
2010043663
isbn 978-1-107-00433-7 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of urls
for external or third-party Internet Web sites referred to in this publication and does not
guarantee that any content on such Web sites is, or will remain, accurate or appropriate.
In memory of my father
Contents
List of Illustrations
Preface
page xi
xiii
Introduction
1. The Roots and Authenticity of the Surrender Agreements
in the Seventh Century
Treaties before the Muslim Conquest
Local Surrender Agreements during the Roman
and Byzantine Period
Surrender Agreements Made following the Muslim Conquest
Conclusion
2. Shuru4t Umar and Its Alternatives:The Legal Debate over

the Status of the Dhimm s
The Dating and the Formation of Shuru4t Umar
Legal Discussions throughout the Eighth and Ninth Centuries
The Composition of the General sulh Documents
Conclusion
3. The Date and the Ideology of the Ghiya4r Code

The Sources regarding Umar b.Abd al-Az3zs Edict
The Ideology behind the Edict
The Date of the Adoption of Iranian Dress Codes
4. The Enforcement of Shuru4t Umar

Restrictions upon the Dhimm3s prior to al-Mutawakkil
The Restrictions Issued by al-Mutawakkil
The Long-Lasting Enforcement of al-Mutawakkils Restrictions
ix
1
8
10
21
32
57
58
60
63
68
85
88
89
92
96
99
100
103
108
Contents
5. The Provenance of the Modes of Subordination

of Non-Muslims
The Status of Minorities in Ancient Near Eastern Societies
The Status of Minorities in the Byzantine and
Sasanian Empires
Muslim and Pre-Muslim Modes of Subordination Compared
The Sasanian Origins of the Social Position of Non-Muslims
in Islamicate Society
The Ideology of the Iranian Class System
The Adoption of Sasanian Aristocratic Ethos by the Muslims
Conclusion
Conclusion
Appendix I: Al-Turtu4sh3s Version of Shuru4t Umar
Appendix II: Al-Sha4fi s Version of the Pact to Be
Accorded to Non-Muslim Subjects
Notes
Glossary
Bibliography
Index
113
114
116
121
130
137
141
162
164
171
173
177
235
239
255
Illustrations
1. Map of the locations of surrender treaties

and vassal treaties
2. The standing caliph from Khirbat al-Mafjar
3. Sha4pu4r II on his horse, hunting boars
4. Iranian gold belt-buckle
xi
page 46
134
151
156
Preface
As many do, this book started from an article. Actually, before that, it
started, as many new ideas do, in the classroom, in a course on nonMuslims under Muslim rule in the first centuries of Islam. In the beginning
it seemed as if when tackling the question of the status of ahl al-dhimma
I would be treading a path that had been trodden by many before me.
Using the fruits of former studies, I was therefore quite sure that I was
covering ground that was new for students, but not otherwise.
It was while reading and re-reading the sources and bibliography that
I discovered that there are still questions that are unasked and unanswered, and there are new avenues of research to follow. I discovered
that though there was ample work on the subject of Shuru4t Umar and
the status of the dhimms, it still did not provide an answer to one main
question:how and why had these Conditions of Umar come about? In
other words, what was their Sitz im Leben, what was their main purpose, and what were their sources of inspiration? I was especially interested in the question of intercultural exchanges that may have played a
part throughout the process of their formation. That is, what were the
cultural traditions that stood at the basis of this development? Were
they mostly Muslim, as had usually been presumed, or did traditions
and institutions of the conquered populations have a meaningful part
in this as well?
Having first tackled the question of the immediate circumstances in
which the document of the Shuru4t was formed (Chapter 2), I was drawn
to try and trace the process that took place between the time in which the
small Muslim minority had taken over huge territories with millions of
inhabitants which had formerly been under Byzantine and Sasanian rule,
xiii
xiv
Preface
and the moment in which a new comprehensive set of rules relating to

this new population had been established.
Upon entering new and untrodden ground, I found myself in need
of much information and advice. Many scholars generously shared their
erudition with me, and I therefore owe great thanks to many who assisted
me throughout this long journey.
I have a great debt to Amikam Elad, who has been guiding and supporting me in my work for many years, and has read the drafts of several
chapters in this book and commented upon them with meticulous care.
I benefited greatly from the comments and suggestions of many scholars who were generous enough to invest their time in reading various
versions of the manuscript. I owe much to Patricia Crone, who read the
complete manuscript from cover to cover and commented on it page after
page originally anonymously, but who has since consented to be identified. Her wise comments and suggestions have greatly improved the final
product.
Several people read the last chapter, which centres mostly upon
Sasanian history and culture and provided me with important insights
and comments. First of these is Zeev Rubin, who gave me much advice
and support while writing this chapter. It saddens me greatly that this
great scholar, who was always pushing the boundaries of knowledge,
relentlessly studying a myriad of cultures and languages, is no longer with
us. I also owe many thanks to Shaul Shaked and to Jamsheed Choksy,
who were kind enough to devote time to read and comment on this chapter. They all assisted me in establishing a stronger basis on issues related
to Iranian culture and society.
Mark Cohen was kind enough to read the whole manuscript, commented upon it, and shared his thoughts with me. I would also like to
thank Robert Hoyland, with whom I had enlightening conversations
on various issues during his stay in Jerusalem in the past year; he also
read several chapters and provided me with important insights. Julia
Rubanovich and Vera Moreen kindly assisted me in the translation of the
Persian. Yohanan Friedmann patiently assisted me with several matters.
Ruth Jacoby generously shared with me the cover illustration which she
and David Yerushalmi found while researching Jewish apparel in Persian
society. Finally, I would like to thank all of those who made the production of this book possible: the people from Cambridge University Press,
including Marigold Acland, Mary Starkey and Helen Wheeler, and Leigh
Chapman here in Jerusalem, who toiled over the index and the proofs. It
Preface
xv
was a real pleasure to work with them, and I thank them all. Needless to
say errors remain mine alone.
I also owe many thanks to the Center for Advanced Judaic Studies
of the University of Pennsylvania, where I was a member of a research
group in 20067, and to its director, David Rudermann, for giving me a
year free of cares. It was during this year that a major part of this book
was written.
My greatest debt is to my father, Moshe Weinfeld, a great scholar of
the Bible and the ancient Near East, who shared with me his great curiosity, his love of knowledge, and, most importantly, his endless pursuit of
historical and cultural ties and contacts, and provided me with a rich and
complex view of the ancient world. He passed away in April 2009, and
I am greatly sorrowed by the fact that he is not here with me to see this
book published.
I want to express my love and gratitude to Buni, my husband and partner, my constant companion, who was always there listening patiently to
my thoughts and my doubts, shared with me his insights and advice, and
supported me all the way to the finish line.
Introduction
The seventh century in the East was a time of major turbulence and
upheavals which culminated in the Muslim conquest and dominion over
what had been Sasanian Iran as well as over a significant part of the
Byzantine empire. When the Muslims gained their first major victories
over the Byzantine forces in the fourth decade of the seventh century
CE, and when in the next few decades they overtook large parts of the
ancient Near East, they were but a small minority among a large and heterogeneous population, made up of various ancient peoples, ethnicities,
cultures, and religions. The encounter between the new, and not yet fully
formed, Muslim religion and society and these ancient societies and cultures was a momentous event for both the conquerors and the conquered,
as is well attested by the contemporary sources that have survived.1 It
raised hopes of freedom and change in some of the conquered peoples,
including the Jews, the Samaritans, and the Monophysites, and generated
great fear and awe in others.
While the immediate effects of the conquest seem to have been mild,
according to the accepted views nowadays,2 and allowed life in the
conquered territories to take its course, its long-lasting effects were farreaching.
This book proposes to investigate the emergence of the regulation of
the status of non-Muslims under Muslim rule beginning with the initial
agreements signed at the time of the conquest and continuing into the
seventh to ninth centuries, a period in which the relationship between
the Muslim rulers and the numerous populations of conquered peoples
was formed. It was during this latter period that endeavours were made
to create a consistent policy regarding the conquered population, and the
1
document of Shuru4t Umar, which was to become canonic, was drawn

up. The book thus tries to track both the immediate initial stages and
the longue dure processes that stood behind the formation of the status
of non-Muslims under Muslim rule for many centuries to come. Unlike
former works on the subject, this research focuses on the origins and on
the process of the formation of the status rather than on the established
fixed status. Moreover, it attempts to do so from an interdisciplinary
viewpoint, integrating long-standing traditions, customs, and concepts
originating in the worlds of both the conquerors and the conquered.
history of research
The policy adopted by the conquering Muslims towards the huge and
heterogeneous population that they now dominated has attracted the
attention of many scholars of Islam. Research of this field has concentrated first and foremost on the examination of the canonical document
called Shu4ru4t Umar, i.e. the Conditions of Umar (also called the Pact
of Umar, or the Petition to Umar)3 which defines the relationship
between the Muslim conquerors and the non-Muslim population and
delineates the status of the latter in Muslim society. The Umar implied is
traditionally believed to be the mythological caliph and conqueror Umar
b. al-Khatta4b (r. 1323/63444).
This document has been discussed again and again in many different
scholarly works for a long time,4 and there has been a continuous debate
over its date and its Sitz im Leben. For several reasons,5 the most obvious
being that the document reflects a state of established Muslim rule and
of the close coexistence of Muslims and non-Muslims, the majority of
scholars who referred to this document believed that, in its present form,
it was not a product of the days of the conquest itself, as its title implies,
and as Muslim tradition claims. Other arguments include its unresolved
inconsistency with reports regarding the early surrender agreements, and
its irrational petition format, in which the conquered ask of their own
initiative that such a series of restrictions be imposed upon them. Rather,
it was justifiably assumed that the existing document was a product of a
later period, and was formulated by the Muslims some time during the
eighth or ninth century. It is thus considered by most scholars a pseudoepigraphic document which was attributed to the mythological caliph
and conqueror.6
It should be emphasized, however, that the probable date suggested by
most scholars, the eighthninth century, is nevertheless a comparatively
Introduction
early date as regards Muslim law. This makes this document (as well as
other parallel documents of its time) especially important when attempting to trace the formation of the status of non-Muslims.
Research regarding the status of non-Muslims as reflected in Shuru4t
Umar treated a variety of issues. The most renowned works, written
by Arthur S. Tritton and Antoine Fattal, provided an extensive survey
of the status of non-Muslims under Muslim rule in the document and
in Muslim legal literature, and reviewed its implementation through the
ages. Regarding the circumstances of its composition, Tritton, followed
by Fattal, believed that given the above considerations, the document
must have been drawn up as an exercise in the schools of law to draw up
pattern treaties.7 Alternatively, Albrecht Noth and Mark Cohen argued
that it was the product of an ongoing process which incorporated early
elements from the time of the conquest, especially those regarding the
security of the conquering minority, with new elements which reflected
the reality of later times.8
Cohen also raised the question of the form of the document, questioning the implausible idea that the non-Muslims actually came asking
their Muslim rulers for such conditions as are listed in the document;
he concluded that the Pact of Umar may be seen as an outgrowth of
the conquest treaties (Noths view) but transformed into the mold of a
petition.9 In this work Cohen also enriched the discussion by adducing
early and previously unknown versions of the Shuru4t.10 Noth raised several issues,11 one of which was the purpose of the Shuru4t. He asked specifically whether its purpose was, in fact, to humiliate the non-Muslims.
Based on an exhaustive analysis of the document itself, he concluded
that the documents intention was not in fact to humiliate, but rather
to differentiate between Muslims and non-Muslims. He reasoned that
the fact that the Muslim conquerors were but a small minority among
the conquered population caused a need for a means of differentiation
between the two groups. This view relied necessarily on his claim that
although the existing document was composed at a later time, it nevertheless reflects the conditions of the conquest rather than those of later
periods and circumstances.
Daniel E. Miller raised the question of the date of the canonization
of the document. In his Ph.D. thesis he followed the various versions
of the petition to Umar, sorted them out according to the various legal
traditions, and traced the development of the document from its nascent
stages to its canonization. Although Miller believes that the document
goes back in its embryonic stages to the second century of Islam, he
nevertheless claims that it became central only during the fourth century
of Islam, and believes that it became the normative document only in the
seventh.12
Until now research has thus focused on various aspects of the status of
non-Muslims under Muslim rule in its early stages, mainly on the canonic
text of the Shuru4t, and on Muslim had th and legal literature. In addition,
historiographic, religious, polemic, and other materials which originated
in the non-Muslim sources of Islamicate society were employed mostly in
order to examine the actual implementation of the restrictions.
An additional subject which attracted less attention was the initial surrender agreements.13 These were examined separately from the Shuru4t,
and their veracity has often been questioned. In addition, although some
scholars raised the question of the incompatibility of these agreements
with the Shuru4t, there has been no serious attempt until now to try and
trace the process of transition from the agreements to the canonical text
of the Shuru4t. Additionally, both the surrender agreements and the Shuru4t
have been examined mainly through Muslim sources or sources emanating from Islamicate society.
Goals of the Present Research
Unlike previous works, which focus mainly on the final product, i.e. the
Shuru4t, its implications and interpretations within Islamicate society, this
book endeavours to look for its roots and origins, searching for these not
only in the Arabian and Muslim world, but in the ancient cultures and
civilizations of the conquered lands and peoples as well. The key working
assumption is that the Muslims did not devise the principles that lay at
the basis of the surrender agreements and Shuru4t Umar ex nihilo. The
agreements, if indeed genuine, as I endeavour to demonstrate, must have
relied on some existing model. Similarly, Shuru4t Umar, which gradually
replaced the surrender agreements, and purporting to be a comprehensive surrender agreement, did not emerge deus ex machina; rather, it was
conceived in a long and complex process, and must have been inspired
by some former patterns and concepts that guided its creation. These
may have originated in the Arab society most familiar to the Muslim
conquerors, but may have also stemmed from the ancient societies of the
conquered peoples, including HellenisticRomanByzantine culture, and
Iranian society and culture. Verification or negation of this approach in
the case of the initial surrender agreements or the Shuru4t naturally entails
an examination of the sources representing these cultures which were
Introduction
dominant in the area prior to the arrival of the Muslims as well, and thus
requires research more interdisciplinary in its nature.
In this respect, this research joins a growing group of scholars who
assume that the development of Muslim society can be better understood
in its wider historical context, rather than as a world apart. In the last
decades attention has been drawn to the significant contribution of nonMuslim sources to the understanding of Muslim history in general, and
to the wide use of non-Muslim sources, including Greek, Syriac, Persian,
Jewish and other sources contemporary with the period of the conquest
in particular. This innovative course of research was led by Patricia
Crone and Michael Cook in their revolutionizing book Hagarism,14 and
was later followed by Michael Morony,15 Lawrence Conrad,16 Robert
Hoyland,17 Chase Robinson,18 and others.
The present book attempts to use not only contemporaneous evidence
originating in sources other than Muslim ones in order to understand
the transition period, but earlier sources as well. Hence, source material
relating to periods preceding the conquest is employed, in an endeavour
to delve into the history, traditions, and culture of the conquered societies in order to gain new insights regarding the concepts that shaped the
status of the non-Muslims in Islamicate society.
Using these varied sources originating in the various cultures that preceded the Muslim conquest, the book aims to draw a continuous fulllength picture of the process of the formation of the relationship between
the conqueror and the conquered from the first encounters and initial
surrender agreements particular to each city or region, through the preliminary endeavours to create a consistent policy regarding the conquered
population, the acceptance of Shuru4t Umar, and, at the end, the question
of its actual enforcement from the ninth to the eleventh centuries.
The Structure of the Book
The opening chapter examines the agreements made between the surrendering cities and their Muslim conquerors. As noted above, these have
been examined separately before, their authenticity was often doubted,
and their content was considered solely in view of the Muslim sources.
This chapter endeavours to prove that the agreements were in fact an
authentic product of the interaction between the conquerors and the conquered, and that they reflected, to a great extent, an ancient heritage of
the conquered societies regarding the customs, the procedures, and the
documents that were part and parcel of surrendering.
The process of the transition from multiple and inconsistent agreements to the creation of one general set of rules to be imposed upon all
non-Muslim populations is dealt with in the second chapter. I propose
here that this process involved an internal discussion within the Muslim
world over the continuing validity of the surrender agreements, and over
various proposed alternatives of such a uniform document. While it is
shown that the Muslims accepted the traditionally sacrosanct character
of these documents, it is argued that the need for a uniform and accepted
policy regarding the non-Muslims living under Muslim rule became
urgent, and overcame the inhibitions and reservations. The discussion
regarding the various alternatives ended, in its turn, in the complete victory of Shuru4t Umar over its competitors.
The third chapter tackles the question of Umar IIs role in the process. It attempts to reinforce a long-standing thesis, already raised by
Tritton and Fattal, that the basis of the Shuru4t was laid by Umar II b.
Abd al-Az3z (r. 99101/71720) by showing that that the principle of the
ghiya4r, i.e. the differentiating signs between Muslims and non-Muslims
via dress, appearance, and public behaviour, which forms the main part
of the Shuru4t, was part and parcel of the ideology of the exaltation of
Islam which was widely promoted by Umar b. Abd al-Az3z.
The fourth chapter of the book contributes to the long-standing discussion regarding the actual enforcement of Shuru4t Umar. I attempt to
show here that in contrast to the generally accepted notion that until the
twelfth century the Shuru4t were enforced only sporadically, Muslim rulers from the ninth century onwards in Egypt and Syria often attempted
to enforce the Shuru4t with varying degrees of success throughout the
caliphate.
Although last, the fifth chapter is central to the thesis of the book. This
chapter, entitled The provenance of the modes of subordination of nonMuslims, aims to trace the origins of the various clauses that make up
the Shuru4t. A large part of the chapter is dedicated to the origins of the
ghiya4r, in attempt to understand what ends they were meant to achieve.
The main thesis in this chapter is that most of the restrictions originated
in rules and customs that were prevalent in Byzantine and Sasanian
societies. There is, however, a significant difference between these two
sources:the restrictions originating in the Byzantine empire revert mainly
to Byzantine law regarding Jews in the empire, a clear and straightforward
transfer of a code regarding members of a dominated religion. However,
the rules originating in the Sasanian realm, mainly those regarding ghiya4r,
revert to the ideal of the Sasanian class system, which was promoted in
Introduction
Iran by the Sasanian aristocracy, and demanded that an external set of

signs including dress, paraphernalia, and public customs distinguish
between the elites and the commoners. This Sasanian ideal of an immobile hierarchic society, where each estate is clearly discernible through its
dress and paraphernalia, was adopted and in fact appropriated by the
Muslims in order to distinguish between Muslims and non-Muslims, as a
way of establishing their own superiority.
I cannot end this introduction without making the following statement:my work is purely academic; at no stage of this research was there
any intention that it should serve any political ends.19 I am aware of
course that its results, especially those in Chapter 4 and Chapter 5, may
be used by some to support claims of an inherent policy of humiliation
towards non-Muslims in Islam a claim that should immediately be
rejected. My opinion is that one cannot compare the ancient and medieval climate regarding social status and hierarchy with the views of the
modern world. In ancient and medieval societies social hierarchy, as well
as discrimination between various groups, was accepted, and was almost
self-evident. (This is true, by the way, even for democratic Athens, where
women, metoikoi, and slaves, who together made up the major part of
Athenian society, did not have the same rights as the Athenian citizens
and were socially inferior.) None of these societies believed in equality
or equal rights the way modern Western societies do. I therefore believe
that judgement of these societies according to our values is anachronistic
and useless. There is no sense in trying to attach these views and concepts
blindly to contemporaneous Islam, which just like any other religious
or political group is made up of diverse views and notions. In sum, in
my opinion historians should remain loyal to their sources and to their
academic disciplines, and the results of their research should remain as
detached as they could possibly be from any current political or social
debates and controversies.
1
The Roots and Authenticity of the Surrender
Agreements in the Seventh Century
The early surrender agreements made between the Muslim conquerors

and the non-Muslim inhabitants of the conquered cities are a common
feature throughout early Muslim historiographic and legal literature.
Surrender agreements are often mentioned, and at times cited in full,
by al-Bala4dhur3, al-Tabar3, al-Yaqu4b3, Ibn Atham al-Ku4f3, Ibn Abd
al-Hakam, Abu4 Yu4suf, Abu4 Ubayd, and Yahya4 b. A"dam; they are also
mentioned sporadically in many other compositions.
These agreements have been studied by various important scholars,
from numerous points of view.1 The main claim against the authenticity
of the surrender agreements lies in the fact that some of the agreements
cited by Muslim authors are detailed and comparatively long documents.
They include not only general conditions concerning payment or taxation
on the side of the conquered and the obligation of protection of people,
property, and prayer-houses, but in fact many intricate details regarding
arrangements concerning public matters as well as peoples rights and
property. The agreements are written, witnessed, and signed usually by
the commander of the Muslim army and at times are reported to have
been sealed. They seem to be too complex and versatile for conquerors
who had recently emerged from the desert and were not yet sure of their
position in regard to the conquered population. Thus Fattal believes that
the early agreements were rather succinct and undetailed. According to
this view, some of these agreements were made verbally, and it is doubtful whether many of them were in fact written down at the time of the
conquest.2
The existence of the famous treaty between Nubia and Amr b. al-A"s
successor, Abdalla4h b. Ab3 Sad b. Ab3 Sarh, concluded in 652 and known
8
The Roots and Authenticity of the Surrender Agreements
as the baqt,3 has been doubted by several scholars. Peter M. Holt claims
that the treaty, almost certainly legendary, represents an attempt to retroject conventions of MuslimNubian relations which had developed
by the 4/10th century.4 Michael Brett too claims that it is a product
of Muslim jurisprudence, and Jay Spaulding presented it not long ago
as a forgery meant to promote Muslim interests.5 However, in 1972 in
the excavations at Qasr Ibr3m in Egyptian Nubia, a papyrus scroll from
the eighth century was discovered, which contained a letter from the
Abba4sid governor in Egypt written in 141/758 to the king of Nubia
and Muqurra, demanding from him what you owe of the baqt about
which a peace agreement was made with you,6 and demanding that the
Nubians fulfil their side of the pact if you wish us to fulfil for you our
compact (ahd).7 Although this letter is not the original conquest agreement, it indicates that the governor in the mid-eighth century believed
that there was a valid surrender agreement, similar in its contents to the
agreement found in the literary sources. The authenticity of the Nubian
baqt is thus well attested by this letter, and as a result provides some
support for the authenticity of surrender agreements that are referred to
or cited by Muslim authors; yet it provides only a single, somewhat late,
piece of evidence.
Two scholars, Albrecht Noth and Wada4d al-Qa4d3, dedicated articles
specifically to the question of the authenticity of these agreements.8 Noth
believes that the reports concerning the agreements reflect authentic
documents, though these have been altered at times by the transmitters.
He nevertheless assumes that some details, such as the obligation not to
revile or hit Muslims, or the obligation to build roads or bridges, were
added on later.9 He adds that since we do not possess any copy of an
original contract, we do not have any secure means of verifying whether
these agreements are authentic, forged, or just a fiction. The question of
the genuineness of these agreements, he notes, can only be examined with
the help of inner criteria, which unfortunately can only be of approximate value.10
Al-Qa4d3 has attempted to further support the authenticity of the surrender agreements.11 She bases her argument on the texts of the agreements
themselves, and following their comparison arrives at the conclusion that
they were drawn up in similar ways, that they are grounded on similar
formats including basically the same elements, that their style is standardized, and that their content is analogous. In addition, she also mentions
that there exist formal Muslim documents of a different nature from
the end of the first century of Islam that include similar phrasing, thus
10
supporting the idea that legal documents exhibiting the same style were
already being written at the time.12
Al-Qa4d3s work does indeed go far in advocating the case for the
authenticity of the agreements. In this chapter I would like to add further support to this argument with the aid of evidence found outside the
corpus of Muslim literature. The evidence I will present here originates
in the realm of the conquered rather than that of the conquerors. This
external evidence is then supported and enhanced by the Muslim sources
themselves. In this chapter it will be argued that the surrender agreements made between the Muslim conquerors and the representatives of
various conquered entities (cities, regions, or groups) have their origin in
an ancient tradition of international diplomacy and law which, mutatis
mutandis, was still prevalent throughout the territories when conquered
by the Muslims. This tradition was not only a norm accepted by the
conquered population at the time of the conquest, but was known to the
Muslim conquerors as well. If this is true, then not only is there no need
to suspect the authenticity of these agreements, there is in fact good reason to acknowledge their validity.
treaties before the muslim conquest

International treaties formed the main basis for international relations
from ancient times throughout the ancient Near East, as well as throughout the Hellenistic, Roman, and Byzantine world.13 I will give here just
a succinct summary of this institution which highlights the predominant
characteristics of this tradition in the early periods. This will be followed
by a more detailed description of such treaties and agreements in the
Byzantine period, prior to the Muslim conquest.
In the Graeco-Roman world international pacts and agreements were
considered part of the ius gentium (the law applicable to all people). The
treaty itself was under the sacred protection of the deity invoked by the
oath. Zeus/Jupiter was called Zeus horkios kai pistios, i.e. the the guardian of oaths and good faith (pisitis/fides). This was the actual basis of
obligation in all agreements under international law. The requirement
that treaties should be upheld (pacta sunt servanda) became a categorical
imperative of international law.14
This tradition of treaty-making was characterized by various common
elements:the pacts or agreements were concluded following preliminary
negotiations; they required ratification of the sovereign body; they were
11
ratified by solemn oaths; they had a similar structure, often including preamble, stipulations, sanctions, provisions for deposit in the temple and
periodical reading, and names of the gods acting as witnesses. It is often
emphasized that they were written down. Copies of the treaty, written
in many cases in two languages, were kept by both parties, usually in
the temple, or published and placed in the archives. A pact or agreement
was valid either for a certain period of time, or throughout the life of the
ruler who signed it. Upon the death of one of the parties it became void,
and needed to be renewed. It should be remarked that stipulations often
included such elements as an obligation of loyalty, military aid, return of
fugitives, giving hostages as guarantee of observance of the treaty by the
vassal, and payment of an annual or lump sum of money. At times these
treaties could be very detailed, and included specific information on various matters such as arrangements concerning the evacuation of people
and territories, the army, hostages, boundaries, provisions, equipment,
payments etc.15
There was a common terminology used in conjunction with the treaties.16 Thus, the covenant was called bond and oath or pact and promise (e.g. Hebrew brit we-ala; Greek horkos kai synthe ke ; Arabic aqd
wa-hilf;17 Persian pasht ud ze nha4r18) or called honesty or confidence
(Akkadian ade ; Aramaic dy; Hebrew edut or amana; Greek pistis;
Latin fides); a covenant was cut, i.e. made (e.g. Hebrew karat brit; Greek
horkia temnein); violated broken (Hebrew hepher; Arabic naqada;
Greek parabainein; Latin frangere); the relationship was described as
love and friendship (e.g. Hebrew twb we-hesed; Greek filia kai symmachia; Latin amicitia et societas) and more.19 The existence of such a
universal vocabulary demonstrates that treaty-making was based on a
well-established set of rules, perceptions, and terms which formed the
basis of understanding and accord between the negotiating parties and
were commonly known and accepted throughout the ancient world.
The following survey will demonstrate that this tradition endured well
into the end of Byzantine rule in the East, and that agreements and treaties, often including a long set of stipulations, securities, sanctions, and
oaths, continued to be signed and witnessed. These agreements served as
a central diplomatic tool throughout Late Antiquity in the relations of the
empire with the competing political entity the Sasanian empire as well
as with groups of lesser political status such as the Barbarians and the
Arabs. As shall presently be shown it is this diplomatic tradition, in fact,
that was to play a major role in the Muslim conquest of the Near East.
12

Treaties with the Sasanians
Throughout the centuries of Late Antiquity the Roman empire made a

series of treaties with the Sasanian empire, including those of 244, 298,
363, 422, 505/6, 533, 562, and the last one, in 628, just a few years prior
to the Arab conquest.20
As a matter of course these treaties were preceded by an offer of peace
by one of the parties, followed by intricate diplomatic negotiations held
by embassies which, on different occasions, were made up of envoys of
various ranks:from those bearing plenipotentiary powers to those who
served solely as messengers. During these negotiations the conditions
of the future treaty were discussed, including boundaries and fortifications, various payments, property, diverse rights (e.g. rights of passage,
or trading rights), withdrawal of soldiers, the fate of prisoners of war,
hostages, etc.21 Often these negotiations were conducted in writing, via
an exchange of letters. Letters from the rulers on both sides also ratified
these agreements.22
To note just the outstanding examples, the making of the treaty of
244 was recorded in the famous Ka4ba-yi-Zardusht inscription at Naqshi-Rustam as well as in various Roman sources.23 In one of the several
rock-reliefs depicting the kneeling emperor Philip the Arab offering
peace to the mounted victorious Sha4pu4r I, Sha4pu4r is seen holding an
object tied by a ribbon most likely the agreement or its draft.24 In
298, following Galerius glorious victory over Narseh and the capture
of Narsehs family and treasures, a detailed treaty was negotiated, with
the Romans having the upper hand.25 The main source in this case is
Peter the Patrician (c. 50064), Justinians master of offices (magister
officiorum), who was himself sent as an envoy to Khusro Anu4sh3rwa4n26
and most probably had access to archival material. Peter adduces a full
description of the negotiations, but does not actually supply the text or
the exact terms of the treaty.27
The famous treaty signed in 363 by Jovian after his great defeat by
Sha4pu4r, in which he ceded large areas in Mesopotamia to the Persians,
is described in detail by Ammianus Marcellinus.28 He explains how the
Persians sent envoys to facilitate the opening of negotiations over a peace
treaty (pax; foedus amicitiae), describes the negotiations in detail, and
lists the terms of this treaty, adding that:
When this shameful treaty (decretum)29 was concluded, lest anything
contrary to the agreements (pactum) should be done during the truce,
distinguished men were given by both sides as hostages:from our side
13
Nemota, Victor, and Bellovaedius, tribunes of famous corps, and from

the opposite party Bineses, one of the distinguished magnates, and three
satraps besides of no obscure name. And so a peace of thirty years was
made (foederata itaque pace annorum triginta) and consecrated by the
sanctity of oaths (eaque iuris iurandi religionibus consecrata).30
Zosimus adduces the terms of this treaty as well, noting that the truce
(spondai) was concluded on these terms and confirmed by contracts
(grammatioi) on both sides.31
This shameful treaty now had to be upheld, to the Byzantines great
dismay. Especially humiliating was the clause regarding the ceding of
Nisibis and Singara and the evacuation of its population to Byzantinecontrolled territory. Despite the repeated supplications of the citizens of
Nisibis to be allowed to fend for themselves rather than evacuate their
city, Jovian refused to break the treaty, claiming that he did not wish
to incur the guilt of perjury,32 stoutly maintaining the sanctity of his
oath.33 This insistence demonstrates how binding these agreements were
considered to be.
Ammianus could not disregard this claim, but insistently argues that
in the past, in similar extreme cases, such treaties were annulled:In fact,
the ancient records teach us that treaties made in extreme necessity with
shameful conditions (icta cum dedecore foedera), even when both parties
had taken oaths in set terms (postquam partes verbis iuravere conceptis),
were at once annulled by renewal of war.34 He then proceeds to adduce
the relevant cases. This not only demonstrates that the institution of the
treaty (in the case of the Persians the equal treaty) was still functioning
according to the same principles, but that it was indeed conceived by the
Byzantines themselves as the same institution, abiding by the same rules,
and subject to precedents of Roman legal history.35
Although the information regarding the technical side of the treaties of
the fifth century is less detailed, there is enough to show that there was no
basic change in the procedure and structure of these treaties.36
In 505/6, after several years of war, a temporary truce was signed.37
According to the chronicle of Marcellinus Comes, who served as cancellarius in Constantinople under Justinian, Celer, the magister officiorum,
resolved to conclude a treaty (foedus) with the Persians, when Armonius,
secretary a secretis, had been sent to him to draft the treaty.38 Procopius
says:So a proposal was discussed between them, according to which the
Persians were to deliver over the city [i.e. Amida] to the Romans upon
receipt of one thousand pounds of gold. Both parties then gladly excecuted the terms of the agreement (ta sunkeimena).39 Joshua the Stylite,
14
who gives a minute eye-witness account of these events, reports that once
both the Persian and the Roman commanders saw that they were losing strength, they agreed to make peace on condition that the deal was
approved and ratified [by both] rulers; if not, the war would continue.40
This condition cited by Joshua the Stylite relies once again on the Roman
tradition mentioned above by Ammianus Marcellinus, according to
which though the generals and envoys on site had the authority to sign
the agreements, and though these signatures were considered valid, the
contracts had nevertheless to be ratified by the rulers. It may be conjectured that in this specific case this condition may have carried more
weight, and the agreement was actually not in force before its ratification
by the rulers.41
The chronicle attributed to Zachariah of Mitylene adduces a somewhat different, yet exceedingly interesting, version regarding this treaty.
According to him, Farzman, a warrior who had distinguished himself in
fighting on the Roman side, came to the city and reached an agreement
(tanway) with the Persians there, after which:
Celer the master of Offices, gave to Kava4dh eleven hundred pounds
of gold for the ransom of the city and for peace. And when the documents were drawn up they brought the drafts [Syriac form of the Greek
apographas] to the king for his signature. And the king fell asleep, and
it was told him in a vision that he should not make peace (shayno); and
when he woke up he tore up the paper, and departed to his own country,
taking the gold with him. But Farzman remained in the city to govern its
inhabitants and the country.42
Although this story may be suspected to be a popular myth, it is nevertheless significant to our discussion since it emphasizes both the authority of
the signed documents and the importance of the ratification by the ruler,
without which an agreement could not be proven valid.43
A significant example of this formal procedure is preserved in the
account of Menander the Guardsman (Protector) describing the conclusion of the fifty-year peace in 562 between the Byzantines and the
Persians. Writing in the second half of the sixth century,44 Menander
recounts in detail in his History the process of the negotiations, the stipulations of the treaty itself, and the technicalities it involved. Following
the negotiations, and before the terms of the treaty were put down in
writing, it was agreed that both rulers should provide the documents
which are called sacrae litterae in Latin, and which confirmed everything
that had been established by the ambassadors. He notes that following
15
a full description of the terms these sacrae litterae were exchanged. He

then provides a rare description of the process of the writing down of the
treaty and its signing:
When these and other issues had been argued out, the fifty year treaty
was written out in Persian and Greek, and the Greek copy was translated into Persian, and the Persian into Greek. For the Romans the
documents were validated by Peter the master of the offices, Eusebius
and others, for the Persians by the Zikh Yesdegusnaph, the Surenas and
others. When the agreements had been written on both sides, they were
placed side-by-side to ensure that the language corresponded.45
Menander then goes on to list in detail the provisions of the treaty. He

continues:
When matters had progressed to this stage of orderly development,
those whose task it was took the texts of the two documents and polished their contents, using language of equivalent force. Then they made
facsimiles of both. The originals were rolled up and secured by seals
both of wax and of other substance used by the Persians, and were
impressed by the signets of the envoys and of twelve interpreters, six
Roman and six Persian. Then the two sides exchanged the treaty documents, the Zikh handing one in Persian to Peter, and Peter the one in
Greek to theZikh. Then the Zikh was given an unsealed Persian translation of the Greek original to be kept as a reference for him, and Peter
likewise was given a Greek translation of the Persian.46
From this outstanding description it may be gathered that the formalities

of treaty-making demanded skill and expertise, and that they followed
an accepted conventional procedure of writing and translating, inspection of the translation by both sides, followed by the formal signing,
which involved an additional editing of the treaty, the making of copies,
the sealing of the original documents and the impression of the seals by
envoys and interpreters of both parties, and the exchange of the originals
and of the copies to be used for reference. As always in the case of Rome
and Persia, the treaty was one made between equals (foedus aequum), as
is quite clear from Menanders account.
Although there is scant information concerning the treaty between
Kava4dh II and Heraclius in 628, there is enough of it to indicate that a
similar treaty designated to restore the status quo ante was negotiated
in a similar manner. The shah sent a letter requesting peace, the emperor
replied and entrusted the tabularius (notary, registrar) Eusthantios with
the negotiations, and the treaty was signed.47
16

Treaties with the Barbarians and the Arabs
These conventions were not limited to these two ancient empires, where
they were long established,48 but were adopted and employed by the
Romans during the period of Late Antiquity in regard to two entities that
entered into formal relationships with the empire:the Barbarians and the
Arabs. In both cases there is evidence to show that they became foederati
of the empire not only in name, but according to the same rules, practices,
norms, and terminology to which the Roman empire had been formerly
accustomed.
Barbarians
Before the case of the Barbarians is demonstrated, it should be noted that
at times they introduced into the process their own traditional mores and
diplomatic practices.49 Thus, they seem to have had their own ways of
surrendering, including not only the familiar supplication on bent knees
and prostration, but also such gestures as throwing aside their weapons
and falling flat on their breasts,50 presenting themselves in the manner
of criminals standing with bended bodies, and begging that swords be
poised at their throats, an act which no doubt symbolized their fate if
they were to break their oath.51 When swearing, they drew their swords,
which they venerate as Gods, and swore that they would remain loyal.52
Ammianus notes in fact regarding the Sarmatians that never before had
they been forced to present pledges for a treaty.53 When the Roman
ambassadors met the envoys of the Huns they held a meeting mounted
on horseback. For the barbarians do not think it proper to confer dismounted, so that the Romans, mindful of their own dignity, chose to meet
the Scythians in the same fashion, lest one side speak from horseback, the
other on foot.54
In spite of all this, these tribes seem to have adapted to Roman diplomatic norms, and the sources indicate that they conducted negotiations,
and signed peace agreements, according to Roman tradition. Most of these
treaties are in fact unequal treaties (foedera) which followed capitulation
to the Romans.55 Thus56 in 360 Valens sent envoys to the Goths offering
them conditions; the envoys sent letters reporting that the Goths agreed
to the conditions, a meeting place was agreed upon, a treaty was struck,
and hostages were exchanged.57 Elsewhere Ammianus reports how in 374,
following a meeting between the emperor Valentian and the king of the
Alamanni, a treaty (foedus) of friendship was confirmed between them by
17
the sanctity of oaths (amicitia media sacramenti fide firmatur). After that
the king of the Goths was referred to as socius of the Roman people the
traditional terminus technicus for a Roman ally.58 In 378, when Fritigern,
leader of the Goths, negotiated with Valentian, he wished to become in
fact rex socius et amicus, the traditional title of a fully recognized allied
king.59 The treaty was finally signed in 382 with Theodosius, following
lengthy negotiations. Although no full description of it survives, the key
terms may be reconstructed:the Goths became foederati or symmachoi;60
they were allowed to settle in certain Roman territories yet to remain
autonomous. They were not, however, given Roman citizenship or conubium. They were obligated to supply auxiliaries for the Roman army, and
in return were to receive gifts and regular payments.61 The fact that such
complex treaties were negotiated indicates that the terms and the gestures
of treaty-making were not just shadowsof a past glorious institution; on
the contrary, it seems that the institution of the treaty played an important
role in contending with the challenge of the invading tribes. In fact, it may
have even received a special impetus in this situation.
The treaty (spondai) made with the Huns when Attila came to power in
434 was sworn to each by his native oath (patrion horkon omosantes).62
It was a complex and detailed treaty, and included the obligation not to
receive fugitives from Scythia, to hand back those who were already in
Roman territory along with Roman prisoners of war who had escaped, or
alternatively pay eight solidi for each; in addition, it stated that the Romans
should make no alliance (me symmachein) with other tribes against the
Huns and that there would be safe markets with equal rights for Romans
and Huns. This treaty was to be maintained as long as the Romans paid a
yearly sum of 700pounds of gold to the Scythian kings. When the Huns
broke the terms of the agreement and attacked the Romans at the market,
the Romans blamed them for contempt of the treaty (oligo4ria to4n spondai).63 Later on Attila sent letters to the emperor Theodosius II claiming the breach of certain clauses, after which he attacked the Romans.
This was followed by the signing of a new treaty, whose terms were still
more advantageous to the Huns. The point here is that the Huns adopted
the Roman system, and turned it quite evidently to their own advantage. They indeed signed written treaties with the Romans, consisting of
many clauses, most of which were clearly to their own benefit and profit.
The Roman foedus, a sophisticated legal institution, was thus adopted
and employed by the invading tribes here to overcome the Romans
themselves.
18
Arabs
Alliances (hilf) and covenants (ahd, aqd) among the Arabs were
recorded in pre-Islamic times.64 There is evidence of alliances of communities through a covenant in southern Arabia in the beginning of the
first millennium BCE.65 Written compacts kepts by the parties are attested
in the sixth and seventh centuries (kita4b, sah fa).66 However, their terminology and contents vary to a great degree from those common in the
Graeco-Roman world.67
Although sources regarding treaties with the Arabs outside the Arabian
peninsula are less abundant than those with the Goths, there is enough
to show that they were just as familiar with this institution as an international political tool as the tribes living on the northern borders of the
empire.
Zeev Rubin has already noted that the bilingual GreekNabatean
inscription found in Rawwafa
4 in Saudi Arabia, written between 166 and
4
169,68 commemorating the dedication of a temple built by the Thamudians
to the Roman caesars Marcus Aurelius and Lucius Verus, clearly points to
the existence of an alliance between these people and the Roman empire.
Another alliance with the Romans is evidenced in Imru4 al-Qays
famous funerary inscription in Nama4rah from 328. Although, as Rubin
notes, this inscription is full of lexicographic, palaeographic, and grammatical problems, the only thing that is quite clear in it is that Imru4 alQays was an ally of the Romans.69 Although we have no specific details
regarding these alliances, it can be quite safely assumed, based on the
cases already examined as well as those that will be presently presented,
that an alliance necessarily entailed a written document.
There is ample information about the alliance made between Mavia,
queen of the Saracens, and the Romans at the end of Valens rule in
378 in the Byzantine sources.70 It was actually a renewal of the Roman
treaty made with the Romans by her deceased husband, who was king
of the Saracens. Sozomen notes that at that time, the Saracen king
having died, the alliance with the Romans was dissolved (eluthesan).
Mavia, his wife, received the authority over the people and devastated
the cities of Phnicia and Palestine, as far as the place inhabited by
those Egyptians well known to those sailing up the Nile, the inhabitants of the region called Arabia.71 The term used by Sozomen here
regarding the treaty, eluthesan, is more rightly translated as dissolved
since such treaties were always between rulers, and had to be renewed
when either of the contracting parties passed away. As noted by R. C.
Blockley, at that point the successor of a deceased could declare that
19
the treaty was ended and act accordingly or he could seek alterations, or simply renew the treaty, if the other party agreed.72 This is well
in line with Sozomens description:the Saracen king dies, the treaty is
dissolved, and Mavia does as she wishes until new terms for an alliance
are negotiated.73
Rubin demonstrates convincingly that alliances with the Saracen tribes
were accepted and continuous. It seems that, following this renewal of
the alliance in 378, Mavias people greatly aided the Romans in their battles against the Goths,74 while later, in 389, the orator Pacatus noted that
a punishment was exacted from the rebellious Saracens for the dishonouring of a treaty.75 Rubin notes that Mavias daughter continued this
tradition, and that the alliance with the Romans was still intact in 425.76
He continues to demonstrate how Mavias rule was replaced by that of
Dujum (called Zokomos by Sozomen77), who headed the Arab foederati
in the following period, and argues convincingly that Zokomos, who had
converted his tribe to Christianity according to Sozomen, was in fact
the ancestor of the Daja4im who according to Hamza al-Isfaha4n3 lived in
Syria and Transjordan, and who converted to Christianity, became allies
of the Romans, and ruled other Arab tribes. Rubin thus claims the existence of a chain of alliances starting with Mavia, and continuing with the
clan of Dujam who in their turn were to be replaced by the Ghassa4nids.
Another case of an alliance with the Saracens is recorded by the historian Malchus. He recounts how a certain Amorkesos, head of the tribe
of Nomalius (?), fled in 474 from Persian territory to the Romans, and
managed to seize Iotaba (an island off the shore of present-day Eilath),
and to become an ally of the Romans and phylarch of the Saracens under
Roman rule on the borders of Arabia Petraea. His tribe too, according to
Malchus, was Christian.78
Yet another case is that described by Nonnosus, a member of a line of
diplomats in the service of the Byzantines who seem to have specialized
in negotiations with the people of Arabia. Excerpts of Nonnosus description are preserved in Photius Bibliotheca.79 Nonnosus grandfather had
been sent by Anastasius to sign a peace treaty (eire ne ) with Arethas, head
of the tribe of Kinda, while Nonnosus, his grandson, negotiated a treaty
(kai eire nikas etheto spondas) with Qays, head of Kinda, at the time of
Justinian. Although the terms of this treaty are not described in detail,
Nonnosus recounts that as part of the treaty, Mua4wiya, son of Qays, was
taken as a hostage to Byzantium, while Qays himself was called to the
emperor, and was appointed by him over the Palaestinas (te n Palaistino4n
hegemonian para basileo4s edeksato), bringing with him a large number of
20
his people.80 It is not clear exactly what office or title Qays was honoured
with; however, the terms which included a royal hostage, an invitation
to the emperors court, and an honourable title and office point to a
very favourable agreement for Qays.81 It is quite evident that this was a
detailed treaty with specific terms that was in all probability recorded in
a document.
It seems therefore, that the Roman (Byzantine) empire employed the
traditional formal alliances in her relationships with the Arab tribes on its
borders. Formal expressions of these alliances, such as skilled diplomatic
envoys, dissolution of treaties following the death of one of the parties
and the need for their renewal, blame for breach of a treaty, and the taking of hostages as security for the upholding of a treaty, all point to the
existence of treaties signed according to the accepted Roman tradition.
As we have seen that this was the case with the Goths and the Huns as
well, this is not surprising at all. In the case of the Arabs, however, this is
especially significant, since it indicates that they were quite familiar with
the minutiae of formal alliances and treaties centuries before the Muslim
conquest.
The point made here is that although conditions were no doubt
much changed, and the superiority of the Roman empire was no longer
unchallenged, it nevertheless continued to employ the same diplomatic
practices.82 In fact, signing treaties was the best way in which the later
Roman empire could continue to effectively control the growing number
of ties and commitments that were being made by the numerous parties.
The Roman tradition of making long and detailed treaties is therefore
evident in the relationship of the empire with the Persians, Goths, Huns,
and Arabs. Of course, changes had occurred through time. Already in the
Byzantine period, treaties and agreements were no longer being inscribed
on stone, or bronze tablets (see above), materials which enabled their
preservation for posterity; they were now being written down on parchment or papyri making them, alas, much more fragile and perishable.
This explains the fact that, from the Byzantine period onwards, we only
have knowledge of agreements rather than actual copies. The witnesses,
who in Antiquity were often the gods themselves, could be replaced by
human witnesses, as exemplified by Menander the Protector.83 The parchment or papyri documents were naturally signed and sealed and no longer placed under the statues of the gods at the temple, and the ancient
imprecations are absent from the treaties.
Nevertheless, the issues raised in the treaties and agreements signed in
the Byzantine period including the prohibition on alliances with other
21
political entities; fugitives; prisoners of war; hostages serving as security;

supplying military assistance; and payment or other remuneration were
central in past agreements of the empire, and continued to play an important part in future agreements to be signed during the Muslim conquest
of the East, as will be demonstrated later on.84
local surrender agreements during the roman

and byzantine period
Since most of the agreements during the Muslim conquest were made
between the Muslim commanders and the cities they conquered, I would
like to treat here the specific case of agreements made on the occasion of
the surrender of a city or territory to its enemies. It seems necessary to
refer here to two major phenomena which have a bearing on the nature
of the Muslim surrender agreements. These are:(1) the position of the
cities in the East throughout the hostilities between the Romans and
the Sasanians in the centuries preceding the Muslim conquest; and (2)
the history of surrender agreements made by cities prior to the Muslim
conquest.
The Position of the Cities in the East during
the Hostilities between the Romans and the Sasanians
The cities of the East, and especially those of Mesopotamia, had lived
through a history of wars and instability in which they were the main
victims, being raided, looted, and transferred from one side to the other
from the beginning of Roman rule in the East. A short survey starting
from the second century onwards will serve to emphasize the fact that
by the seventh century this situation was considered almost natural by
the inhabitants, who probably could not imagine any other way of life.
Moreover, since the ruling empire could not be depended upon often
enough to supply them with adequate fighting forces, a citys inhabitants
were quite accustomed to taking responsibility for its fate. This meant
that often they had to choose whether they wished to fight the advancing
forces or to negotiate the citys surrender. All of this was often enough
done independently, without consulting the central government or ruler
at the time, and was therefore self-understood for many of the cities that
were confronted by the approaching Muslim forces.
In the days of the emperor Trajan (r. 98117) the Persians captured
many cities and plundered many districts.85 According to John Malalas
22
(mid-sixth century) the Persians captured Antioch through an agreement

with its inhabitants. Later Trajan managed to convince the people of
Antioch though a secret communiqu to attack the Persians, promising
them the support of the Romans.86
In 256 and in 260 Antioch was captured again by the Sasanians. The
emperor Valerians attempt to drive them back ended in a humiliating
defeat. During Justianians rule, in 538/9, Antioch was captured once
again by the Sasanians.87
Around 350 the Persians attacked the cities along the Roman frontier, including Nisibis and Amida. As related above, the defeat and death
in battle of the emperor Julian in 363 were followed by the surrender of Nisibis and Singara to the Persians, cities which had previously
exchanged hands several times. On this occasion the newly appointed
emperor, Jovian, concluded a shameful treaty88 with the Persians, ceding
Nisibis to them. This was supposed to be mitigated by Sha4pu4rs agreement to allow the inhabitants of the city to leave the city for Roman
territory. The people of Nisibis and the surrounding area begged Jovian
to repeal his decision, but he refused; the inhabitants then begged for the
right to fight the Persians themselves, but were refused on the grounds
that the oath and the treaty could not be broken. After this rejection,
Jovian marched through the mourning cities, who could not bring themselves to offer any cheer of pleasure, contrary to their usual practice. The
inhabitants of Nisibis were then transferred to Amida.89
Although there was comparative peace from the end of the fourth
century and throughout the fifth, by the beginning of the sixth century
trouble started brewing up again. In 502/3 the cities of Mesopotamia
Harra4n, Amida, Tella, and Edessa suffered from harsh Persian attacks.
Thus, when Amida was captured by the Persians a great number of the
inhabitants were killed (Pseudo-Joshua and Pseudo-Zachariah mention
80,000!),90 the city and its churches were looted and destroyed, and all
the survivors apart from the old and disabled were taken into captivity
and resettled in Singara.91 During this attack it is first and foremost the
people of Amida, as well as those of Tella and Harra4n, who played a
crucial part in the fighting over their cities and the surrounding territory,
while the Roman army was not always to be depended upon.92 In fact,
the local militias in this area, supported by provincial vexillations, were
together responsible for the maintenance of their pedaturae, or sections
of the wall.93 It is also the inhabitants of these cities who obviously paid
the terrible price of defeat.94 This aroused great tension between them
and the emperor. Thus, it is claimed in the chronicle of Pseudo-Joshua
23
that while there was terrible hunger among the remaining inhabitants of
the conquered cities, the Roman troops lacked nothing. There were more
things for sale in their camps than could be found in the cities, whether
food, drink, shoes or clothing.95 Elsewhere he says that those who came
to our assistance ostensibly as our saviours looted us in a manner little
short of enemies.96
During the Persian Wars in the days of Justinian, many a city was
attacked by the Persians:attacks on the cities of Mesopotamia and Syria
were recurrent from 530 onwards; among these are Gabboula, Batnae,
Martyropolis, Sura, Beroea (Aleppo), and finally, in June 540, Antioch.
When the Persians arrived at Antioch the Roman army retreated, and
left the inhabitants to fend for themselves. They fought heroically to the
end, and paid a dear price once defeated.97 The Persians raided the city,
and looted its cathedral and churches. The war was renewed in 572 for
another twenty years, and Dara, an important bishopric on the Tigris,
was invaded and raided by the Persians. On that occasion many prisoners
were captured and enslaved.
The final attack before the Muslim conquest occurred between 603
and 614, when the whole East, including Mesopotamia, Syria, Armenia,
Cappadocia, Paphlagonia, Galatia, and later Antioch (611), Damascus
(613), and Jerusalem (614), was taken over by the Persians and held until
the end of the third decade of the century.98
As before, throughout this conquest the residents of the cities were
the main victims of these attacks, paying with their lives, their freedom,
and their public and private possessions for each defeat. Thus, for example, some time between 604 and 60699 the city of Dara was besieged
by Khusro Anu4sh3rwa4n for a year and a half; when the walls fell and
the city was captured all its inhabitants were put to death.100 The same
picture emerges from the Chronicon ad 1234 and the Chronicle of Sert.
The decision to fight to the end involved a high risk, and the Chronicon
asks:Was there any place which resisted him, which he did not destroy
and ravage, whose men he did not kill, and whose people he did not
deport?101 Sert recounts that Phocas was so busy killing his rivals that
he neglected to resist the Persians who were moving into his empire, and
ruining large parts of his country.102 It is evident, therefore, that the cities of the eastern Roman empire had suffered, especially throughout the
sixth and beginning of the seventh centuries, an almost endless series
of calamities, paying a heavy price for the tense and unstable relations
along the RomanSasanian border. Often enough they were left to fend
for themselves and to lick their wounds following heavy battles and cruel
24
defeats. It comes therefore as no surprise that these cities, so used to taking their fate in their hands, would negotiate their own surrender and
conclude their own agreements independently and separately once the
Muslim conquerors arrived on the scene.
The History of Surrender Agreements Made by Cities
Prior to the Muslim Conquest
Graeco-Roman tradition differentiated clearly between conquest by force
(Latin vi,103 Greek kata kratos104) and conquest by surrender (Latin deditio; Greek homologia105). While cities captured by force might be subjected to the killing and enslavement of their citizens and the loss of all
their property, cities that capitulated could negotiate the terms of their
surrender.106 The surrender usually involved an appeal by the surrendered
and a promise by the conqueror of pistis (in Greek) or fides (in Latin),
both meaning faith, trust, or good faith. As noted above, pistis (pl.
ta pista / fides) stood at the basis of all treaties and agreements,107 and
came to have the meaning of an assurance that that produces confidence,
a promise, engagement, word, assurance, confirmation, in particular a
given promise of protection or security; a guarantee.108 In Greek there
is a clear connection between the ancient horkia pista or faithful oaths
taken on the occasion of the surrender and pistis.109
The Roman term for surrender was deditio in fidem110 i.e. giving oneself up to the good faith or trust of the conqueror. The exact implications
of the classical deditio in fidem are unclear, and there has been an ongoing discussion of this question among classical scholars.111
Another option was the deditio per pactionem,112 i.e. surrender
accompanied by an agreement or pact (pactio, pactum, or sponsio in
Latin; pakton (pl. pakta),113 spondai (truce, treaty), and homologia in
Greek). This was, as defined by Coleman Phillipson, a covenant usually entered into by a general, usually on his own authority, engaging
to secure ratification by his government of the terms to which he had
consented.114 This act was based on giving words of promise or commitment, as is indicated in Latin, Greek, and Syriac. Thus the pact is
also called in Latin verba deditionis (lit. words of surrender); in Greek
homologia from logos (word), and in Syriac melta4 d-qya4ma4, word of
covenant. It is interesting to note that in Greek there is no word for surrender other than homologia. The implication of this is that surrender
was automatically accompanied by terms or an agreement, all of which
are enfolded in this one term.115
25
Yariv Shahal, whose work concentrates on this institution in Republican

Rome, demonstrates that at that time the order was the following:116 (1)
the pactio was agreed upon by the conqueror and the defeated; (2) this was
followed by the act of the deditio or formal surrender; (3) the pactio was
implemented; (4) the pactio was ratified by the Roman Senate. Although
a pactio had to be ratified by the Senate or the emperor, it was considered
a legal commitment on the part of the conqueror, and its breach was con
sidered a wrong (Greek adikema
)117 or injustice (Latin iniuria or iniusti118
tia). Cicero lists the pactio along with the foedus (treaty or covenant),
sponsio (agreement), amicitia (friendship) and societas (alliance) as one of
the rudiments of the legal relations between nations.119
It is thus clear that written agreements made between the conquering
general and the representatives of the surrendering city were an integral
part of the process of surrender in the Roman Republic, ensuring that
once the surrender was implemented, the conditions agreed upon would
be upheld and honoured by the conquerors.
As has been shown previously, regarding the treaties with the Sasanians,
the main elements that stood at the basis of the institution of treaties
remained intact, and were applied in a similar manner in the later Roman
empire. This also holds for the institution of the deditio, and specifically
the deditio in pactionem. Peter Heather adduces several examples dating
from the fourth century, which demonstrate that the usual mechanism
was in fact deditio, foedus, and then restitutio (restitution of the former
social order).120 In fact, he says:
As far as the diplomatic theory of the Roman state was concerned, the
Persians excepted, foederati were created by an act of surrender (deditio) on the part of the people involved, followed by a restitution of the
existing social order (restitutio) and the making of a negotiated agreement (foedus) To judge by its widespread appearance in a whole variety of texts, this theoretical framework was not only well developed and
widely applied, but knowledge of it also widely disseminated among the
literate classes of the Empire.121
Heather notes that while the framework remained fixed, the conditions
of the agreements varied according to the circumstances:
Roman diplomacy was a sensitive instrument for frontier management, which could be adapted to suit a wide variety of circumstances.
Its treaties could express quite different degrees of domination, different practical demands were made of its subordinate allies according to
need and possibility, native forms were utilized to add to the solemnity
26

of ceremonies, and it was standard, and eminently sensible policy to
endeavour, via payments, to create a class of amenable, semi-client kings
along its frontiers. The iron fist of imperialism was thus couched as
most successful imperialistic fists through history have been in a practical, well-informed velvet glove.122
What I strive to demonstrate below is that this same mechanism was later
to be employed by the Muslim conquerors, who adopted the frame of the
surrender agreement, including the surrender, the agreement that followed
it, and the restitution of the former social order, offering separate conditions in various circumstances according to their interests and possibilities. It is thus highly probable that the adoption of this mechanism was
initiated by the conquered populations, by whom it had been employed
for centuries. Additionally, although they were not as well versed in this,
the Muslim conquerors, as has been shown above, were also familiar
with this mechanism, and were thus readily willing to adopt it.
Such events of offers to surrender, negotiations, surrender of cities
and the signing of pacta/pakta or surrender agreements by individual
cities are indeed recorded by Byzantine historians in the period preceding
the Muslim conquest, although these are not as numerous as one would
expect. This is most probably due at least partially to the fact that ancient
historians tended to concentrate their efforts on the main events, in this
case on the main pacts and agreements signed between the rulers, as has
been shown above, rather than report about local surrenders of one city
or another. Thus they may often mention the latter in passing, or not
mention it at all, while the major pacts are often described in great detail.
Since obviously during the Muslim conquest no such global pacts either
with the Byzantines or with the Sasanians were signed, it was necessary
to adduce the local surrenders of cities one by one.
Having noted this, there are however several reports of such events.
The treaty concluded independently between the city of Antioch and the
Persians during the days of the Roman emperor Trajan has been referred
to above:123
The Persians captured Antioch the Great and occupied it not indeed by
force of arms but by an amicable agreement and a treaty (kata syntaxin
phyliken kai pakta) by which they controlled and guarded it for the
Persian Emperor Sanathroukios. For the Antiochene dignitaries had of
their own accord set terms for peace and submission (kata idian proairesin pakta eirenes kai upotages stesanto4n) through an embassy to
the Persian Emperor.124
27
In this case the dignitaries of Antioch initiated the signing of the treaty,
and signed it without the consent of the emperor.
A brief mention made by Eutropius relates to a series of surrenders of cities (deditio) during Julians expedition against the Sasanians.
Unfortunately, however, he does not supply any details:
Accordingly Julian took possession of the state and after vast preparations waged war against the Parthians. I was also a member of this
expedition. He accepted the surrender of or forcibly seized several of
towns and fortresses of the Persians (aliquot oppida et castella Persarum
in deditionem accepit vel vi expugnavit).125
This unsatisfactory mention is filled in by Ammianus Marcellinus, who

adduces some interesting cases of offers to surrender, negotiations, and
agreements. He describes the case of two towns on the Euphrates on
Julians route:
After these successful operations we reached a fortress called Thilutha,
situated in the middle of the river, a place rising in a lofty peak and fortified by natures power as if by the hand of man. Since the difficulty and
the height of the place made it impregnable, an attempt was made with
friendly words (as was fitting) to induce the inhabitants to surrender (ad
deditionem incolae temptati mollius); but they insisted that such defection then would be untimely. But they went so far as to reply, that as
soon as the Romans by further advance had got possession of the interior, they also would go over to the victors, as appendages of the kingdom (regnorum sequelas victoribus accessuros). After this, as our ships
went by under their very walls, they looked in respectful silence without
making any move. After passing this place we came to another fortress,
Achaiachala by name, also protected by the encircling river, and difficult
of ascent; there too we received a similar refusal and went on.126
Another relevant case is that of the walled city of Pirisabora described

later on Julians journey. During the siege laid on this city there was fierce
fighting for some time until the breaking-point when Julian decided to
build a special war-machine:
To this huge mass, which would rise above the battlements of the lofty
towers, the defenders turned an attentive eye, and at the same time considering the resolution of the besiegers, they suddenly fell to their prayers,
and standing on the towers and battlements, and with outstretched
hands imploring the protection of the Romans, they craved pardon
and life (fidem Romanam pansis manibus protestantes vitam cum venia
postulabant). And when they saw that the works were discontinued,
28

and that those who were constructing them were attempting nothing
further, which was a sure sign of peace, they asked that an opportunity
be given them of conferring with Ormizd. When this was granted, and
Mamersides, commander of the garrison, was let down on a rope and
taken to the emperor, he obtained (as he besought) a sure promise of
life and impunity for himself and his followers (vita cum inpunitate sibi
consortibusque suis firmiter pacta), and was allowed to return. When he
reported what he had accomplished, all the people of both sexes, since
everything that they desired had been accepted, made peace with trustworthy religious rites (pace foederata cum religionum consecrationibus
fidis). Then the gates were thrown open and they came out, shouting
that a potent protecting angel had appeared to them in the person of a
Caesar great and merciful. The prisoners numbered only 2,500; for the
rest of the population, in anticipation of a siege, had crossed the river
in small boats and made off. In this citadel there was found a great
abundance of arms and provisions; of these the victors took what they
needed and burned the rest along with the place itself.127
Here the mechanism of surrender and the agreement are described in

detail: the inhabitants seek assurance of protection from the Romans,
that is the fides Romana, and ask for pardon and for their lives; work
on the war machine is ceased a sign of receptiveness on the part of the
Romans; the besieged then request a meeting with the Roman general (of
Persian origin), Ormizd; this is granted. The city gates however, remain
closed, while Mamersides, commander of the garrison, is let down via a
rope; surprisingly, rather than meeting the Roman general he is brought
to Julian himself, who promises them life and impunity; Mamersides then
returns to the city, and only after having passed this information on and
conferred with the people of the city does he announce the surrender of
the city. It is then that peace of foederati is made (pace foederata), i.e.
unequal peace, sealed by oaths of fides.
Unlike cases in which the surrender was made early on, here it followed a long period of fierce fighting; the conditions of the sworn agreement were therefore minimal, and included only the life and impunity
of the remaining inhabitants. It may be that in such a case the terms are
agreed upon orally and there is no need for a signed document, although
it is quite possible that the promise of the protection of their lives and
their impunity was given in writing.
During the rest of the fourth century and most of the fifth, no doubt
as a result of the relative peace that prevailed throughout this period,
there is no information regarding surrender of cities. Further information is found regarding the fifth and sixth centuries. Several examples
29
of peace agreements made by cities in Mesopotamia with the Sasanians

are found in Procopius Wars. After Khusro Anu4sh3rwa4n besieges Edessa
in 544, attacks it, and finds it impregnable, he offers its inhabitants and
the Roman general an agreement or treaty (symbasis; homologia). His
interpreter Paul calls for Martinus, the Roman commander in order
that he might make arrangements for the agreement (symbasis). Thus
Martinus came to a conference with the commanders of the Persians, and
they concluded an agreement by which Khusro Anu4sh3rwa4n received five
centenaria from the inhabitants of Edessa, and left them, in writing the
terms of the peace (en grammasin autois te n homologian apelipe) not to
inflict any further injury upon the Romans.128 Although this description
is short, it is nevertheless clear that it follows the accepted norms according to which one side offers to hold peace negotiations, the generals of
both parties meet and confer, and the terms are agreed upon. The terms
are then set down in writing and the agreement is signed.129 The term
used here, homologia, is the accepted term employed to convey terms of
peace or surrender, as mentioned earlier.
A similar attempt, made at the beginning of the same siege, failed
when the conditions offered by the Persians seemed unacceptable to the
Romans, as Khusro Anu4sh3rwa4n had demanded an amount unacceptable
to the Edessans, who had offered to pay as much as they had provided
previously, when he threatened them following his capture of Antioch.130
It is thus evident that Khusro Anu4sh3rwa4n was interested in accumulating
wealth and property rather than in controlling the area, and employed
the Roman mechanism of surrender and agreement in order to achieve
this end.131
Another case in point is that of Beroea (540). Just like in the case of
Edessa, Khusro Anu4sh3rwa4n was prepared to negotiate for peace in return
for a lump sum of money. While the inhabitants of Edessa had managed to
purchase for themselves a surrender agreement which included their lives,
as well as their property and freedom, actually a restoration of the previous state of things, the people of Beroea were unable to do so. Therefore
their bishop, Megas, informed him that they had no money and entreated
him to grant him only the lives of the men. Khusro Anu4sh3rwa4n fulfilled
his request and binding himself by an oath, gave pledges to all on the
acropolis (diomosamenos hapasi tois en akropolei ta pista edoke
4 ).132 In
this case, it is clear that the pista, i.e. the assurances or guarantees, meant
their lives only; there was no written agreement including additional
stipulations, and, as Procopius goes on to say:The Beroeans left the
acropolis free from harm, and departing went each his own way.
30
In the case of the city of Sura (540), the inhabitants, believing themselves defeated, sent their bishop to Khusro Anu4sh3rwa4n to beg that the
town be spared. Bringing along a gift of fowls, wine, and loaves, the
bishop promised that the men of Sura would give him ransom worthy of
themselves and the city which they inhabited.133 As in the previous case,
this is once again an offer of a surrender agreement. Khusro Anu4sh3rwa4n
pretended to consider the offer, but in fact carried out a surprise attack,
burning the city to the ground, killing many of its inhabitants and taking
the rest as hostages.134
It was only afterwards that an agreement was made between Khusro
Anu4sh3rwa4n and the bishop Candidus of neighboring Sergiopolis. The
latter vouched for the hostages and promised to pay Khusro Anu4sh3rwa4n
for their release, the payment to be made at a future time.
Khusro therefore requested him to set down in a document the terms of
surrender (ho Chosroes exiou en biblidio4 ten homologian aphenta) that
he would give the money at a later time Candidus did as directed
and swore the most dire oaths (horkous deinotatous omo4moko4s) specifying that he should receive the following punishment if he should not
give the money at the time agreed upon, that he should pay double
the amount and should himself be no longer a priest as one who had
neglected his sworn promise.135
The same mechanism was applied during the Sasanian conquest of the
East at the beginning of the seventh century. Sebeos notes that:
The [Edessans] because of the multitude of [Persian] troops and their
victory in the engagements and since they had no expectation of salvation from anywhere, parleyed for peace, and requested an oath that
they would not destroy the city. Then, having opened the city gate, they
submitted. Similarly Amida and Tella and Resheina and all the cities of
Syrian Mesopotamia willingly submitted and were preserved in peace
and prosperity. They went to the city of Antioch, and these too willingly submitted with all the cities and their inhabitants, fleeing from the
sword of Phocas.136
The people of Edessa asked for peace and negotiated the terms; only
after this had been done did they open the gates of the city and submit.
Other cities that submitted apparently did the same, as they were preserved in peace and prosperity. The agreement reached in Edessa, as
well as in the case of the rest of the cities of conquered Mesopotamia,
thus allowed for a restoration of the previous conditions following the
surrender.
31
When describing these same events the Syriac chronicles chose to use
the idiom they gave their word (yahbu4 melta4) or word of oath, or word
of assurance for fides/pistis. Although this is not a literal translation it
does convey the same basic meaning of promise of assurance which is
referred to by the Latin and Greek sources. It should be noted that in
Greek the word logos can signify promise as well.137 The Chronicon
ad 1234 thus reports that after the Euphrates had been crossed by the
Persians, Antioch, Apamea, and Emesa were conquered. The latter
city was given the word by them and they submitted to the Persians.
Damascus chose to submit, and the Damascenes gave their word that
they will raise tribute to the Persians.138 Later on, when the Byzantines
managed to defeat the Persians within the city and they surrendered, they
gave them their word of covenant (melta4 d-qya4ma4) that they will leave
and go to their land.139 When describing the events of 506/7, Joshua the
Stylite reports that the Persians and the Byzantine magistros drew up an
agreement (qyama
4 4) and made peace (shayna4). They composed written
terms (kta4be ) between them.140

Although in both of these cases the description is very succinct, the
previous evidence regarding the events of the Persian conquest leaves no
doubt that these reports refer to the same institution of surrender and
agreement.
The same mechanism was once again used upon the reconquest of
these territories by Heraclius towards the end of the third decade of the
same century. According to the Melkite Alexandrine Patriarch Eutychius,
the Jews of Tiberias and the Galilean mountains welcomed the victorious
Heraclius with gifts and asked him for an ama4n, which he granted them,
giving them a written agreement. Here it was not a city but a religious
community that was requesting an agreement.
However, somewhat later, the Christians begged him to retract this
agreement and to kill the Jews for having assisted the Persians in the mass
murder of Christians and in the destruction of churches in Jerusalem as
well as in Tyre. Heraclius refused, saying:How could I consider it lawful
to kill them after I have given them ama4n141 and written an agreement to
that effect? You know what is incumbent upon one who breaks the agreement (ma yajib an man naqada al-ahd)! If the agreement (ahd) and the
ama4n142 will be broken [by me] it will be a disgrace and terrible defamation for me.143 Nonetheless, he finally consented to break the agreement,
having been persuaded by the Christians that they would help him atone
for the sin. Heraclius statement fits in well with the Roman concept as
expressed by Polybius already noted above144 that the breach of such
32
a surrender agreement was considered a wrong (Greek adike ma)145 or

injustice (Latin iniuria or iniustitia).
To sum up this discussion:both the grave and traumatic experiences
suffered repeatedly by the cities, especially throughout the sixth and early
seventh centuries, and the established tradition of concluding surrender
agreements in the Roman and Byzantine East generated the signing of
surrender agreements between the Muslim conquerors and the inhabitants of the conquered cities. Throughout the Muslim invasion, as in the
century that preceded it, the empire was unable, most of the time, to
protect the inhabitants of the cities. The signing of surrender agreements
thus seemed the most reasonable option, if the right terms of surrender
could indeed be obtained.
surrender agreements made following

the muslim conquest
We come now to the agreements made between the Muslim conquerors
and the conquered cities and territories. As will presently be shown, these
agreements were actually inspired by the long-existing tradition of agreements that was prevalent in the territories conquered by the Muslims in
the seventh century.
Several characteristics inherent in the agreements with the Muslims
demonstrate that they indeed continue this tradition:
The Terminology
Ama4n
The most common term in Muslim literature defining the relationship
between the Muslim conquerors and the surrendered city is ama4n. The
term is translated as safety, protection, safe-conduct, and an assurance
of safety and security. The relevant idioms are talaba minhu al-ama4na
(he demanded from him an assurance of safety) and atawhu ama4nan
(hegranted him an assurance of safety), as well as dakhala f ama4nihi (he
entered into his protection).146
The meaning of ama4n is thus parallel to the Greek pistis and the Latin
fides, all of which denote faith, trust, protection, and assurance of security. It is most interesting therefore that the term ama4n does not appear
as such in the Qura4n, and that it actually developed and came into usage
at the time of the conquest, as noted by Joseph Schacht:The institution
of ama4n continues, in fact, the pre-Islamic Arab institution of jiwa4r by
33
which a stranger, who was in principle outlawed outside his own group,
received for his life and property the protection of a member of a group
to which he did not belong, and therefore the protection of the group as
a whole.147
Emile Tyan notes that the institution of ama4n played a very important
role in international law; it made it possible to avoid some of the consequences of the basic principle of a permanent hostile state between the
Muslim world and the rest of the world, permitting the sojourn and the
free activity of strangers in Muslim territory.148
Sura 9 of the Qura4n, discussing the relations between the believers
and the polytheists, chooses to use the terms ahd, dhimma, and jiwa4r
rather than ama4n.
The term ama4n is also absent from the first agreements referred to and
cited by the sources, which were made by the Prophet with cities in the
Arabian peninsula such as Khaybar, where the term consistently used is
sa4lahu4huala4 haqn dima4ihim (wa-tark al-dhuriyya).149 The same terminology is used in relation to Fadak.150 In reference to Ayla the root mn
appears in a verb:wa-kataba lahum kita4ban bi-an yuhfazu4 wa-yumnau4
(and he wrote them a document that they will be protected and will
be secure).151 Regarding Najra4n the text of the agreement cited on the
authority of Yahya4 b. A"dam reads: wa-li-najra4n wa-ha4shiyatiha4 jiwa4r
Allah wa-dhimmat Muhammad al-nab rasu4l Allah (and to Najra4n and
its dependants the protection (jiwa4r) of God and the covenant (dhimma)
of the prophet Muhammad, the Messenger of God).152
In two other cases the root mn appears as a conjugation of the verb:the
agreement with Bahrayn, cited by Abu4 Ubayd, says that if they accept
Islam they will be safe (in a4minu4 fa-innahum a4minu4n),153 and in the
agreement with Maqna4 the version cited by al-Bala4dhur3, dictated to him
by people from Egypt, reads: fa-innakum a4minu4n walakum dhimmat
Alla4h wa-dhimmat rasu4lihi (From the time this letter reaches you will
be safe, and you have the protection of Allah and the protection of his
messenger); yet immediately following this comes:wa-inna rasu4l Alla4h
yuj rukum mimma yuj r minhu nafsahu (Against whatever the Prophet
of Allah protects himself, he will protect you).154 It is quite clear in these
cases that the dominant terms of protection here are jiwa4r and dhimma.
Only in later agreements does the noun ama4n start to appear regularly as the assurance of protection and, most important, the term jiwa4r
disappears almost completely in connection with the protection of the
non-Muslims.155 It does, however, continue to be in wide use concerning
the Muslims. The term dhimma, on the other hand, does continue to be in
34
use, although it is much less common than ama4n in this connection.156 It

thus seems that the later sources here were indeed following the terminology of the transmitters, which in their turn apparently reflected the usage
of the early agreements.
How did the term ama4n come into usage? Nouns stemming from
the root mn, which carry the connotation of fidelity, firmness, trust, are
recorded from antiquity in Hebrew, Aramaic, and Arabic.157 The word
qya4ma4 used in the Syriac idiom melta4 d-qya4ma4 for word of oath or
promise of assurance is actually equivalent to the Aramaic amn, both
meaning literally stability, firmness which may be trusted, and therefore
also assurance and covenant.158
The word ama4na appears in the Qura4n with the meaning of trust
(23:72, 2:283, 4:58, 8:27), and twice, citing the same passage, it appears
as a synonym of ahd, i.e. covenant (23:8, 70:32).
It is thus quite evident that the root amn served in the Semitic languages as faith, trust, assurance, and treaty. Just a small adaptation was
therefore needed to create in Arabic a new term, ama4n, which acted as
a specific unique legal term equivalent to pistis/fides bearing the specific
meaning of assurance or promise of safety and protection, rather than
just a general trust and differentiated from the Arabic jiwa4r, whose implications were too far-reaching and were too suggestive of insider status
where non-Muslims were concerned. The verbs then used in Arabic with
ama4n also point to the influence of the outside world: thus the ama4n
is given (Arabic ata4hu al-ama4na = Greek pistin didonai, Syriac yahbu4
melta4; and one enters into the ama4n:Arabic dakhal f ama4nihi = Greek
eis pistin elthein). Apparently the people who surrendered asked for pistis, fides, or melta4, as this was the long-accepted technical term. This was
translated into Syriac as giving word (melta4 or melta4 d-qya4ma4) and into
Arabic as ama4n.
Ama4n seems therefore to be a loan translation of the term pistis/fides,
its meaning being a basic assurance of protection as was usually granted
in such agreements in the long-standing tradition of the Graeco-Roman
world.
Baqt
The word baqt comes from the word pakton, which in its turn is a
Hellenized form of the Latin pactum. In the Hellenistic world this term
was used for both a compact of mutual obligations and its connected
payments. The Arabs used this expression for the tribute that was raised
by Christian Nubia.159
35
As noted in the beginning of this chapter, the letter discovered in the

excavations in Egyptian Nubia, written in 141/758160 and sent by the
Abbasid
4 governor in Egypt to the king of Nubia and Muqurra, contained
a demand to keep the baqt according to which you were given the agreement.161 It is already clear in the letter that the word had evolved to mean
the payment to be made according to the pakton. Al-Maqrz
3 ,3 who adduces
the text of the agreement in a chapter entitled Dhikr al-baqt , takes trouble to explain the meaning of this term, speculating on its origin.162
It is obvious that the terminology used here for the agreement or
the obligations rooted in it originated in the world of the surrendering
party, i.e. the Nubians, who chose to use the common and well-known
Greek term, and, as I will attempt to show, also contributed their perception of a surrender agreement which originated in their case in the
Graeco-Roman culture embedded in Egypt. The concept of the baqt was
so strong that it survived for centuries, long after its original meaning
was forgotten.
The Verb qa4taa ala4
This unusual usage of the verb, which is used rarely by Bala4dhur3 as a
synonym for sa4laha ala4, denoting peace made on condition that ,
thus:wa-qa4taahu ala4 bila4dihi,163 or wa-qa4ta ahu ala4 ita4wa164, talabu4
al-ama4n ala4 an yatara4jau4 ila4 ardihim fa-qu4tiu4 ala4 khara4j.165 This special usage may also originate in the term to cut a covenant in Greek
(horkia pista temnein),166 Hebrew karat brit.
The terminology used in the Syriac text regarding the events of the
Arab conquest in the Chronicon ad 1234167 reflects the common usages
noted until now. Thus the traditional idioms used regarding agreements made prior to the conquest, which echo the Greek/Latin terminology, remain the same when agreements made with the Muslims are
described:168 giving assurances is yahbu4 melta4; an agreement or covenant is kta4ba4 or qya4ma4.169 Kta4ba4 here refers to a written agreement,170
while qya4ma4 w-ma4wma4te i.e. covenant and oaths, is an ancient expression in Aramaic171 parallel to the Greek horkia kai pistis; Umar writes
the inhabitants of Jerusalem letters patent as they wished concerning
their churches and their customs (w-ktab l-hu4n sig liy(u4)n ayk d-baw
al-ida4tayhu4n w-na4mu4sayhu4n)172 the word sig liyu4n originating in the
Greek sigillon. The stipulation that they be allowed to hold onto their
customs (nomoi) is a common one in the ancient treaties where the
expression is kata patrious nomous (according to the customs of their
ancestors).173 All of this shows that in the view of Syriac historiography
36
the agreements made with the Muslims were part of the same ancient
mechanism with which they were familiar.
It should be added here that when drawing up these documents in
the Arabic language, the Muslims introduced their own terminology as
well, including terms such as dhimma, which became the most dominant expression in their relationship with non-Muslims, the expression la4
yasullu4 wa-la4 yaghullu4, used in the vassal-treaty type (see nn. 25697),
an expression which appears in the Hudaybiyya agreement, the provision of nas ha (advice), which appears in ahd al-umma, and others, thus
adapting these agreements to their own language, tradition, and mentality when it suited them.
The Procedure of Surrendering
Like their Graeco-Roman predecessors,174 Muslim commanders made a
distinction between conquest of a city by force (anwatan) or through an
agreement (sulhan).
It has been claimed by Noth that many reports, especially regarding
the conquest of Egypt and Iraq, did not originate at the time of the conquest, but were rather a product of Umayyad political agendas.175 While
it is of course possible that specific reports regarding the conquest were
modified to suit certain interests at later times, the institution that stands
at the basis of these reports was nevertheless one that originated in the
Graeco-Roman world and was adopted by the Muslim conquerors.
Muslim sources describe a process analogous to the process known
from the pre-Islamic Near East:once a city surrendered, it received an
ama4n = pistis/fides/melta4. The granting of an ama4n was usually accompanied by a document listing the conditions that were agreed upon in
the sulh (peace agreement) (kita4b, kita4b ama4n, ahd, sulh, and the verbs
wa-sa4lhu4hum ala4; ishtaratu4hum ala4; qa4tau4hum ala4, atu4hum ama4n
ala4, a4manu4hum ala4 followed by the conditions).176 This was therefore a conditional surrender, which was often accompanied by a written
agreement:homologia, pakton/pactum, qya4ma4, or kta4ba4, Arabic kita4b,
ahd, aqd.
The process of obtaining a surrender agreement is well exemplified by
the case of the city of Marwaru4dh, located 260 km south-east of Marw,
which, having rebelled against the Muslims, surrendered in 32 H/652
CE. According to al-Tabar3s description177 the inhabitants were besieged
by the Muslims and finally surrendered. An ambassador representing
the marzuba4n of Marwaru4dh emerged from the city with a letter, asking
37
the Muslims for safe conduct, and they gave their consent (a4minu4n
fa-a4manu4hu). The letter expressed submission to the Muslims:We praise
God, in whose hands are the turns of fortune, who transfers kingship as
he pleases. The letter then proceeded to ask for peace terms with you
(musa4lahatuka wa-muwa4daatuka)178 on the [same] lines as my grandfathers submission. The marzuba4ns letter then proceeded to lay out specific terms including the payment of 60,000 dirha4ms to the Muslims, the
confirmation of his control over the territory and possessions given to his
great-grandfather by Khusro as a reward for killing the serpent that was
intimidating the inhabitants of the area at the time, the exemption of all
the members of his household from tribute, and an assurance of his title
as marzuba4n of the area. I have sent you my nephew Ma4hak to seek a
pact with you concerning my requests (li-yastawthiq bika bima saaltu).
The surrender was thus offered depending on acceptance of the conditions stipulated in the letter. Al-Ahnaf,
the commander representing the

Muslims here, replied in a letter again cited fully by al-Tabar
.3 He emphasized that the letter was written on behalf of all the Muslims who were with
him, and gave his consent to the terms presented, citing them one by one.
He also added his own condition:You, together with the heavy cavalry
who are with you, are likewise obliged to aid the Muslims and fight their
enemies in return you are owed the aid of the Muslims. Once both sides
had consented, the agreement was witnessed, dated, signed, and sealed.
A similar, although less detailed, description is adduced by al-Bala4dhur3
in the case of Jurja4n, transmitted by the elders of Dab3l:179
As Hab3b [b. Maslama] was advancing against the patrician of Jurja4n,
he was met by a messenger of the patrician and the inhabitants of the
town, and presented a written message and asked for a treaty of peace
and security (wa-saalahu kita4ba sulhin wa-ama4nan). Accordingly,
Hab3b wrote to them: Your messenger, Nuqla, came to me and my
companions the Believers saying on your behalf that we are a nation
whom Allah has honoured and given superiority You also stated that
you would like to make peace with us (innakum ahbabtum silmana). As
for your present, I have estimated its value and considered it a part of
your tax. I have written you an ama4n (wa-katabtu lakum ama4nan) and
inserted one condition in it. If you accept the condition and live up to it,
well and good. Otherwise, then be you apprised of war [that shall come
upon you]180 from Allah and His Messenger.
Here again, the non-Arabs approached the Muslims through an envoy

carrying a letter, admitting to the superiority of the Muslims, and asking
for a peace agreement. On the basis of this letter, the contents of which
38
are not revealed here, Hab3b wrote an agreement, adding a condition

considered imperative by him.
We thus see here a case of conquest through surrender (sulhan) in
which the procedure of the surrender on terms, familiar from the GraecoRoman tradition, is clearly followed. It is evident here that these are written agreements, made on a basis of a written letter of surrender.
As already noted above,181 Muslim sources report that such agreements were indeed being written in the days of the Prophet. Thus, in
the cases of al-Ta4if,182 Bahrayn,183 Ayla,184 al-Jarba4 and Maqna4,185 and
Najra4n,186 the format of the agreement is a kita4b, i.e. a document. In the
cases of Maqna4, and particularly Najra4n, especially detailed agreements
are adduced. However, these early agreements were not written in the terminology and format that would characterize the later agreements, and
seem to indicate that at that early date the tradition, which was common
outside the Arabian peninsula, had not yet infiltrated Arabic culture as it
did later, at the time of the conquest.
Actual Copies of the Agreements
Early Muslim literature often mentions the existence of copies (nusakh,
sing. nuskha) of these agreements, and some of the versions cited are
claimed to have been copied from such copies. Thus, regarding the agreement of Maqna4, al-Bala4dhur3 says:An Egyptian told me that he saw with
his own eye their agreement (kita4buhum) on a red parchment, the writing
partly effaced, and he copied it and he dictated it to me as follows.187
As for the copy of the agreement with Najra4n, al-Bala4dhur3 reports that
al-Husayn heard from Yahya4 b. A"dam, who said:I copied the copy of
the agreement of the Prophet (nuskhat kita4b rasu4l Alla4h) to the people of
Najra4n from the kita4b of a man who took it from al-Hasan b. Sa4lih.188
Although these reports should not be taken at face value, they nevertheless indicate that the actual existence of written copies was taken for
granted.
Referring to the existence of written agreements in Egypt, Ibn Abd
al-Hakam cites a tradition in the name of Ubayd Alla4h b. Ab3 Jafar, who
said that there were three people who kept a copy of the ahd with Amr,
stating their names and their dwelling places.189 He also adduces a conflicting tradition transmitted by Zayd b. Aslam according to which Amr
had a wooden box (tabu4t) in which he held all the agreements he made,
and there was no agreement from Egypt there.190
39
This contradiction illustrates well enough that information regarding

these copies was often undependable. Yet it does emphasize the fact that
agreements were indeed written, and that the documents were expected
to be kept by both parties.
Al-Qa4d3 notes that traditions regarding written versions of agreements
cannot serve as proof of the existence of these copies, and believes that at
least some of the agreements made were oral agreements.191 This cautious
attitude should, however, be reconsidered in the light of the prevalent
tradition in the Near East:as shown above, beside the official documents
exchanged by both sides, additional copies of the agreement were indeed
made and kept by both sides for administrative use. Thus, although we
cannot of course take all these reports at face value, it cannot be doubted
that copies of agreements existed, and above all that the inhabitants of
the city that received the agreement preserved both the original and its
duplicates carefully, as these served as their insurance policy. This does
not exclude the possibility that in exceptional cases the agreement was
an oral rather than a written one. This must have been true, during the
Byzantine period as well as during the Arab conquest, in cases when the
surrender came after heavy fighting and those surrendering were granted
their lives only, or in cases of extremely insignificant or marginal places.
Written agreements, however, were clearly the rule.
Al-Qa4d3 herself notes, in fact, that these agreements were renewed
when a new governor was appointed, or when a new caliph came into
power.192 This, too, is in line with the Graeco-Roman model, in which
agreements were to be renewed upon the death of one of the parties.193
Al-Qa4d3 adduces several examples:the new governor of Egypt, Yaz3d b.
Abdalla4h al-Hadram3 renewed the agreements of Barqa and Anta4bulus.
In the process of the renewal, the former agreement was brought to
him by Ibn Diyya4s al-Nasra4n3 al-Qubt3 (a4ta4hu bi-kita4b ahdihim);194 the
Nubian baqt and the agreement with Ruha4 were renewed by Umar b.
Abd al-Az3z.195 In the case of both Ruha4 and BarqaAnta4bulus, it was the
locals who produced a copy of the agreement (in Ruha4 it was the bishop
of the city who was asked to produce one). The same is true of Tifl3s,
whose people brought the agreement given to them by Hab3b b. Maslama
to al-Jarra4h b. Abdalla4h al-Hakam3, who renewed it.196 Al-Qa4d3 infers
that the Muslims did not have a sufficiently developed administrative
system in the days of Umar and Uthma4n for the authorities to keep a
copy, and therefore depended on the copy held by the conquered.197 This
does not necessarily follow. It is more probable that the dhimm s concerned simply wanted to be sure that it was indeed the same version that
40
was being renewed, while the governor or caliph who renewed it could
inspect the terms before signing the new agreement. If we are to judge
by the agreements cited by al-Bala4dhur3, al-Tabar
3, or Abu4 Ubayd, there

must have been ample copies of such agreements. Having said this, it is
of course quite likely that not all the agreements were zealously renewed.
Rather, it may be assumed that this often occurred when there were relevant issues at hand which prompted one party or the other to insist
upon a renewal of the contract.
The fact that the original agreements were considered legally binding documents is illustrated by an incident in Damascus around 800
CE related by Ibn Asa4kir.198 He cites Ibn al-Mualla4 (d. 286/899)199 who
reported that he had read a document (kita4b sijill) written by the qa4d of
Damascus, Yahya4 b. Hamza (d. 176/792).200 The document recounts how
the Christians of Damascus came to the qa4d claiming that the Muslims
had taken over their churches (ghalabu4hum ala4 kana4isihim); they asked
him to fulfil the covenant (ahd) given them, and the agreement (kita4b)
written for them by Kha4lid b. al-Wal3d. The qa4d then cited the agreement
produced by them. He checked the agreement and found that it was their
own proper agreement (wa-qaratu kita4bahum fa-wajadtuhu kha4ssatan
lahum), he also checked the jizya and confirmed that this was indeed
their proper jizya. Once he was satisfied, he ruled that the people they
had sued should if possible return the property itself to them, or if not,
compensate them adequately. The same view is also expressed by Abu4
Yu4suf in his Kita4b al-khara4j, and by al-Sha4fi3 in Kita4b al-umm,201 who
both insist that the uhu4d were binding legal documents. This is especially
striking if we consider the fact that the agreements were first and foremost documents assuring the protection of the dhimm s:their lives, their
personal and public property, and their religious freedom. In fact, as shall
be demonstrated in the following chapter, by the time of Abu4 Yu4suf and
Yahya4 b. Hamza the qa4d of Damascus, these agreements often seemed
an encroachment upon the rights of the Muslims. Nevertheless, they were
respected by the Muslim authorities.
The Structure of the Agreements
Al-Qa4d3 has already demonstrated that the structure of the agreements
is uniform, and usually includes most of the following elements:the basmala; the names of the giver/s and the receiver/s of the ama4n; the territory
included; the stipulations; the witnesses; the scribe; the date; the signatures; and, in several cases, the seal. Al-Qa4d3 notes that the uniformity of
41
the structure supports the authenticity of the documents.202 Indeed, this

formal legal structure of the ama4n agreements as well as their elaboration
and sophistication, which made them so suspect in the eyes of many, was
in fact not a late anachronistic invention of Muslim jurists, but rather
an adaptation of the common Near Eastern tradition, specifically the
Graeco-Roman tradition in the East.
As we have seen above, in the Byzantine period formal agreements
were indeed being written down, dated, witnessed, signed, and sealed,
and copies were made for the use of both parties. They usually included a
detailed list of stipulations as well as terms of validity. Rather than making them suspect, the structure of the Muslim agreements, as well as their
uniformity, confirms their connection to the ancient treaties and supports
their authenticity.
The Characteristics of the Agreements

The extant versions of agreements made by the Muslim conquerors comprise several elements which are common in pre-Islamic Near Eastern
treaties too.
Formulae of Oaths
Ancient treaties were witnessed by the gods, and were secured through
invocations of the gods, oaths, and imprecations.203 In several Muslim
agreements we find the formulae shahida Alla4h wa-mala4ikatuhu (God
and his angels have witnessed)204 or shahida Alla4h wa-mala4ikatuhu,
wa-kafa4 bi-Alla4hi shah dan (God and his angels have witnessed, and
God is a sufficient witness).205 In many others, the dhimma is not dhimmatu al-muslim n but that of Allah and his Prophet. Thus for example in
the agreement with Misr:206 For the terms of this document (ala4 ma4 f
hadha4 al-kita4b) the covenant of Allah and his protection and that of his
Messenger (ahd Alla4h wa-dhimmatuhu wa-dhimmatu rasu4lihi), and that
of the Caliph, the Commander of the Faithful, as well as the protection
of the Believers, are guarantees.207
The involvement of the deities is even more accentuated in the case
of the Nubian baqt as cited by al-Maqr3z3; in case of the breach of the
covenant:
this hudna and ama4n will be revoked (bariat minkum) and we both will
return to a state of parity until God judges between us. He is the best
of judges. In this matter [we take] upon ourselves the covenant of Allah
42

and his promise (ahd Alla4h wa-mitha4qihi) and his obligation and the
obligation of his Messenger Muhammad and you owe us the most
esteemed obligation of what you profess, that is the obligation of Christ,
and his Apostles and the obligation of those you venerate of your religion and people. God is the witness between us and you as to this matter
(Alla4hu al-shahd baynana wa-baynakum ala4 dhalika).208
Here the need to secure the covenant through a binding oath by both parties, each to his own deity, seems especially significant and noteworthy.209
It may well be that the commitment of the deity in the treaty was also
demanded by the surrendering party who, as demonstrated above, saw
the oaths as an inseparable part of the agreement, which guaranteed its
fulfilment.
As stated above, in the later Roman and Byzantine periods many of
the treaties were no longer witnessed by the gods.210 This is also true of
many Muslim agreements, which merely have the basmala invocation at
the head of the document.
The Stipulations
The stipulations themselves may be divided into two different types, both
of which seem to have had their origins in the ancient traditions of the
surrender treaty and the vassal treaty.
The Surrender Treaty.211 This model represents the total surrender
of the inhabitants of the city, which would thenceforth be ruled by
the Muslims. It is found in most of the agreements regarding the cities in Egypt, Syria, and Mesopotamia, including agreements relating
to Ruha4,212 Damascus, Misr,213 al-H3ra, Ardab3l, Bihquba4dh, Baalbakk,
Jerusalem, Ludd, and Filast3n, and it appears to emanate from a Roman
Byzantine model. It is basically a kita4b ama4n enumerating the obligations of protection that the Muslims take upon themselves, including (in
various combinations) the protection and safety of the inhabitants of
the city, their children, monks, priests, property, mills, churches, monasteries, and crosses; and the obligation to allow them to hold onto their
ancestral customs, and not to coerce them to accept Islam,214 in return
for the payment of jizya.
In some cases it is phrased simply as we owe you protection and you
owe us jizya:lakum al-dhimma wa-alaykum al-jizya,215 or, in the case
of al-H3ra:he made an agreement with them in return for the payment
of 190,000 dirha4ms and [their] protection: a4hadahum ala4 tis n
wa-miat alf dirha4m wa-ala4 al-mana. This payment was to be made
on a yearly basis. In addition, the agreement expressly says that if they
43
are not defended as promised, they need not pay their dues until their
defence is restored.216
This model does not, in fact, demand anything from the surrendering
city other than the recognition of Muslim rule in the form of the payment
of jizya, while the obligations of the Muslims are in most cases (though
not in these last two) enumerated in detail. Such agreements, which were
probably the most any surrendering city could wish for, are attested in the
Graeco-Roman tradition.
We have already cited the treaty between Rome and Pharos in which
the polis, the ancestral customs, and the lands were given back to the
citizens of Pharos following their deditio.217 A similar agreement is cited
in a letter of the consul Gnaeus Manilius Vulso in 189 BCE to the inhabitants of Heraclea. Following the defeat of Antiochus III at the battle of
Magnesia in 190 BCE, large parts of Asia Minor were transferred from
Seleucid to Roman rule. Cities that had surrendered to Rome prior to this
battle were awarded their liberty, lands, and laws.218 The Roman Senate
provided guidelines based on which a commission headed by Vulso acted
in 188 BCE.219 The process of surrender described here is very similar to
the one described above (see The Procedure of Surrendering), as are the
conditions agreed upon:
Greetings from Gnaeus Manilius son of Gnaeus Consul of the Romans,
president of the ten envoys to the council and the people of Heraclea.
Your envoys, Dias, Dies, Dionysius, Anaximander, Menedemus,
Moschus, Aristides and Menes, honorable men have appeared before
us, presented your decree and in person have discussed favourably the
matters set forth in your decree, omitting no respectful sentiment.
We are well disposed to all the Greeks, and since you have come over to
our allegiance (pistis), we shall try to take due thought for your interests, since we are always contrivers of some benefit. We concede your
freedom (eleutheria) to you, as also to other cities that have conceded a
protectorate (epitropia) over themselves to us, and are allowed to keep
all their possessions under their own control; we allow you to govern
yourselves by your own laws (kata tous hemeterous nomous) and in
other respects we shall try to benefit you and to further your interests at
all times. We accept from you your gifts and your pledges (philanthropa
kai pisteis) and we shall try not to be surpassed in requiting favours. We
have sent to you Lucius Orbius to be curator of your city and your territory, that no one may molest you. Farewell.220
There are of course significant differences between this letter and the
Muslim agreements. This is a unilateral statement rather than a signed
bilateral agreement. There is no mention of a regular tribute to be paid
44
(like the jizya); although it is highly unlikely that the city was exempt from
paying taxes, the exemption is not mentioned explicitly. Nonetheless,
despite these differences, several central elements of this letter are found
in the ama4n agreements. Like in many of the latter, the Roman commitment in the letter is given in answer to a formal embassy which presented
a decree of the city council and discussed its requests with the representatives of the conquerors. The decree states that the citizens of Heraclea
asked the Romans for pistis or fides, buttressing this requst with gifts. On
their part, the new rulers, wishing to be benevolent, gave the inhabitants
assurance of their (supposed) freedom,221 possessions, and the right to
govern their city according to their own rules, just as the Muslims would
in the kita4b ama4n. The Roman conquerors also stated their obligation,
just as the Muslims were to, to defend their new protgs. Unlike the
Muslims, the Romans appointed a Roman curator to the city (all in the
name of freedom).
Despite the significant gap in time between the two, the similarity in
form and content between this letter and the ama4n documents is quite
impressive. However, while this and many other similar documents222
from the second century BCE were luckily preserved, I know of no such
formal documents from the later Roman empire that have survived.
Nevertheless, enough evidence from the chronicles has been adduced
above223 to demonstrate that the same mechanism continued to be in use.
As we have seen, the same terminology, including the deditio, pistis/fides,
conditiones, pactum/sponsio, homologia, pax, foedus, restitutio, and
their equivalents or loan-translations in Greek and in Syriac continued to
be employed, and there are numerous examples of negotiations leading to
terms and agreements that follow the Roman model. Embassies exchange
letters, the terms continue to be set down in writing in documents and
letters, and the agreements are sealed by dire oaths.
In the case of the treaties with the Sasanians we have information
regarding the terms themselves, and in several cases the stipulations of
the treaty are described by the sources in detail. Regarding surrendering
cities, however, the sources are not as generous; thus, while we can attest
to the continuity of the model, the sources do not supply detailed descriptions of the content of the agreements. This is not surprising since a survey of the written sources in the Republican period provides, as the later
sources do, myriad reports concerning agreements, but does not adduce
the texts, or the full content of the agreements, as the inscriptions do.224
This said, the main elements of the agreements, including the security
of the civilians lives, their property, and their right to continue living
45
according to their ancestral laws without interference, as well as payment

in some form or other, are reflected in the reports.
This may be buttressed by a somewhat different, yet relevant, example
also concerning the protection of community rights which is recorded in
an appendix to the ByzantinePersian treaty of 562, cited by Menander
Protector. The document concerns the rights of the Nestorian Christians
living under Persian rule. Unlike the previous cases that have been discussed until now this is not a surrender agreement, but rather a paragraph
seeking to secure the religious rights of a community already living under
Sasanian rule. It does, however, seem to echo the classic clause central
to surrender agreements, which allowed the surrendering inhabitants to
live according to their ancestral laws (Greek kata tous patrious nomous;
Syriac na4mu4sayhu4n), and therefore to go on with their lives without any
major interruption or change225:
When these matters had been agreed and ratified, they turned to a
separate consideration of the status of the Christians in Persia. It was
agreed that they could build churches and worship freely and without
hindrance sing their hymns of praise, as is our custom (kata nenomistai
hemn). Furthermore, they would not be compelled to take part in
Magian worship or against their will to pray to the Gods that the Medes
believe in. For their part, the Christians would not venture to convert
the Magians to our belief. It was also agreed that the Christians would
be permitted to bury their dead in graves, as is our custom.226
The right to worship freely according to their custom (nomos) is found

in most of the ama4n agreements (e.g. ama4n to you, and your children,
and your families, and your monasteries, and your houses of prayer, and
your religion (milla/d n) and your prayers).227 The term nomos/nomoi is
translated literally in many of the agreements as milal wa-shara4i.228 This
seems in fact to have taken the place of the former ancestral customs
(patrioi nomoi).
All in all, it seems that the similarities between the Graeco-Roman
model and the Muslim model of the surrender treaty are striking enough
to demonstrate convincingly the resemblance of the two.
The Vassal Treaty. This model acknowledges the continuity of the local
leadership under Muslim sovereignty. It is attested in the agreements in the
area of Iran, including Jurja4n, Tifl3s, Ma4h D3na4r, Ma4h Bahra4dha4n, Isfaha4n,
al-Rayy, Qu4mis, Tabarista4n wa-J3lj3la4n, A"dharbayja4n, Marwaru4dh,
Armenia, Hara4t with Ba4dgh3s and Bushanj, and Mu4qa4n. It includes all
46
Aral
Sea
Black Sea
Tiflis
Constantinople
Caspian
Sea
Ar m ani yya
Ardabil M u q
Amida
al-Ruha
Eu
ph
ra
te
Emessa s
O
xu
s
an
Jilji
is
gr
Ti
jan
Jur
Ta b a r is ta n
Q u mis
al-Rayy
Tripoli
Mah Dinar
Ba albakk
Beirut
Mah Bahradhan
Dimashq
Bihqubadh
Isfahan
Ludd
(Damascus)
al-Hira
Bayt al-Maqdis
(Jerusalem)
Mediterranean
Sea
Alexandria
lan
Adhar bayj an
Merv
Nishapur
Marwarudh
Harat
Misr
250
le
Ni
Vassal treaty
Surrender treaty
Major city
Red
Sea
Persian
Gulf
500
km
Illustration 1. Map of the locations of surrender treaties and vassal treaties (only
where the actual text of the treaty is cited)
the elements of the previous model, and contains further stipulations

imposed on the conquered inhabitants rather than upon the conquering
Muslims, giving the agreement a more reciprocal character.
These stipulations include several obligations:(1) to give sincere counsel
and military aid to the Muslims (wa-lana nas hatukum wa-dalaukum;229
230
wa-inna alayka
ala4 ada4 Alla4h; wa-lana nashukum wa-nasrukum;
nusrata al-muslim n wa-qita4l aduwwihim)231 or, in another form, to

give counsel and hospitality (wa-la yughayyar shay min dhalika ma
adaw wa-nasahu4 wa-qaraw al-muslim n)232 or wa-ala4 an yansahu4 wa-la
yaghshu4, i.e. to give advice and not conceal anything;233 (2) to be loyal
(wa-alayna al-wafa4);234 (3) to serve as guides to the Muslims (irsha4d
al-tar q; arshidu4 ibn al-sab l; hida4yat al-tar q; dala4lat al-muslim);235 (4)
to give hospitality to a Muslim in need of it (qira4 al-muslim al-muhta4j);236
and (5) to take care of the roads and bridges in their territory (an yuslihu4
jusu4runa4;237 aslihu4 al-turuq).238
Clauses regarding loyalty, military assistance, additional aid, the return
of fugitives, and the obligation to show the way feature in ancient Near
Eastern treaties.239 These elements are also common in defensive alliances
made during the Hellenistic and Roman periods, often between cities of
47
equal standing (isopoliteia), which promise to assist each other in case

of attack or war (symmachia),240 or between Rome and cities or countries that had submitted to Roman rule (foedus iniquum).241 As already
noted above, the information regarding the Byzantine period is not as
detailed, since no copies or versions of treaties have come down to us.
Yet the sources indicate that this tradition continued to exist throughout
the Byzantine period, and that issues of military assistance and loyalty
played a central role in treaties with the Goths and the Huns,242 as well
as with the Arabs.243
However, the fact that the Muslims used the vassal treaty specifically
in regions that had previously been part of the Sasanian empire and, as
far as can be detected, not elsewhere may indicate that the direct model
was a type of treaty used by the Sasanian shahs in their dealings with
their feudal lords.244 Thus, this type of treaty may have represented an
Iranian branch of the vassal-treaty model.
The obligations included in this type of treaty indicate that the relevant
party ruled a wider territory than just a city and its vicinities. The most significant feature is the existence of an army which could be of assistance to
the Muslims; in addition, the area comprised roads and bridges (plural), a
feature characteristic of a region rather than a city. Most important, however, the majority of the names included in this list are of regions rather
than cities, i.e. Jurja4n, Ma4h D3na4r, Ma4h Bahra4dha4n, Qu4mis, Tabarista4n
wa-J3lj3la4n, A"dharbayja4n, Armenia, and Mu4qa4n, while the names of the
cities mentioned (Tifl3s, Isfaha4n, al-Rayy and Marwaru4dh) may well be
the capitals of their regions.
The existence of Sasanian signed documents which constituted contracts of investiture has been convincingly shown by Widengren,245 who
has noted that signed documents of investiture given to the feudal rulers
are mentioned in Firdaus3s Sha4h-na4ma in several places: thus, Rustam
speaks to the shah on behalf of Aula4d, asking that he be given the land
of Ma4zandara4n: First he [Aula4d] should be awarded the Kings robe of
honour, and given an edict and a seal (ahd-u muhr).246
A few lines later it is Rustam who is awarded such a written document
by the shah:The hand of a scribe wrote an edict for him [i.e. Rustam]
on silk, with musk, wine and aloe.247 In this edict he was awarded the
throne of N3mru4z. A series of similar investitures, all awarded through
such documents, is described later on in the Sha4h-na4ma, when KayKhusro, on his death-bed, grants decrees of territorial possessions to
Rustam: Gaudarz, G3v, Tu4s. Here the words manshu4r (letters patent)
and ahd (pact) seem to be used interchangeably, and each ahd is sealed
48
with a seal (muhr) of gold.248 Such documented investitures are also

recorded by the Armenian historians Moses of Chorene and Thomas
Artsruni.249
The central position of the feudal lords in the Sasanian empire stands
out clearly in Sha4pu4r Is (r. 240/370/3)250 inscriptions of Hajjia4ba4d and
Kaba-yi-Zardusht,251 as well as in the Pa4iku4l3 inscription from the end
of the third century (2934) where alongside the shah appear always the
feudal chiefs, the Great Ones, the Nobles, and the Princes.252 It has indeed
been shown recently by Parvaneh Pourshariati that the noble Parthian
families continued to carry significant weight, and that the Sasanian family was in fact dependent on alliances with these noble families throughout its rule.253
A series of events adduced by the Armenian historian Sebeos demonstrates well the relationship of Armenia, which had traditionally been such
a vassal state of the Iranians, with the shah. Pru
3 z4 (r. 45984) had persecuted the Armenians; they rebelled against him, and a great battle ensued
in which Pru
3 z4 himself was killed. What followed is described by Sebeos:254
He (Kawat)255 also made a treaty with the Armenians, summoned
Vahan256 to court, and greatly honoured him. He bestowed on him
the office of marzuba4n of the country and the principality of the
Mamikoneans. He received an oath of full submission, and dispatched
him peaceably to his own country.
Vahan, an Armenian noble, took an oath of submission to the shah, and

received a vassal treaty appointing him marzuba4n of Armenia. When
the office of marzuba4n was later taken away from the Armenians and
given to the Persians they rebelled again, turn[ed] their allegiance to the
Greeks, and the Byzantine emperor Justin II (r. 56578 CE) made an
oath with the Armenians and confirmed the same pact which had been
made between the two kings the blessed Tradat and Constantine.257
As shown above, these appointments were conferred through written
documents.
The political allegiance of the vassal state is thus established through
such treaties. Nothing is known regarding the contents of these documents. It is quite plausible, however, that they contained such obligations
as those included in the vassal-treaty model, i.e. loyalty, military assistance, counsel, and hospitality.
It is most probable that, when conquered by the Muslims, the feudal
lords who were accustomed to this system offered the Muslims just the
same kind of agreement. This fits in well with our knowledge that the
49
territories whose agreements were shaped after the model of the vassal
treaty were indeed conquered by the Muslims, but not ruled and administered by them; rather, the local leadership continued to run them as before,
only this time under Muslim jurisdiction. To cite Pourshariati:And thus,
at the expense of the Sasanians, one after another, the Parthian dynastic
families of the ku4st-i-khwara4sa4n and ku4st-i-a4durba4daga4n made peace with
the conquering Arab armies Their motive: retaining defacto control
over their territories.258
The model used in the agreements with the Muslims was not completely uniform, as the demand to supply military aid does not appear
in all cases. Most probably there were different models of vassal treaties
(e.g. alliance vs. non-aggression treaty), or we may suppose that in some
cases the model had to be altered. Thus, in some of the agreements there
is a demand for counsel only, although sincere counsel or advice seem,
according to the phrasing, to be relevant to military issues as well (see
above). In addition, there seem to have been some situations in which
the Muslims were willing to receive military aid, as in Armenia, where
the inhabitants were obliged to aid in raids (an yunfiru4 li-kul gha4ra),259
or Marwaru4dh, where the marzuba4n was obliged together with the
cavalry with him, to aid the Muslims and fight their enemies.260 The
term used here for cavalry, asa4wira4, is actually an Arabicized form of
the Persian asavara4n, the cavalry that constituted the backbone of the
Sasanian army.261 In some cases military assistance substituted for the
payment of tax. Thus, those people in Jurja4n who took part in its protection were exempt from paying taxes.262 The people of Darband actually
asked for their tribute to be military assistance, and in return that the
Arabs should not humiliate [them] with the [payment of] jizya.263 In
many cases, emphasis was put on other sorts of assistance such as guiding
the Muslim soldiers, according hospitality to Muslims (probably Muslim
warriors), and keeping up the roads and bridges, also vital to Muslim
control of the territory. On the other hand, in the case of Khura4sa4n and
Tabarista4n the agreement was that they would not be obliged or render
help or assistance against anyone.264
It is interesting to note that elements from this model serve as the basis
of the treaty between Abu4 Ubayda and the people of al-Sha4m adduced
by Abu4 Yu4suf,265 citing Makhu4l al-Sha4m3:it stipulates that the inhabitants shall be obliged to guide anyone who lost his way, to build bridges
over the rivers out of their own funds, to accommodate any Muslim who
passes by for three days to light the fires for those following Allahs
Way (sab l Allah
4 ), and not to reveal the weak points of the Muslims.266
50
It thus seems probable that the type of treaty employed by the Sasanids
for appointing their feudal lords was used as the basis for the agreements
made by the Muslims with rulers of the different regions of Iran. The latter continued indeed to be run very much as before. If this is indeed the
case, then again, the system adopted by the conquerors had its origins in
the world of the conquered.
Detailed Agreements
As already mentioned,267 most scholars claim that the minute details
appearing in the agreements make them suspicious: in fact, the more
detailed the agreement, the less likely they deem it to be authentic and
trustworthy. Yet, as noted above, long and detailed agreements were quite
common in the pre-Islamic Near East. The treaties between Rome and
Carthage, for example, included a definition of a maritime boundary
between the two, unless driven by storm or by their enemies. If anyone
is compelled to land, he shall not be permitted to buy anything or to
take anything except such as is necessary for the ships supplies or for
sacrifice; and within five days he shall depart.268 Another example is the
comprehensive agreement signed between Rome and Antiochus III:the
territories he was obliged to evacuate are defined, and it is specifically
stated that he could take his weapons only, and nothing else. A time limit
was set for all those who wanted to return to the liberated territories (just
as would be done in the Alexandria and the Jerusalem agreements, as will
be shown below). The size and equipment of his army was limited and
defined, and a large fine was set upon him.269
This continued during the Byzantine period. To mention just a few
examples, the treaty with Attila, mentioned above, was a very detailed
one;270 the treaty made between Rome and Persia in 298 included a minute description of the territories to be handed over to the Romans, a
detailed definition of the boundary line, and the fixing of Nisibis as a
sole trading point.271 The treaty made between the Byzantines and the
Persians in 562 is also long and detailed, and includes among others the
following stipulations: the Persians are to prevent the Barbarians from
entering the Roman empire; the agreement will include the Arab allies on
both sides; merchants will conduct their business through specified custom points; ambassadors will travel freely back and forth; all Barbarian
merchants shall travel through Nisibis and Daras with special permission; deserters and fugitives will be returned, etc.272
It should not surprise us, therefore, that such comprehensive agreements were drawn up following the Arab conquest, especially in the case
51
of important cities. These listed precisely the stipulations presented both

by the inhabitants of the city and by the Muslim conquerors, depending
on the specific conditions in each case.
The Treaty of Alexandria. John of Nikiu describes the capitulation of
Cyrus, bishop of Alexandria to the Muslims on 8 November 641, and
enumerates the conditions of the agreement they reached.273 Although
he does not actually cite the treaty, he does adduce its contents in detail.
The treaty of Alexandria includes specific matters and arrangements that
were made:
And they fixed the amount of tribute to be paid. And as for the
Ishmaelites, they were not to intervene in any matter, but were to keep
to themselves for eleven months. The Roman troops in Alexandria were
to carry off their possessions and their treasures and proceed (home) by
sea, and no other Roman army was to return. But those who wished to
journey by land were to pay a monthly (?) tribute. 19. And the Moslem
were to take as hostages one hundred and fifty soldiers and fifty civilians and make peace 20. And the Romans were to cease warring against
the Moslem, and the Moslem were to desist from seizing Christian
Churches, and the latter were not to intermeddle with any concerns of
the Christians. 21. And the Jews were to be permitted to remain in the
city of Alexandria.
None of these terms are in fact surprising or incredible when compared

to stipulations found in earlier agreements. Moreover, in this case, as
well as in the following ones, evidence supporting the authenticity of
these detailed agreements may also be found within the documents
themselves.
In the treaty of Alexandria several of the stipulations had no significance except at the moment of the signing of the agreement:this is true
of the arrangements concerning the departure of the Byzantine army; the
commitment of the Muslims not to intervene in any matter for eleven
months; the special clause stipulating that the Jews of Alexandria might
remain in the city; and the clause referring to the hostages taken by the
Muslims as a guarantee for the implementation of the agreement. This
last clause has parallels in pre-Islamic treaties following wars, as in the
case of Aetolia in 189 BCE or Antiochus III in 188 BCE, in the Roman
Persian treaty in 363, or the Goths in 360.274
Similarly, stipulations mentioned by John of Nikiu in the agreement
made with Babylon/Memphis (Manf) include the promise that they
should not be put to the sword, that they undertake to deliver up to him
all the munitions of war, which were considerable, and the evacuation
52
of the citadel with permission to take with them only a small quantity
of gold.275 All of these are relevant only for a short period and have no
long-term implications.
The Treaty of Jerusalem. The same can be said for the surrender agreement made with the inhabitants of Jerusalem (Il3ya4), adduced by
al-Tabar3, citing Sayf b. Umar, from Kha4lid and Uba4da.276 The agreement
includes, besides the usual clauses, specific provisions such as an obligation to expel the Byzantine army and the brigands, and an undertaking
by the Muslims to vouch for their safety until they reach their haven; the
possibility for the soldiers to remain as inhabitants in the city if they so
wish; and permission for any inhabitant who so wishes to leave with the
Byzantine army. The text goes here into precise detail, stating that those
among Il3ya4s inhabitants who wished to take their possessions and leave
the city alongside the Byzantines, thus abandoning both their churches
and crosses,277 (they, and their churches and crosses) will come to no
harm until such a time as they reach a place of safety. Another clause
states that the peasants who were in city before the murder of a certain
person (fu4la4n) may remain in the city as inhabitants. The taxes, it is specified, shall not be levied before the harvest.
Just as in the previous cases, here too we find clauses that have no
future implications, and are, in fact, practical arrangements whose purpose is to provide the necessary means for a peaceful transfer of power
and for the maintenance of stable rule in the city. Similar clauses pertaining to the opportunity given by the Muslims to leave with the Byzantines
appear, for example, in the cases of Fihl and Tiberias.278
Especially disputed in the Jerusalem agreement is the clause prohibiting the Jews from living in Jerusalem. Samuel D. Goitein claimed that not
just this clause, but the whole version of the agreement, was a forgery.279
There is, in fact, no reason to believe that this clause is untrustworthy.280
The absence of Jews from Jerusalem, their former capital and the place of
their venerated Temple, was considered by the Christians an undeniable
proof of the victory of Christianity over Judaism. The possible presence
of Jews in Jerusalem was therefore a great threat from the Christians
viewpoint. Given that prohibiting the residence of Jews in Jerusalem was
a key element in Christian policy, it is more than possible that the representatives of the Patriarchate of Jerusalem, headed most probably by the
Patriarch himself, demanded that the Muslims include such a clause in
the surrender agreement.
The fate of non-Christian inhabitants in Christian cities was also
raised in other agreements. As cited above, in the agreement concluded
53
in Alexandria there was a clause stating specifically that the Jews were to
be permitted to remain in the city. This was no doubt due to the fact that
the negotiations on the side of the surrendering party were handled by
Cyrus, the Chalcedonian Patriarch of Alexandria. It needed to be stated
clearly that whatever was concluded was applicable to the non-Christian
population of the city as well. A similar clause is to be found in the agreement of Dab3l (Dw3n), where the majority of the inhabitants were also
Christians. The ama4n given to Dab3l opens thus:This is an agreement
(kita4b) given by Hab3b b. Maslama to the Christian inhabitants of Dab3l,
to its Magian [inhabitants], and to its Jewish [inhabitants].281
In both Alexandria and Dab3l the agreements preserve the status quo
ante. This is in fact just the case in Jerusalem, where Jews had not lived
since the Bar-Kokhba revolt, which had been quelled in 136. The existence of a clause regarding the Jews in the agreement of Jerusalem is thus
no exception; in fact, quite the opposite. It preserves the status quo ante
which is firmly in the interests of the Christians. The actual presence of
Jews in the city, just a short while after the conquest, is most probably
a result of further negotiations made some time after the signing of the
agreement.282
The Treaty of Marwaru4dh. Another case in point, supporting the authenticity of the detailed agreements, is the Marwaru4dh agreement.283 The
marzuba4n of the city asks for a ratification of the status quo ante, thus
preserving the special status and rights that had been conceded by Khusro
to his great-grandfather when he killed the serpent that was feeding on
the people and cutting the roads that connected the lands and villages
along with their inhabitants.284 The myth of the great-grandfather and
the serpent here corroborated the rights of the marzuba4n, basing them
on ancestral privileges. These ancestral privileges could only have been
relevant at the time of the conquest, when they were still viable and could
be confirmed by the local population, who were apparently familiar with
the story. The manner in which the marzuba4n had obtained his status
was of no interest to the Muslims, and certainly did not in itself create
an obligation on the Muslim end. What did the Muslims care if his greatgrandfather had killed a serpent (real or metaphorical) a hundred years
before? It is therefore a local inner code, relevant only at the time of the
conquest itself to the marzuba4n and to the inhabitants of that region, thus
supporting the authenticity of the document.
The Treaty of Najra4n. Another long and detailed agreement is that of
Najra4n, cited by al-Bala4dhur3 as transmitted by Yahya4 b. A"dam.285 This
agreement differs greatly in wording and style from the agreements
54
made outside the Arabian peninsula after the Prophets death. Its special phraseology, which does not yet follow the accepted vocabulary of
the agreements, supports its early date and therefore its authenticity.
It does looks as if the original wording of this long and detailed document has been preserved to a large degree in al-Bala4dhur3s text.286 It
uses the word jiwa4r, which was abandoned in the later agreements, as
seen already.287 Other expressions used are: la4 yughayyaru haqq min
huqu4qihim wa-amthilatihim rather than the milal wa-shara4 used later,
or la4 yuftanu uskuf min uskufiyyatihi wa-la4 rahib min ruhbaniyyatihi
wa-la4 wa4qih min waqa4hiyyatihi288 rather than ama4n ala4 kana4isihim, or
biyaihim, wa-sulba4nihim/sulubihim,289 or wa-biyaihim wa-salawa4tihim
wa-d nihim.290 The issues are similar, yet it seems that the phraseology
that later became quite distinct had not yet developed.
Contentwise, this agreement seems to be a distinguished sample of the
detailed agreements. The content of the agreement displays clearly that
there were sensitive issues raised here, and that a meaningful process
of negotiations took place. The Muslim demands are quite simple, and
include a lavish payment in goods, which are listed precisely; a demand
for hospitality; and the provision of military equipment in the war
against al-Yaman. A sensitive point is the prohibition on taking interest
on loans.
The Christians insist here on their religious and civil autonomy; the
protection of all their property; their total religious freedom and the
promise not to coerce them into converting to Islam; their exemption
from military service; and the guarantee that no army should set foot on
their land. An especially interesting clause is one that exonerates them
from any responsibility for crimes committed before the rise of Islam. It
is evident that this was a disparity treaty, in which the Najra4nites were
willing to acknowledge the sovereignty of the Muslims, and to pay them
tribute, pending the acceptance of their stipulations, while the Muslims
were willing to accept the latter on condition that their own stipulations
(beside the tribute) were accepted as well.291
Here again, there are extremely minute details that would not have
been drawn up unless they were actually relevant, such as the number
of robes that the Najra4nites agreed to transfer to the Muslims; the fact
that they were to be handed over in two stages, a thousand in Safar and
a thousand in Rajab; that each should have the value of one auqiya; and
that if the price of a robe was more or less than one auqiya this would be
taken into account. Other listed goods could be given in place of robes.
The agreement lists specifically the items that the Najra4nites must supply
55
to the Muslims in case of a war with al-Yaman, including thirty coats of

mail, thirty mares, and thirty camels. These the Muslims were committed
to return, and to compensate for any losses. It thus seems that although
the terminology was not yet set at the time, the concept of such treaties
was well known to both parties.
The treaty of Najra4n is therefore a prominent example demonstrating that disparity treaties were recorded and were well known to the
inhabitants of the Arabian peninsula, and that they could indeed be very
detailed. This treaty indicates, however, that the terminology that would
later be adopted in the treaties following the conquest was not accepted
during the Prophets lifetime.
The Nubian Baqt. As already noted,292 the Nubian baqt has been dismissed by many as completely untrustworthy. This detailed contract,
reported by Ibn Abd al-Hakam and cited by al-Maqr3z3,293 was drawn up,
according to the latter, in 6512. This treaty is echoed, as noted above,
in the letter sent by the Abba4sid governor of Egypt to the king of Nubia
and Muqurra in 141/758, whose existence testifies not only that such
a treaty indeed existed during Umayyad times, but that its stipulations
were indeed quite similar to those reported by the historians.
The letter itself is replete with technical details relating to the breaking of the treaty, and concerns matters relevant to the specific time of
writing. It raises complaints regarding a certain tradesman by the name
of Sad who had run away with someone elses fortune and had not
been extradited, as agreed; it goes on to complain at length about the
mistreatment of one Muhammad b. Zayd, who had sent a merchant of
his to Nubia. Not only was this merchant arrested by the Nubians, but
Muhammad himself was sent for. The latter arrived in Nubia, was himself badly beaten, and he and an accompanying group of Muslims were
detained. All of this is adduced as evidence of the Nubians breaking of
the compact by hosting fugitives (you are to return any Muslim engaged
in hostilities against the other Muslims),294 not guarding the safety of
merchants, and detaining envoys and other Muslims (you are to look
after the safety of any Muslim or ally [of the Muslims] who lodges in
your territories or travels in them, until he departs from you).295 In addition, the governor complains about the Nubians neglect of the baqt i.e.
the delivery of Nubian slaves, who were to be supplied yearly according
to the compact, and about the fact that those who had been sent included
the one-eyed, the lame, weak old men or young boys296 (Each year you
are to deliver 360 slaves which you will pay to the Ima4m of the Muslims
56
from the finest slaves in your country and in whom there is no defect
Among them is to be no decrepit old man or woman or any child who
has not reached puberty).297
The close correlation between the letter and al-Maqr3z3s text is quite
impressive, especially given al-Maqr3z3s late date (d. 845/1442), and provides strong evidence as to the existence of a detailed treaty with the
Nubians from early times.
The clause regarding the payment in slaves was the cause of heated
discussion from the second century of Islam among Muslim jurists, who
believed that there could be no slave-trade with those who had an ahd.298
This discussion, in fact, supports the authenticity of the clause, as already
noted by Forand and by al-Qa4d3.299
In sum, detailed agreements, should not therefore be considered
suspicious and untrustworthy simply for being detailed, for two main
reasons:
1. Such agreements containing long series of clauses were prevalent
in the Near East from early times onwards. It is quite clear that the
inhabitants of the conquered cities and territories were familiar
with such documents, and demanded to be given such assurances
at the time of capitulation.
2. Many of the clauses of the detailed agreements are bona fide
clauses, reflecting specific issues raised at the time of the conquest,
some of them having no relevance whatsoever just a short time
after the agreement had been signed and implemented. There is
no reason therefore to believe that they were invented at a later
date.300 On the contrary, the detailed agreements strengthen rather
than weaken the case for the authenticity of the agreements in
general.
Payments and Gifts Accompanying Surrender
Another feature appearing in agreements with Muslims, which is typical of the traditional agreements, is the custom of handing over to the
conquerors a large sum of money or gifts.301 Thus the capitulation of
the inhabitants of Alexandria was accompanied by a generous payment
of gold, in addition to the tribute.302 In the case of Jurja4n and Qa4l3qala4
gifts sent to the Muslims are mentioned.303 Apparently, in some cases, the
surrendering inhabitants continued their long-standing custom of sending, along with their envoys, large sums of money or gifts to appease the
conquerors.
57
conclusion
Many scholars have suspected that the agreements made by the Muslims
with various cities, regions, and groups were forged, fully or partially, due
to their unexpected sophistication and to the late date of the Muslim
sources that report them. Using non-Muslim sources from the world of
the conquered populations, I have attempted to support the claim thatthe
versions of the surrender agreements cited by Muslim authors from the
second half of the eighth century onwards may well be based on authentic documents preserved both by the Muslim conquerors and by the local
conquered populations.
The claim presented in this chapter is that the formal agreements concluded between the Muslims and the conquered populations in Syria and
Palestine, Mesopotamia, and Iran were an additional link in a long tradition of such agreements common throughout the pre-Islamic world. The
Arabs were in fact familiar with this tradition, and had already in previous centuries signed foedera with the Romans. In addition, the Prophet
had before the conquest signed individual surrender agreements with
cities in the Arabian peninsula. However, with the advancement of the
conquest, the ancient Graeco-Roman tradition was brought to the fore
by the conquered populations, dictating the character of the agreements
made between the Muslims and the inhabitants of the cities. The negotiations, the format of the agreement, the terminology, and the nature
of the stipulations contained in the Muslim agreements all reflect this
long-standing tradition, while the famous term ama4n seems to have been
the Arabic form of the basic tenet of all agreements in the pre-Islamic
East:the pistis/fides. There is therefore no need to suspect these reported
agreements; rather, there is good reason to believe that they reflect the
original agreements made by the Muslim conquerors with the local populations, both sides being well aware that this was the accepted mechanism
employed on the occasion of surrender to a conquering army.
We cannot, of course, ignore the fact that the versions of the agreements handed down to us by Muslim historians may, at times, be inaccurate, and at others, distorted and tampered with in order to serve
certain purposes. Yet the existence of such documents, drawn up
between the Muslim conquerors and the conquered populations of the
East, cannot be doubted. The inhabitants of the conquered area knew
well from their own centuries-long experience that drawing up such
documents at the time of surrender was the only means to assure their
safety and well-being.
2
Shuru4t Umar and Its Alternatives
The Legal Debate over the Status
of the Dhimm 3s
The signing of surrender agreements allowed many of the inhabitants

of the newly conquered territories to go on with their lives as they had
before, even if this may have involved at times a certain measure of interruption or change, such as the Muslim occupation of houses deserted by
Byzantines or the erection of a Muslim house on a plot allotted especially
for this purpose.1 The signed agreements were considered valid and binding by both the conquerors and the conquered.
This arrangement seems to have been sufficient for the initial period, a
period in which the Muslims were occupied with ongoing conquests, trying to adjust to their new role as conquerors, and familiarizing themselves
with the conquered territories and their various cultures and languages.
Following this initial period, which may have lasted several decades or
more in certain areas, a new process commenced. The Muslims, who
were at first mostly conquerors on the move, were now settling down;
although in some cases they established their own settlements and cities,
the most prominent of which were the amsa4r such as Ku4fa or Basra, they
often settled in existing towns and cities, thus creating close proximity
between the Muslim conquerors and the non-Muslim conquered inhabitants. Moreover, non-Muslims soon settled even in the newly founded
Muslim settlements, bringing about a situation in which Muslims and
non-Muslims were coexisting side by side.
An initial version of this chapter appeared in my article Shuru4t Umar and its
Alternatives:The Legal Debate on the Status of the Dhimm3s, Jerusalem Studies in Arabic
and Islam, 30 (2005), pp. 170206. See appendices for translation of Shuru4t Umar and
al-Sha4fi3s pact.
58
59
This situation was not accounted for in the surrender agreements, and
there was a growing need to work out and define these newly created
encounters, which increasingly permeated the daily lives of both Muslims
and non-Muslims. The state of affairs was especially threatening to the
Muslims, who had previously agreed not to interfere with the lives of
the conquered, and now found themselves in insupportable situations
where pigs and wine were sold in the open markets, Christians held loud
and colourful liturgical processions, in which a processional cross with
the figure of Jesus, as well as embroidered banners and icons of saints,
were displayed throughout the main streets, and non-Muslims held high
government positions and exercised formal authority over Muslims. This
situation apparently became insupportable for the now-settled conquerors, who could not tolerate this situation where they could not check or
restrict the behaviour of non-Muslims or prevent them from filling the
public space with customs, ceremonies, and practices that often offended
the Muslim inhabitants.
This chapter aims to follow the process by which the old arrangements documented in the surrender agreements were pushed to the background while new ones were slowly being discussed and introduced by
the Muslims. This slow process ended with the acceptance of the wellknown document Shuru4t Umar, which eventually became the canonical
text summing up the principles and rules applied by the Muslims to the
dhimm s (non-Muslim inhabitants) under their rule.
The text of the Shuru4t exists in various versions. It is not my purpose
in this book to compare these or to follow the changes through time.
Meticulous research covering these issues has already been undertaken
by Daniel Miller.2
I will list here the best-known versions as well as the earliest
ones. Antoine Fattal used the most famous version, which appears in
al-Turtu4sh3s Sira4j al-mulu4k from the beginning of the twelfth century.3
Ibn Asa4kir4 in Tar kh mad nat dimashq (twelfth century) adduces five
different versions of the document with minor variations. Ibn Qayyim alJawziyya (first half of the fourteenth century), in Ahka4m ahl al-dhimma,
cites two versions with further minor variations.5 He devotes a very
large chapter to the interpretation of this document, saying that it was
so famous that he did not need to give its isna4d, that it was accepted
by ima4ms, inscribed in their books, and referred to by them; the text of
the Shuru4t was continuously on their tongues and in their books, and
the caliphs executed it and acted according to it.6 However, the Shuru4t
appears in much earlier texts. Several different versions, with some
60
interesting variations, appear in a text that was published several years

ago by Mark Cohen entitled Juz f h shuru4t nasa4ra4. The work is attributed in the text to Abu4 Muhammad Abd Alla4h b. Ahmad b. Zabr al-Qa4d3
(d. 940), and he appears in one of the versions as the earliest transmitter.7
This text is an important addition to the few early known versions of the
Shuru4t. Another early version appears in Abu4 Bakr al-Khalla4ls (d. 923)
collection of responsa by Ibn Hanbal.8 Cohen also mentions an early
version attributed to Ibn Hibba4n (884965) whose work is cited by later
authors.9 Ahmad b. al-Husayn al-Bayhaq3 (9941066) cites the Shuru4t
in his al-Sunan al-kubra4.10 Another comparatively early version of the
Shuru4t appears in Fada4il bayt al-maqdis written by Abu4 al-Maa4l3 alMusharraf b. al-Murajja4 (d. c. 103040).11 These early versions provide
us with a terminus ante quem for the Shuru4t, indicating that by about
900 the text was well known, and that there were already several versions
of it that were going around.
At a certain point the text of the Shuru4t attained the position of a
canonical document regarding the status of the dhimm s under Muslim
rule. In this chapter I will attempt to elucidate how this came about, who
were the promoters of the text, and what position it represented. I will
try to demonstrate that in the initial stages of the process this text did not
reflect a consensus concerning the status of the dhimm s but in fact represented only one of several existing approaches to the issues in question.
the dating and the formation of
SHURU 4 T
UMAR
When was the text of the Shuru4t compiled and what was its Sitz im
Leben? The reasons for rejecting the attribution of this treaty to Umar
b. al-Khatta4b have been extensively expounded by Tritton and Fattal,
and need not be repeated here. It can hardly be doubted that the treaty
in the form we know it belongs to a period much later than the conquest.12 Noths suggestion to the contrary will be considered later in
this chapter.
Tritton, and Fattal in his footsteps, believed that it would seem that it
was an exercise in the schools of law to draw up pattern treaties.13 Fattal
assumed that it was the work of mujtahids of the third century AH [i.e.
the ninth century], who could not resist the temptation to assemble in
one document the successsive restrictions on the liberties of the dhimm s,
without taking into account circumstances of time and place.14
I will attempt to show that the eighth and ninth centuries appear to
have been a period in which the regulations concerning the dhimm s were
61
the subject of a lively debate. If these treaties were a subject of study in

the law schools, this was not because the jurists needed to exercise how
to draw up treaties, as Tritton suggests, or simply to collect successive
regulations in one document, as Fattal assumes, but rather because this
was a burning social and religious issue in the caliphate and therefore
also among Muslim jurisprudents.
It may well be that Shuru4t Umar existed even earlier than the ninth
century, if the isna4ds that appear in the different versions are to be trusted.
Ibn Asa4kir, for example, cites three transmitters in three of his versions
of Shuru4t Umar, presented as the sulh agreement with the people of
al-Sha4m:15 Sufya4n al-Thawr3, al-Wal3d b. Nu4h, and Sar3 b. Musarraf.16 Of
the three, Sufya4n al-Thawr3 is the best-known figure; he lived between 97
and 161 H (71678 CE) in Ku4fa, Khu4rasa4n, the Hija4z, and Basra, visited
al-Sha4m, and belonged to the first group of scholars in Ku4fa who arranged
the Had3th.17 All three received the Had3th from Talha b. Musarraf (d.
c. 11213/7301).18 The isna4d of the Shuru4t in Ibn al-Murajja4s Fada4il
bayt al-maqdis wa-al-khal l wa-fada4il al-sha4m goes back to Muhammad
b. Kab al-Quraz3 (d. 10819/72637).
If we date the text according to these isna4ds it goes back to the first
half of the eighth century CE. If we are more cautious and remain within
the methodological limitations of recent isna4d research, it is dated to the
second half of the second Islamic century, i.e. the middle of the eighth
century, the time of Sufya4n al-Thawr3.19
In his article on Shuru4t Umar, Albrecht Noth argued that the text
is in fact a collection of early understandings based on treaty traditions
and on the praxis between the Muslim conquerors and the non-Muslim
conquered population. He believed that the text is composed of three
basic layers:
1. Clauses that go back to the time of the conquest. These comprise
the basic ama4n, including the obligation of the conquerors not to
cause damage to prayer-houses of the dhimm s, and the obligation
of the conquered to cooperate with their new rulers e.g. act as
guides for them in unknown territory, accommodate Muslims in
their houses for three days, and allow them to enter their churches
day and night. Noth also shows convincingly that the priniciple of
khila4f, or ghiya4r, the differentiating signs, goes back to a very early
date.20
2. Clauses that belong to the period when Muslims settled in the
newly conquered territories, and which serve the need to define the
62

practical arrangements between the two groups. To them he attributes the prohibition on the loud use of the na4qu4s (the clapper or
gong used in the East for the purpose of calling to prayer), and the
prohibition on conducting processions and exhibiting Christian
(or other) religious symbols publicly.
3. Other variations of these prohibitions that developed later, such
as the prohibitions on praying loudly in Muslim neighbourhoods,
on conducting funerals in Muslim areas, and on displaying signs
of polytheism (shirk).21 Noth thus believed that the Sitz im Leben
of the earliest material in the Shuru4t was in fact the time of the
conquest, and that it reflected the spirit of that time, when the
object was not to humiliate but to separate the Muslim minority
from the dhimm population, which constituted the majority at
the time.
Noths argument that some of the clauses reflect the conquest environment is convincing. Yet if it reflects different periods, as Noth claims, was
its crystallization a process that occurred almost naturally and unconsciously, generation after generation, without any discussion?
I would like to suggest that this was not the case; rather, the text
of the Shuru4t was drafted and formed as part of an ongoing discussion concerning the status of the dhimms which started some time
in the eighth century, most probably at the time of Umar II b. Abd
al-Az3z,22 and culminated around the turn of the eighth century and
the beginning of the ninth. If the isna4ds are reliable, the text as we
know it existed already in some form in the mid-eighth century; yet it
was not the sole voice concerning the status of the dhimms. I will try
to show that before the Shuru4t became the canonized text, there was
no single formal legal document concerning Muslimdhimm relations
which was agreed upon. Rather, there were alternative versions which
disagreed on several major issues. Tritton, who was in fact aware of the
other versions, cited them all under the title the Pact of Umar, and did
not attribute any importance to the differences between them. Fattal,
who was also familiar with other versions, noted that Abu4 Yu4sufs text
reminds one strangely of the Pact of Umar, and constitutes, in fact, an
abridged version of it, the oldest that we possess.23 I believe that rather
than constituting an abridged form of this document, these versions
represented alternative approaches concerning vital matters which seem
to have been in dispute before the Shuru4t became the established and
canonical text.
63
legal discussions throughout the eighth

and ninth centuries
Muslim tradition strongly emphasized the need to respect the individual
sulh agreements that were concluded by Muslim commanders with cities and other entities. This is repeated time and again in many sources.24
Although the need to mention this repeatedly raises the suspicion that
this was not always so in reality, the agreements seem to have been a legal
barrier, stopping the Muslims from freely modifying conditions according to their needs, as they would no doubt have liked to.
Thus, on the face of it all, the cities and settlements that had sulh treaties were living under clear agreements concluded at the time of conquest.
This, however, was not enough to regulate relations between the Muslims
and the dhimm s.
There were (a) places conquered anwatan, meaning that they did not
have an agreement; (b) amsa4r al-muslim n25 cities founded by the Arabs
after the conquest in which the dhimm s were newcomers; and (c) the
most complicated case cities that had previously been populated mainly
by dhimm s, and even had sulh agreements, but which were gradually
changing their characters to include more and more Muslims. How
would this affect the status of the dhimm s in those cities?
The last two cases were undoubtedly the product of changes and
developments that were taking place from the second half of the seventh century and throughout the eighth and ninth, when Muslims were
becoming an increasingly dominant section of the population. All these
special cases raised many legal questions. Jurists became highly involved
in the matter, presenting different opinions concerning the principles of
Muslimdhimm relations each supported, of course, by legal proofs
and traditions.
The debate revolved around the issue of Muslims and dhimm s living
in the same neighbourhoods, and its implications. The discussions concerning this question usually invoked a had th attributed to Ibn Abba4s.
Abu4 Ubayd cites the following version: In any city established by the
Arabs, the dhimm s are prohibited from building a prayer-house, wine
is not to be sold, a pig shall not be bought, and a na4qu4s shall not be
sounded. Anything that was [agreed upon] beforehand, it is the Muslims
obligation to fulfil.26 A more detailed version of this had th is cited by
Abd al-Razza4q:As for what the Muslims established, a church cannot
be raised in it, nor a synagogue, nor a cross, nor a spear-head. A horn will
not be blown in it, a na4qu4s will not be sounded, and neither wine nor
64
pig will be admitted into it. As for a place where a peace agreement was
made the Muslims are obligated to fulfil the [terms of the] sulh they
made with them.27
This had th differentiates between cities established by the Arabs, perhaps also those taken by them anwatan, in which dhimm s held no previous rights or privileges, and between cities and towns that were originally
inhabited by dhimm s and were conquered sulhan. As we have seen in the
previous chapter, the latter had treaties that usually gave them ama4n for
their persons, their property, including prayer-houses, and allowed them
to carry on with their lives as before on condition that they pay the jizya
and provide hospitality and certain other services for the Muslims.
Ibn Abba4ss had th is in fact quite emphatic concerning Muslim obligation to adhere to these ahds and fulfil their terms. It seems clear-cut,
according to this had th therefore, that only in a city established or settled by the Arabs, where there was no ahd, could the rules of the Muslim
misr be applied.
This interpretation could not, however, contend with the more complex situation that was gradually unfolding. What of a city that was conquered sulhan but was gradually being populated by Muslims? Could the
dhimm s still carry on as before? Could they keep their prayer-houses?
Could they go on practising their religious customs, especially those
involving the public sphere, such as processions, or sounding the na4qu4s?
Could they exhibit their religious symbols and icons in public? Could
they go on growing, selling, and consuming wine and pigs? All of these
actions obviously affected the citys everyday life and created its atmosphere. Could a city under Muslim rule be of Christian character? Should
the Muslims put up with public behaviour that contradicted Muslim law
and thereby offended Islam? The narrow interpretation of Ibn Abba4ss
tradition prohibits any change or intervention on the part of the Muslims;
but by the eighth century this, as shall be seen presently, was, it seems,
becoming unacceptable to them.
In fact, much depended on the interpretation of the expression misr
massarathu al-arab (a city established by the Muslims) since, if a city was
defined as such, many restrictions could be imposed upon its Christian
inhabitants. Abd al-Razza4q (126211 H) interprets it to mean whatever
was part of the land of the Arabs or taken anwatan.28 It is not completely clear what is meant here by ard al-arab; however, since what was
taken anwatan (and, by implication, sulhan) is a separate category, it is
most likely that it refers to the Arabian peninsula. In the same chapter,
however, Abd al-Razza4q cites traditions concerning the status of churches
65
in amsa4r al-muslim n. He adduces traditions supporting the actual

destruction of churches by Umar b. Abd al-Az3z in both the old and
new amsa4r:al-amsa4r al-qad ma wa-al-had tha.29 One tradition expressly
claims that according to the sunna churches should be destroyed in both
the old and new amsa4r. This expression, al-amsa4r al-qad ma wa-alhad tha, apparently refers both to cities built by Muslims and to others
that existed before the Muslim conquest, and were probably populated
in later times by Muslims.
Thus interpreted, any city that had a considerable Muslim population was considered a Muslim misr , and its dhimm population was to be
restricted accordingly. This interpretation of the had

th
clearly supported
the less tolerant views that proposed prescribing new rules diminishing the
rights of the dhimms wherever large numbers of Muslims were present.
Beside these traditions, Abd al-Razza4q adduces others which support the view that if these amsa4r had a sulh, the churches could not be
touched.30 It would thus seem that Abd al-Razza4q reflects here a situation of disagreement concerning the question whether a city protected by
an agreement (ahd) could fall into the category of amsa4r al-muslim n,
which in effect changed its status and imposed many restraints on its
dhimm inhabitants, and even according to some opinions the destruction of their churches.
Abu4 Ubayd (d. 224 H) also defines the expression misr massarathu
al-arab. According to him, there are three cases that fall into this category:(1) a place whose people accepted Islam, such as Medina, al-Ta4if,
and al-Yaman; (2) any city built by the Muslims in a place that was
previously not settled, such as Ku4fa and Basra; (3) any city conquered
anwatan and not returned by the ima4m to its inhabitants.31 He goes on
to list those cities that were conquered sulhan and those that were conquered anwatan. Those conquered sulhan include Jerusalem, Damascus,
all the cities of al-Sha4m (but not the surrounding countryside), the cities
of al-Jaz3ra, the Copts in Egypt, and the cities of Khura4sa4n.32
Abu4 Ubayd clearly states that while the dhimm s were not allowed
to exhibit any of their religious customs in amsa4r al-muslim n, in those
that were conquered sulhan the ahd is valid, and the privileges that
were accorded to the inhabitants could not be revoked or retracted.33
Here the definition of amsa4r al-muslim n is clear-cut and the terms of
cities conquered sulhan are non-negotiable, even in the case of changing
circumstances.
Al-Sha4fi3 (d. 204 H) in Kita4b al-umm defines the term misr min amsa4r
al-muslim n as a city that has been spared and conquered anwatan,34 or
66
one established in a place that did not belong to the dhimm s.35 This is
contrasted with a city that was conquered sulh
an and conditions were
agreed upon, which were to be respected and maintained. Al-Sha4fi3 ennumerates all the restrictions that apply in amsa 4r al-muslim n but emphasizes that in a city in which the dhimm s lived separately, or a city that
had a sulh agreement allowing such things as the building of churches
or the public exhibition of wine and pork, the amsa4r restrictions were
not applicable. Like Abu4 Ubayd, he draws a clear line between the two
groups.
There were other interpretations of amsa4r al-muslim n. Abu4 Ubayd
was familiar with views much less tolerant than his own. Thus, under the
title What are the ahl al-dhimma allowed to establish in ard al-anwa and
in amsa4r al-muslim n and what not? he adduces traditions that could
apply to all territories under Muslim rule such as la4 khisa4 f al-isla4m
wa-la4 kan sa (there is no castration in Islam and no church), or it is not
fitting for a house of mercy to adjoin a house of chastisement.36 The promoters of such views obviously believed that all territories under Muslim
rule should be considered amsa4r al-muslim n, and that there should be no
church where there was a mosque.
Muhammad b. al-Hasan al-Shayba4n3 (13287 H) writes in his Kita4b
al-siyar that the dhimm s are allowed to keep their prayer-houses, and
even to rebuild them in any city that was conquered sulhan, but if the
Muslims establish a city (misr) [for themselves] in that place, they should
tear down the synagogues and churches there, but the dhimm s should be
allowed to build similar ones outside the city.37
The view represented by al-Shayba4n3 is that the sulh was valid just as
long as it suited the Muslims. Once the Muslims took over a city there
was an immediate change of rules; there were no scruples about destroying prayer-houses once the Muslims had established themselves in the
city. Dhimm s were also not allowed to go on residing in cities inhabited
by Muslims, and were to be expelled just as they had been expelled from
Medina, and as Al3 b. Ab3 Ta4lib had expelled them from Ku4fa. This is
also al-Tabar3s view.38 According to al-Shayba4n3, however, they were to
be given a place of their own outside the city where they could resume
their former way of life.39
Al-Sarakhs3 (d. 483 H) in his Sharh to al-Shayba4n3s Siyar regards any
place where public prayers are held and where hudu4d punishments are
in effect as a misr.40 Al-Sarakhs3s definition follows that of al-Shayba4n3,
and encompasses, in fact, any city where a Muslim presence has become
predominant. On the other hand, he says that a city where the majority
67
of the population are dhimm s, such as al-H3ra, where no public prayers

are held and hudu4d punishments are not applicable, the restrictions are
not in effect.41 These definitions have nothing to do with the history of
the conquest; rather, they refer to the reality of the time. There are no
secured rights and privileges; there is only the matter of a Muslim majority. Once a city changed its character and the Muslim population became
predominant, no sulh was valid.
The discussion among Muslim jurists concerning the dhimm s revolves
therefore around two main questions:
1. What cities fall under the definition of amsa4r al-muslimn? This
definition was crucial since the dhimms of a city so defined
might be subjected to restrictions that would not otherwise be
allowed. A narrower definition, such as that accepted by Abu4
Ubayd and al-Sha4fi3, would obviously represent a more tolerant
approach, leaving all dhimms who had been conquered sulhan
free to live and practise their religion as promised to them in the
sulh agreements. A broader definition, on the other hand, such as
that adopted by al- Shayba4n3, al-Tabar3, and al-Sarakhs3, would
encompass many more cities, allowing the Muslims to override
and annul earlier promises made to the dhimms, in effect abrogating their uhu4d, in any place where Muslim population had
become dominant.
2. What is allowed or prohibited in a Muslim misr? The main problem seems to have been the dhimm prayer-houses. While it was
agreed that new churches could not be built, it was clearly a matter of general disagreement whether existing churches in amsa4r
al-muslim n should be destroyed or not. There are contrasting traditions on this matter. Abu4 Ubayd adduces a tradition attributed to
Umar b. Abd al-Az3z that you are not to destroy a church, a synagogue or a fire-temple, nor to build a new one. This appears alongside traditions that condone such actions. Abd al-Razza4q adduces
traditions stating that Umar was intent on destroying churches in
amsa4r al-muslim n as well as others saying that he refrained from
doing so.42
Ibn Ab3 Shayba (159235 H) cites a tradition attributed to Hasan, most
probably Hasan al-Basr3 (d. 110 H) saying that he disapproves of letting the prayer-houses (biya) exist in amsa4r al-muslimn. Al-Shayba4n3
also believed that churches in the amsa4r should be destroyed (see
above). Al-Sarakhs3, on the other hand, cites Umar b. al-Khatta4b as
68
saying that no new churches should be built, but no ancient ones were
to be destroyed.43
Some jurists even claimed that dhimms should not be allowed to
live permanently within amsa4r al-muslimn.44 A passage related to this
issue is found in an unspecified work by al-Tabar3, which is cited by
the Mamlu4k jurist Taq3 al-D3n al-Subk3 and is not found elsewhere. If
this passage is indeed authentic,45 then al-Tabar3 was also of this opinion, defining amsa4r al-muslimn as towns where the Muslim population was predominant. His concrete proof was that Al3 b. Ab3 Ta4lib
decided that dhimms could not live in Ku4fa, allowing them to live only
in al-H3ra, which was populated by dhimms, or in Zura4ra, across the
river.46 Al-Shayba4n3 claims that dhimms could buy houses and live in
Muslim amsa4r. Al-Sarakhs3 explains this, saying that they were allowed
to do so, only as long as they remained a minority, so that they could
observe the beauty of Islam.47 They could not, however, rent houses
from Muslims.48
the composition of the general S ULH documents

The presence of dhimm s in areas populated by Muslims became more
problematic as time went on. While immediately after the conquest
Muslims were few and far between, by the eighth century they were to
be found in many big centres, and in many cases probably formed a predominant part of the population. As areas of contact and friction, both
geographical and social, between Muslims and dhimm s were increasing,
it was becoming more urgent to find solutions that would be generally
acceptable and applicable. It would seem that these changing circumstances created an incentive for formulating a uniform set of regulations
applicable to all dhimm s living under Muslim rule. This uniform set of
rules had, however, to be formulated and drawn up, and, more important, to be agreed upon.
In addition to the pressing dilemmas of the times, the composition
of a general and uniform legal document regarding the non-Muslims
towards the turn of the eighth century may also have been inspired by
the creation of a novel legal system, as has been claimed recently by
Benjamin Jokisch.49 According to Jokisch, this new law code was based on
Justinians Corpus Iuris Civilis, and was designed to replace the existing
heterogeneous law.50 This ambitious venture was being led, according to
Jokisch, by Ha4ru4n al-Rash3d, and by his leading jurisprudents, Abu4 Yu4suf
and al-Shayba4n3. As we have seen, al-Shayba4n3 was deeply involved in
69
the issue of the dhimm s, while Abu4 Yu4suf, as will presently be seen, was
engaged in the formation of a general document regarding the dhimm s.51
Thus the idea of cutting the Gordian knot by creating a single, uniform
legal document rather than by making complicated circumstantial differentiations may have been a byproduct of the new codification scheme
undertaken by those jurists.52 It should also be added that all the other
figures participating in the argument over the place of the dhimm s in
Muslim society, including al-Sha4fi3, Abd al-Razza4q, and Abu4 Ubayd,
were active in that same period, and may well have been aware of the
attempt to create such a uniform document. This view is well in line with
Cohens theory regarding the petition form that was given to the Shuru4t.
Cohen suggests that the literary frame of the Pact of Umar be seen as
that of a petition to the Caliph, approved by him in the form of a decree
and furnished with directions to the Caliphs delegate to implement it for
the Christian petitioners.53 Thus, once the decision was made to codify
one set of laws pertaining to the non-Muslims, the format chosen, the
petition to the caliph, was one that was characteristic of the Muslim
administrative system of the time.
How could the aim of drawing up such a uniform document be
achieved? Certain restrictions concerning dhimms who were living in
Muslim cities, such as the prohibition on building new prayer-houses,
the sale of pigs and wine, and the rules of ghiya4r,54 seem to have already
been widely accepted at that time among the jurists. Yet there were still
many issues that were not agreed upon, such as the definition of amsa4r
al-muslimn, which in fact decided where these regulations apply, the
fate of churches within these amsa4r, and even the fate of dhimms living
inthem.
Moreover, the uhu4d or sulh agreements given to cities conquered
sulhan immediately after the conquest constituted a great obstacle. The
uhu4d were commitments that had to be kept. The obligation to observe
these treaties is straightforward and unambiguous in Ibn Abba4ss tradition, and he is followed by many of the jurists in the eighth and ninth centuries.55 Although some of the jurists, such as al-Shayba4n3 and al-Tabar3,
solved this problem by changing the definition of a Muslim misr, this
was certainly a problematic solution which cannot have been acceptable
to all, since it entailed the arbitrary annulment of prior commitments,
whenever the Muslims saw fit to do so.
What probably aggravated the situation was the fact that the original
uhu4d were generally not very favourable from the Muslim viewpoint.
They were, it would appear, extremely tolerant, mostly just demanding
70
submission and taxes or other payments in return for an ama4n, which

allowed life to go on as before without any restriction or interference.
How could the Muslims solve the problem of these treaties? Could
they actually be pushed aside by a document that would arbitrarily
annul them? An ingenious solution for this problem was found in the
form of a general sulh agreement. While all the original ones were
specific agreements given to individual cities and differing from one
another, here there would be a document that allegedly represented a
set form of sulh agreement applicable to all dhimm cities and settlements under Muslim rule. This agreement would include certain conditions regarding Muslimdhimm coexistence, which were not included
in the original agreements, and, since it would be a general agreement,
it could override the individual agreements. The contents of this general sulh agreement had, however, to be agreed upon, and this was
no simple task. I would like to show that Shuru4t Umar was, in fact,
only one approach of several that were suggested in the course of this
process.
I will treat here four different formal versions of such a general sulh.
None of these is a newly discovered document; they are all known and
cited throughout the literature, yet they are all treated indiscriminately as
Shuru4t Umar. Although there are elements common to all versions, there
are nevertheless meaningful differences between them; I suggest they each
represent a different version of a uniform sulh reflecting the different
opinions voiced in this public discussion.
Abu4 Yu4sufs Version
A significant approach was voiced towards the end of the eighth century
by Abu4 Yu4suf. In his Kita4b al-khara4j, a book which represents the formal
legal position emanating from Ha4ru4n al-Rash3ds court,56 there is a version of a general sulh agreement as well as several texts related to the subject. These offer an approach that is quite different from that of Shuru4t
Umar. Abu4 Yu4suf does not cite or mention the Shuru4t at all. Either he
did not know the text, whether because it did not exist yet or had not
gained wide currency, or he did know it, but chose not cite it because he
did not approve of it. Noth refers to Abu4 Yu4suf in a short note only, concluding that the Shuru4t existed in his days, but that he was not aware of
its existence.57 If this is true, the text must still have been marginal, given
Abu4 Yu4sufs position. Noth does not refer to the substantial differences
between the two texts.58
71
Abu4 Yu4suf presents his position on the subject of the relations between
the Muslims and ahl al-dhimma in the chapter On the churches, synagogues and crosses (F al-kana4is wa-al-biya wa-al-sulba4n).59 At the
beginning of this chapter he makes the following statement:
As for what you asked me, O Am r al-Mumin n, concerning the matter
of ahl al-dhimma, and why they were allowed to keep the synagogues
and churches in the cities and in the metropoles (amsa4r) at the time of
the Muslim conquest of the countries, and these were not destroyed, and
why they were allowed to parade the crosses on their holidays:That is
because the agreement between the Muslims and ahl al-dhimma was
based upon the payment of jizya, and the cities were conquered on condition that their synagogues and churches inside the city and outside
it would not be destroyed, and that they [i.e. the Muslims] would prevent the shedding of their [i.e. the dhimms] blood, and that they [the
60
Muslims] would fight any enemy who attacked them [the dhimms],
and that they [the dhimms] would [be allowed to] parade their crosses,61
and that they [the Muslims] would protect them.
The question underlying this text seems to be the grounds on which special privileges were given to the dhimm s, specifically their right to keep
their prayer-houses and parade their crosses. Abu4 Yu4suf states that the
present liberties of the dhimm s all emanate from the sulh agreements
at the time of the conquest. His wording (how they were allowed to )
suggests that these privileges of the dhimm s had been questioned, and so
had to be not only clearly stated, but safely anchored in some early traditions. For this purpose Abu4 Yu4suf advances a series of historical proofs
and texts. First he adduces a version of the sulh agreement made between
Abu4 Ubayda b. al-Jarra4h and the inhabitants of al-Sha4m, an agreement
which was supposed to have been a uniform sulh agreement for all of
the cities of al-Sha4m, i.e. greater Syria with its five districts including
that of Filast3n. This agreement, transmitted from Makhu4l al-Sha4m3 (d.
113/731),62 was later applied, according to Abu4 Yu4suf, to the rest of the
cities conquered as well:
Certain learned men related to me from Makhu4l al-Sha4m3 that Abu4
Ubayda b. al-Jarra4h granted a peace agreement to the people of al-Sha4m
upon entering it, on condition that their churches and prayer-houses
(kana4isuhum wa-biyauhum) would be left [standing], that they would
not build any new prayer-house or church, that they would be obliged
to guide anyone who lost his way, that they would build bridges over the
rivers out of their own funds, that they would accommodate any Muslim
who passed by for three days, that they would not defame a Muslim or
72

strike him, that they would not raise a cross amid the Muslims, they
would not take a pig out of their dwellings into the courtyards of the
Muslims, that they would light the fires for those following Allahs Way
(sabl Alla4h), that they would not show the weak points of the Muslims
(wa la4 yadullu4 lil-muslimn ala4 awra), that they would not strike their
nawa4qs before the Muslim call to prayer, or at the times of the Muslim
call, that they would not take out their banners (ra4ya4t)63 on their holidays, that they would not wear arms on their holidays nor keep them in
their houses. If they did any of these things they would be punished and
it [i.e. the ahd] will be taken from them. These were the conditions of
the agreement (sulh). But they said to Abu4 Ubayda:set a day in the year
for us upon which we can take out our crosses without banners, namely
the day of our great holiday. He did this for them and consented to this,
so that they could not avoid the fulfilment of the conditions stipulated.
The other cities were conquered according to the same [conditions].64
The version of the treaty adduced here combines elements appearing

in the uhu4d of specific cities with new elements which contend with
the problem of ChristianMuslim coexistence. It is in fact a sulh treaty
adapted to the contemporary situation in which Muslims were living in
cities conquered sulhan.
Makhu4ls version of a general sulh agreement, as cited by Abu4 Yu4suf,
presents a series of stipulations having to do with loyalty, aid, and hospitality to the Muslim rulers, and respect of Muslim religion and ritual. Yet
it accentuates the privileges of the dhimm s. Their rights to their prayerhouses and to freedom of liturgy and other practices and habits, i.e. the
right to call publicly to prayer at all times other than the adha4n, or to
keep pigs and wine anywhere other than in the immediate vicinity of
Muslims. Especially accentuated is their special privilege to hold an open
procession with crosses on Palm Sunday, a privilege granted to them,
according to this tradition, by Abu4 Ubayda himself in order to establish a
relationship of mutual honour and trust. The special emphasis put on the
latter privilege may indicate that the sanctioning of the procession with
the crosses on Palm Sunday aroused opposition among certain Muslim
circles, an opposition which will be referred to later on, and which Abu4
Yu4suf may have been trying to subdue by referring to such a significant
agreement made by Abu4 Ubayda himself.
The same version, with only minor variations, is cited by Ibn Asa4kir
in his Tar kh mad nat dimashq.65 The isna4d goes back to Isha4q b. Bishr
Abu4 Hudhayfa al-Ha4shim3, who is reported by Ibn Asa4kir to have been
active in the reign of al-Mamu4n (d. 206/821).66 Ibn Bishr was the author
of a book on futu4h and therefore must have been familiar with many
traditions of sulh agreements.
73
There are several differences in Ibn Asa4kirs version, the most important being the omission of the story concerning the special permission
given to the Christians by Abu4 Ubayda to conduct an open procession
with crosses on Palm Sunday. Also, in contrast to Abu4 Yu4sufs version, Ibn
Asa4kirs version does not cite the edict that their churches and prayerhouses should be left (standing) unharmed (an tutraka kanaisuhum
wa-biyauhum).67 We thus have here the same text of the ama4n given to
the people of al-Sha4m by Abu4 Ubayda, in two versions reported by two
different transmitters whose dates suggest that it was well known around
800 CE. Yet the seemingly minor differences between the two versions
are significant, Abu4 Yu4sufs stand representing a more tolerant and permissive position than that of the parallel version.
Abu4 Yusuf
4
later adduces two reports of sulh
agreements which support
this version of the general sulh
transmitted by Makhu
l4 al-Sham
4 .3 One is
a summary of the conditions made by Khalid
4 b. al-Wald
3 with the ruler
of al-H ra,
3
Iyas4 b. Qabs3 a.
68 Abu4 Yusuf
4
briefly reports here that Khalid
4
made a treaty with them agreeing that neither a synagogue or church of
theirs, nor a fortress of theirs in which they fortified themselves against an
enemy will be destroyed; that they will not be prevented from sounding
the nawaq
4 s , or from parading their crosses on the day of their holiday;
that they will refrain from committing offences, and that they will accommodate those Muslims who passed by them and supply them with food
and drinks permitted [to Muslims].69 He then quotes the text of the treaty
itself, but the text lacks most of the terms he has just enumerated. It makes
no reference to the status of prayer-houses, the nawaq
4 s , or parading the
cross, nor for that matter about their fortresses, but emphasizes only the
duty not to give aid to a kafi
4 r against a Muslim, and not to guide the
enemy towards the weak points of Muslim territory.70 The text cited is
significantly different from the other agreement versions in content and
in style, and raises a heavy suspicion that it was fabricated by those who
Abu4 Yusuf,
wanted to do away with the rights of the dhimms.
4
here as
elsewhere, seems to be attempting to defend dhimm rights while participating in the general enterprise of standardizing dhimm status.
The second sulh he refers to is that with the people of al-Nak3b and
al-Kawa4thil,71 who demanded the same conditions as Ana4ta4. In this
agreement, similar elements are adduced by Abu4 Yu4suf: no church or
synagogue is to be destroyed; they will be allowed to sound their nawa4q s
any time they want, day and night except during Muslim prayer times
(exactly the same stipulation as in Makhu4l al-Sha4m3s text); they will
be allowed to parade their crosses on the days of their holidays (here,
74
it seems that it is not limited only to Palm Sunday). In return, they will
have to accommodate Muslims for three days, and serve as their guides.
The same agreement exactly was given, according to Abu4 Yu4suf, to the
people of Qarq3siyya.72 Abu4 Yu4suf underlines the authenticity and validity of these agreements, drawing their veracity and authenticity from
the support of the Ra4shidu4n: These arrangements of Kha4lid were not
annulled by Abu4 Bakr, nor were they annulled after Abu4 Bakr by Umar,
by Uthma4n, or by Al3.73 He then goes on to say that in fact, when certain caliphs thought of demolishing dhimm prayer-houses they had to
contend with these sulh agreements, and with the clear opposition of the
jurists and the Successors. He closes by saying:The sulh is effective as
it was in the days of Umar b. al-Khatta4b until the Day of Resurrection
(fa al-sulh na4fidh ala4 ma anfadhahu Umar b. al-Khatta4b ila4 yawm
al-qiya4ma). The prayer-houses and the churches were left to them [i.e. the
dhimm s], as I informed you.74
Abu4 Yu4suf emphasizes that he adduces these agreements in order to
support his position on the subject of the dhimm s. He openly and repeatedly says that a tolerant approach should be adopted towards them. This
attitude had been adopted by the conquerors because they understood
that they stand to gain from this policy. When the ahl al-dhimma saw
that the Muslims fulfilled their promises to them, and treated them well,
they became adverse towards the enemy of the Muslims and helpful to
the Muslims.75 In order to substantiate this claim he tells the well-known
story, also told by both al-Bala4dhur3 and Dionysius of Tell-Mahre4,76 writing in the first half of the ninth century, about the agreement with the
people of Hims and Damascus, though he does not name a specific city.
According to this tradition the Muslims returned the jizya paid to them
by these cities on the eve of the battle of Yarmu4k because they were not
sure they would be able to protect the inhabitants of these cities. The
latter were so grateful that, according to Abu4 Yu4suf, they retorted:May
God bring you back to us and give you victory over them, for if they
were in your place, they would not give us back anything, but would
take anything we had left, leaving us without a thing.77 This, Abu4 Yu4suf
explains, was Abu4 Ubaydas strategy, in order to draw the other cities
to him. Once they saw how generous the Muslims were, they all came
of their own volition asking for a sulh, on these terms.78 Al-Bala4dhur3
and Dionysius of Tell-Mahre4 adduce a similar version of the story. It is
meant to disseminate a sense of fair play, a tolerant approach towards the
dhimm s which reaps its reward in the form of support of the new rulers.
Whether the story has any truth in it does not matter for our purposes.
75
It does reflect a spirit that was apparently prevalent in certain Muslim

circles around 800. The disseminators of this story believed, it seems, that
a tolerant attitude on the part of the Muslims was the right approach,
and that it was the only one that would gain the trust and the support
of the dhimm population necessary for the existence of a stable Muslim
state in the long run.
Abu4 Yu4suf further substantiates his view by citing in full Umars
alleged statement on the treatment of the dhimm s.79 The tolerant attitude advocated by Abu4 Yu4suf is well supported by this speech: Umar
is cited as saying to Abu4 Ubayda:You are to fulfil all the terms of the
agreement I have given them. As for the parading of the crosses on their
holidays, you shall not prevent them from doing this, without banners
(ra4ya4t) or standards (bunu4d), as they requested from you for one day each
year. Inside the city, among the Muslims and their mosques, the crosses
shall not be shown.80
The main issue tackled here is thus the terms of ChristianMuslim
cohabitation. Abu4 Yu4sufs main goal seems to have been to be a form
of coexistence that would be satisfactory to both sides. That is, that
Christians would be allowed, as far as possible, to continue leading their
lives as they had before as long as they did not openly offend the Muslims
by disrupting their adha4n, or by displaying pigs, wine, or shirk in their
neighbourhoods. This approach is openly and emphatically attributed by
Abu4 Yu4suf to Umar b. al-Khatta4b, and to the Ra4shidu4n after him. There
is no reference whatsoever to the text known as Shuru4t Umar.
Evidence of Restrictions on the Dhimm3s during
the Period Discussed
As noted above,81 Abu4 Yu4suf mentions that attempts made by certain
caliphs to take over or demolish dhimm prayer-houses were prevented
by jurists who saw themselves bound by the sulh agreements. Abu4 Yu4suf
was in fact contending with problems that were actual and relevant to
his time.
There is evidence of attempts at destruction, and of actual destruction,
of churches by caliphs and other Muslim rulers in the later Umayyad and
early Abba4sid periods. Abd al-Razza4q cites a tradition according to which
Umar b. Abd al-Az3z ordered the destruction of both the ancient and the
newly built churches in amsa4r al-muslim n;82 and another according to
which Umar ordered the destruction of churches, but was challenged by
the claim that they were protected under the sulh agreement (hadha ma
76
sulihu4 alayhi). Hearing this, Umar withdrew his order and left them
standing.83 The caliph Marwa4n II plundered and destroyed monasteries and churches in Egypt;84 in 191 H Ha4ru4n al-Rash3d destroyed the
churches in the awa4sim the frontier areas;85 the anonymous Syriac
chronicle reports that during the first half of the ninth century Abdalla4h
b. Ta4hir refused to grant the request of the Muslims of Harra4n, Edessa
(Ruha4), and Samosata to destroy newly built churches and to stop the
sounding of the nawa4q s since the poor Christians have not rebuilt
one tenth of the churches which have been ruined and burnt by them
[i.e. the Muslims].86 His brother ordered the destruction of churches in
Mesopotamia, but was stopped by Abdalla4hs orders.87
As in the case of the churches, Abu4 Yu4sufs persistent position in this
chapter is in fact a reaction to more intolerant positions on the matter
of displaying the cross adopted by Muslim rulers throughout the seventh
and the eighth centuries.88 Both Theophanes and Dionysius of Tell-Mahre4
report that the mosque being built on the Temple Mount by Umar could
not be stabilized until the cross was taken down from the church on the
Mount of Olives.89 This story informs us that probably at the time of
the Dome of the Rocks building, or soon afterwards, the cross from the
Church of the Eleona, or perhaps from the Church of the Ascension, was
removed. Michael the Syrian adds here that the Arabs became the enemies
of the cross and the persecutors of the Christians because of their veneration of the cross.90 It is related by Dionysius of Tell-Mahre4 that as early
as the time of Uthma4n b. Affa4n (r. 64456) the governor of Damascus
ordered all crosses to be removed and effaced91 from walls and streets and
places open to view, and he forbade the standard of the cross to be shown
on days of feasting and supplication.92 The Jews, overjoyed, carried out
the decree enthusiastically, causing great distress to the Christians. A supplication to the am r to stop the vandalism was followed by the following
statement: I merely commanded that the crosses which we always see
when we pass through the streets should be effaced.93 Michael the Syrian
reports that Abd al-Malik ordered that crosses should be taken down
and pigs slaughtered. Theophanes reports that during Umar al-Khatta4bs
days the haters of Christ brought down many crosses.94 This phenomenon is confirmed by Ibn Asa4kir, who adduces a tradition that UmarII
b. Abd al-Az3z ordered all painted crosses and images displayed in public
to be whitewashed.95 Abd al-Razza4q cites a tradition concerning Umar
b. Abd al-Az3zs policy regarding the Christians in al-Sha4m. Among
other prohibitions pertaining to their dress and appearance there is also a
prohibition on displaying the cross on top of their churches.96
77
Abu4 Yu4sufs strong insistence on these issues underlines the fact that
they were matters of great contention among Muslim policy makers.
It can thus be claimed that his adamant position stands in opposition
to other opinions which condoned or even justified the taking over or
destruction of dhimm prayer-houses and the imposition of severe restrictions on dhimm religious liberties.
Abu4 Yu4sufs Tradition and Shuru4t Umar
The text cited by Abu4 Yu4suf97 includes a detailed list of stipulations. It
is tempting to consider this, as did Noth,98 as a text parallel to Shuru4t
Umar. Yet the attitude exhibited in this text, as well as in the additional
reports adduced by Abu4 Yu4suf, is quite different from that of the Shuru4t.
It is quite evident that given the fact that the prohibitions cited are minimal in number, and limited in their scope, it is a much more tolerant
document.
The opening statement, emphasized by Abu4 Yu4suf again and again,
declares that their churches and synagogues will not be touched. This
statement may also be inferred from Shuru4t Umar, yet it is not proclaimed explicitly. Not only that, but while Abu4 Yu4sufs version of the
treaty contains a prohibition on building new prayer-houses, in Shuru4t
Umar there is an express obligation not to restore those that had deteriorated or fallen into ruin.99 In addition, many versions of the Shuru4t
add a prohibition on congregating (ijtima4) or visiting churches located
in Muslim neighbourhoods, thus implying that churches that were
in areas that had become Muslim could not continue functioning as
such.100
Most stipulations in Abu4 Yu4sufs Kita4b al-khara4j belong to what Noth
calls the stage of the conquest, and suit the conditions immediately following the conquest. There are the obligations to cooperate with the conquerors:provide them with accommodation for three days; build bridges
over rivers for them; refrain from insulting or hurting Muslims; light fires
for the fighters of Allahs Way; refrain from misleading Muslims; and
refrain from carrying arms or keeping them at home. The other stipulations are few. They include the demand not to raise the cross in a Muslim
neighbourhood, not to sound their na4qu4s before the adha4n or at the time
of the adha4n, or to introduce pork into Muslim areas, and they are clearly
meant to avoid offending Muslim feelings. The limitation on crosses is in
Muslim neighbourhoods only, while that on the na4qu4s is during Muslim
prayer times only.
78
It follows that both the adha4n and the display of pork were permitted
anywhere and any other time as, in fact, Abu4 Yu4suf himself chooses to
emphasize.101 Although there are seemingly similar demands in Shuru4t
Umar, in fact they are quite different and exhibit a strict and much less
tolerant attitude. In Shuru4t Umar, parading the cross and the palm leaves
on Palm Sunday is expressly prohibited, and there is an additional prohibition on exhibiting the cross on top of the churches (wa-an la4 nuzhira
al-sal b ala4 kna4isuna4) at all times in all places.102 No attempt at differentiation is made. As for the na4qu4s, the demand is to sound the na4qu4s
quietly, in some versions within the church only,103 at all times.
In addition, Abu4 Yu4sufs version lacks many of the restrictions that
appear in Shuru4t Umar, such as those prohibiting dhimm s from raising
their voices in the presence of Muslims, entering their dwellings, showing
lights on Muslim roads or in Muslim markets,104 or the demand that they
show respect towards the Muslims, rise in their presence, and offer them
a seat.105 Abu4 Yu4sufs version also lacks the much-expanded concept of
the ghiya4r we find in the Shuru4t. Though he does dedicate a different
chapter to this issue (see Chapter 4), he does not attach it to the general
sulh document, nor does he expound on it as much as Shuru4t Umar.
In the latter document dhimm s not only have to appear different from
Muslims, but are also prohibited from adopting Muslim kunyas, from
teaching the Qura4n to their children, and from having Arabic inscriptions on their seals. Dhimm houses should be lower than Muslim houses
(compare al-Mutawakkils decrees:dhimm graves cannot be higher than
Muslim ones).106
In short, the difference between the approach in the texts adduced by
Abu4 Yu4suf and that of the Shuru4t is great. The former merely demands
from the dhimm s proof of loyalty towards the conquerors, limits their
freedom only where it clearly and openly offends Muslim feelings and
ritual, and shows consideration towards the dhimm s. The latter, on the
other hand, is occupied with drawing a clear and emphatic line between
the rulers and the ruled in all areas of social life so that the first always
have the upper hand, and this discrimination is not limited to places
where MuslimChristian coexistence causes friction. The Shuru4t, in other
words, encumbers dhimm life a great deal more.
Al-Sha4fi3s Version of the Agreement
Another text that proffers a formal document defining the terms of
MuslimChristian coexistence appears in al-Sha4fi3s (150204/767820)
79
Kita4b al-umm, which was written slightly later than Abu4 Yu4sufs treatise.107 Al-Sha4fi3s document is a well-known text. It was cited and translated by both Tritton and Fattal.108 Tritton even regarded it as the most
detailed document. Yet they both regarded al-Sha4fi3s text as a version
or an extension of Shuru4t Umar, and neither of them made an effort to
explain the differences between the two texts. Cohen did note the difference and claimed that Sha4fi3s version is better understood as a juridical
elaboration of the actual pact (which probably already existed in the
form we know it) preserving and amplifying its contents with details.109
It is indeed quite certain that Sha4fi3s version was a response to other such
formulas which were being drawn up. It is difficult, however, to discern
which text al-Sha4fi3 was referring to. In my opinion there were several
versions which seem to have been in existence throughout the eighth century and the beginning of the ninth, one of which was al-Sha4fi3s.
Al-Sha4fi3s text is entitled:If the Ima4m wishes to write a document
of agreement for the poll-tax he should write . The next sentence
makes it clear that the following text is a version of the agreement
between the Commander of the Faithful (Amr al-Muminn) and the
Christians of an anonymous city. It is a blank version of a document,
to be filled in at the time of signing by the current Amr al-Muminn,
on the one hand, and the Christians of the city in question, on the other.
The only detail that is filled in is the date:the second of the month of
Rab3 al-Awwal. This, according to some Muslim traditions, is the date
of Muhammads birth, possibly a symbolic date which enhanced the
significance of the pact.110
Al-Sha4fi3s version seems to have been an intermediate one, its approach
being situated somewhere in between Abu4 Yu4sufs versions and Shuru4t
Umar. Like the former, it seems to aim primarily at ensuring submission
to Muslim law and protecting Muslim society from direct insult or injury
as well as from unwanted influences of Christian society. As shall be demonstrated below, the pact adduced by al-Sha4fi3 epitomizes the authors
concept of dhimm Muslim relations. A comparison of his legal discussions concerning the various questions treated in the pact shows that
the opinions presented in the document are representative of the views
expressed by al-Sha4fi3 in his Kita4b al-umm.111 In contrast to the other
versions surveyed, al-Sha4fi3s text does not claim to go back to to Umar
or the Ra4shidu4n, and in fact does not lean on any isna4d whatsoever. It
seems that it was composed by al-Sha4fi3 himself, on the basis of texts
and documents familiar to him. It would not be presumptuous, I believe,
to assume that the issues al-Sha4fi3 chose to contend with, the order in
80
which he brought them, and the space he allotted to each of them, are all
significant. While in his legal discussions he could argue extensively and
encompass wide issues without being limited, the ahd formula confined
him to succinct and unequivocal phrasings, and compelled him to make
careful choices.
Al-Sha4fi3s treaty encompasses a large variety of issues which do not
appear in any of the other treaty formulas, yet are discussed by jurists of
the period.112 As a matter of fact, it seems that he wished to encompass
all issues with a bearing on dhimm s, not only those relating to daily life
and contacts between Muslims and dhimm s. The document he presents
may therefore be regarded as representing the main difficulties that lay at
the heart of Muslimdhimm relations, in his view.
The text opens with the validity of the application of Muslim law to
dhimm s and the rejection of any other authority:You will be subject to
the authority of Islam and to no contrary authority. You will not refuse
to carry out any obligation which we think fit to impose upon you by
virtue of this authority.113 It immediately goes on to warn against the
defamation of Muhammads name, his book, and his religion,114 presumably with reference to the many dhimm s who defied the Muslim religion
and execrated the name of Muhammad in the eighth and ninth centuries,
according to Christian sources.115
Other events perceived by al-Sha4fi3 as imminent threats are the fornication, or marriage, of a dhimm male with a Muslim woman, and the
subversion of a Muslim from his religion.116 In comparison, the issue
of fornication with or marriage to a Muslim woman is absent from the
Shuru4t altogether. The prohibition of causing the apostasy of a Muslim
is followed by the obligation:We shall not prevent any of our kin from
entering Islam if they wish it.117 This addition reflects a higher degree of
intervention in the communal life of the dhimm s than the prohibition on
apostasy from Islam.
There is a group of prohibitions related to Muslim security, an issue
which appears in the surrender agreements of the vassal-treaty type:118
the case of dhimms who rob Muslims on the highways; dhimms who
assist the enemy by fighting the Muslims or by showing the weak
points of the Muslims (dala4la ala4 awrat al-muslimn) (also in Makhu4l
al-Sha4m3s version cited by Abu4 Yu4suf119); and the harbouring of
spies.120 Of these, only the last appears in Shuru4t Umar.121 The Shuru4t
also requires the dhimms to assist the Muslims by giving them accommodation (also a clause in the surrender agreements122), and opening
the churches for them.123 This demand, interestingly enough, does not
81
appear in al-Sha4fi3s version; there is only a repeated demand, towards

the end, not to deceive a Muslim or give aid to their enemies by word
or deed.124
Another issue treated widely by al-Sha4fi3 is the marketplace. Here the
dhimm s might tempt Muslims with wine, pork, blood, carrion, or other
foods forbidden to them. The text forbids dhimm s to sell any of these to
Muslims under threat of annulment of the sale, confiscations, fines, and
punishment.125 This subject is treated extensively; it is especially revealing that the document reverts to it towards the end, repeating the points
and emphasizing that such actions nullify the protection owed to the
dhimm s.126 This is mentioned only briefly in Shuru4t Umar:We will not
keep pigs in their vicinity and we will not sell wine.127 Many versions of
the Shuru4t omit the reference to pigs and mention only the prohibition on
selling wine.128 The omission of the issue of pigs, to which both Makhu4l
al-Sha4m3s text and al-Sha4fi3 devote much attention, may indicate that the
danger had gradually faded away, or had even become irrelevant, in the
sense that the dhimm s observed this prohibition strictly, as the penalty
was well known and understood.
This opening section of al-Sha4fi3s text is much more intent on defending Muslim society than on offending ahl al-dhimma. Its author was obviously concerned with the protection of Muslim society from the harmful
influences of the dhimm s rather than with dictating the minute details of
everyday dhimm life.
The next two paragraphs refer to the legal status of the dhimm s. For
al-Sha4fi3, nothing in the relationship was self-evident. The legal status had
to be defined from the beginning:in principle, Muslims allowed dhimm s
to retain their social and legal institutions, and refrained from intervening
unless specifically asked to do so by litigants. There are certain cases in
which Muslim law was applied:when it came to murder or manslaughter,
the laws of qisa4s and diya were put into effect, while in the case of theft
the Muslim penalty for theft was applied if the thief was brought before
a Muslim judge. One hadd offence specifically mentioned here as applicable to the dhimm s is the qadhf (slander). Here again, it is quite clear
that this was an actual and relevant problem at the time, as the author
elaborates and explains:If a hadd is due, it shall be inflicted on him; if
there is no legal penalty, he shall be punished at discretion so that the
laws of Islam may be applied among you in these matters, both specified
and unspecified.129 It is apparent here that while the essential position
was to avoid intervention unless necessary, when it came to qadhf, the
honour of Islam and its adherents was at stake. Again, as in the previous
82
cases, the dominant situation is one in which Muslim honour and culture
seem to have been under threat.
The next paragraph not a very long one is, in a nutshell, a brief
summary of the restrictions applied to the dhimm s:
You may not display crosses in Muslim cities, nor proclaim polytheism,
nor build churches or meeting places for your prayers, nor strike clappers, nor proclaim your polytheistic beliefs on the subject of Jesus, son
of Miriam, or any other, to a Muslim. You shall wear the girdle (zunna4r)
over all your garments, your cloaks and the rest, so that the girdles
are not hidden. You shall differentiate yourselves by your saddles and
your mounts, and you shall distinguish between your and their headgear (qalansuwa) by a mark which you shall place on your headgear.
You shall not occupy the middle of the road or the seats in the market,
obstructing Muslims.130
It is important to emphasize that this paragraph, listing the restrictions

upon the dhimms, applies only to amsa4r al-muslimn. Although the
text could also be otherwise understood so far (the reference to amsa4r
al-muslimn may apply only to the display of crosses), this point is made
quite clear when we read al-Sha4fi3s section dedicated to the amsa4r a
few pages later.131 It is here that he adduces his unequivocal definition of amsa4r al-muslimn as already noted above,132 explicitly stating that the uhu4d must be respected under all circumstances. Only in
amsa4r al-muslimn, he claims, does this list of restrictions apply. Thus,
for al-Sha4fi3, the general sulh agreement does not abrogate the individual uhu4d; rather, it contains them. His agreement provides a general
description of the terms under which dhimms live under Muslim rule
and a list of restrictions pertaining to amsa4r al-muslimn (according to
his definition) only, a list which does not affect the cities possessing sulh
agreements.
A comparison of these restrictions to the other formulae shows the
following: his stipulations are less tolerant than those presented by
Abu4 Yu4suf, yet they are less restrictive than the ones appearing in the
Shuru4t:here, as in the Shuru4t, crosses are not to be displayed in Muslim
cities at all times there is no exception made for Palm Sunday, as there is
in Abu4 Yu4suf; here, as there, too, the manifestation of polytheism (shirk)
is prohibited, and al-Sha4fi3 even specifies that shirk includes belief in
the divinity of Jesus, son of Miriam. (This prohibition does not appear
expressly in Abu4 Yu4suf, yet it is implied in the question of parading the
crosses without the standards (la4 yukhriju4 al-ra4ya4t f ayya4m dihim)
which in all probability displayed images of icons.)133 The building of
83
new churches is prohibited, as in Abu4 Yu4suf and the Shuru4t; yet al-Sha4fi3
explicitly says that if the sulh allowed the building of churches, then it is
allowed!
Like Abu4 Yu4suf, al-Sha4fi3 explicitly says that existing churches in
amsa4r al-muslim n should not be destroyed. He does not prohibit the
gathering of Christians in such churches or the repair of churches that
have fallen into ruin, as do most of the Shurut4 texts.134 In the chapter on
the amsa4r al-Sha4fi3 states that if the Christians live separately in a city of
their own, they are not to be prevented from building a new church, nor
from erecting tall buildings, nor are they hindered from [exhibiting] their
pigs and their wine and their holidays and their assemblies. It is only
demanded from them that they will not offer wine to a Muslim who has
come to their town.
As for the ghiya4r, they are quite similar to those appearing in Abu4
Yu4suf, but in Abu4 Yu4suf they apply to all dhimms,135 whereas al-Sha4fi3
limits them to the amsa4r, emphasizing their important role in differentiating between Muslims and non-Muslims. Al-Sha4fi3 also adds a prohibition on occupying the middle of the road or seats in the market, the
latter appearing in the Shuru4t along with additional stipulations such
as respecting Muslims in general, and honouring them in their seats.
This list of stipulations is apparently closer both in content and attitude to the Shuru4t than Abu4 Yu4suf, yet it applies specifically to the amsa4r,
while the Shuru4t tends to do away with this differentiation, applying
many of the restrictions to dhimm s everywhere under Muslim rule. In
addition, al-Sha4fi3s list lacks many stipulations that appear in the Shuru4t,
such as the prohibition on using the kunya, speaking as the Muslims do,
teaching the Qura4n to their children, possessing seals in Arabic, raising
their voices in church, or in the presence of Muslims, taking their funerary processions through Muslim roads, raising voices or lighting fires in
funerals, or burying their dead near Muslims. All in all, the author dedicates only a short paragraph in quite a long document to the restrictions
applied to the dhimm s.
Al-Sha4fi3s approach is thus far removed from that of the Shuru4t. All
the restrictions he lists are applicable only in the case of Muslimdhimm
coexistence in the same city; and even then, only if it was not agreed
otherwise in the sulh. In some cases, therefore, Muslims were bound to
accept the building of new churches, and the public exhibition of pigs and
wine in their own metropoles, without objection.
This is not true of Shuru4t Umar, where all of these prohibitions
are absolute, and there is little differentiation between Muslim and
84
non-Muslim cities:the dhimm s are expected to adhere to the conditions

stipulated in all situations. It should be noted that there are some prohibitions referring only to amsa4r al-muslim n which appear in different
versions of the Shuru4t, but these are inconsistent most of the time.136 The
text of the Shuru4t does not relate to the specific sulh agreements, nor does
it imply that they were still valid. It was the text of the Shuru4t that was
applicable to the dhimm s, wherever they were, at all times.
Al-Sha4fi3s tolerance towards the dhimms is also expressed in the
following sections. One is dedicated to the jizya. The payment of the
jizya, i.e. the obligation enforced upon the dhimm as a condition for
the covenant, is parallel in the text to the Muslim obligation to protect
the dhimms:
We owe you protection, for yourselves and for property which it is lawful for you to hold according to our laws, against anybody, Muslim
or other, who seeks to wrong you, as we would protect our own persons and our own property, and we administer justice to you in matters
under our jurisdiction as we do with our own property.137
Al-Sha4fi3 chooses to elaborate here the rights of the dhimms side by

side with their obligations. In contrast, the text of the Shuru4t refers to
the ama4n only in passing: it is something the dhimms ask for when
accepting the Shuru4t, yet the text does not take the trouble to confirm
that it was indeed accorded to them.138 More detailed than Abu4 Yu4sufs
obligations, it is apparent that this is a bilateral agreement rather than
a unilateral list of obligations, as in the Shuru4t. In the Shuru4t, as Mark
Cohen has indicated,139 a legal form is used in which the dhimms ask
for the ama4n and provide the list of obligations themselves. The Muslim
side in this agreement is tacit and passive. This is exactly the opposite
approach from that of Abu4 Yu4suf and al-Sha4fi3, where the Muslims are
the speakers in the text, and so the active initiators of the pact.
The Nusha4tiruhum Version
Another version of the agreement with the dhimm s is cited with minor
variations by both Ibn Asa4kir 140 and Gha4z3 b. al-Wa4sit3.141 Ibn Asa4kir
claims that this treaty was signed by Iya4d b. Ghanm with the people of
al-Jaz3ra who assembled in Ruha4; they were familiar with the conditions of Umar b. al-Khatta4b to the people of al-Sha4m (shuru4t Umar
b. al-Khatta4b ala4 ahl al-sha4m) and agreed to surrender on the same
85
conditions. Ibn al-Wa4sit3 reports that it was given by Umar to Constantin

(?) in al-Sha4m.
This version bears some resemblance to both Makhu4l al-Sha4m3s and
Shuru4t Umar. Like Makhu4l al-Sha4m3s it is quite short and ends with a
mutual obligation to observe its terms. The issues covered are also much
the same, i.e. the building of churches, the cross, the na4qu4s, accommodation of Muslims by dhimm s, guiding the Muslims, not keeping pigs in
Muslim neighbourhoods, advising the Muslims, not deceiving them, and
not assisting the enemy. Both Ibn Asa4kirs text and Ibn al-Wa4sitis seem
to be simple treaties, most probably amplified versions of some of the
original uhu4d. Some of the elements of this version here are reminiscent
of the Shuru4t:thus it allows the sounding of the na4qu4s within the church
only illa4 f jawf al-kan sa exactly the same wording as the Shuru4t; as
for display of the cross, Ibn Asa4kirs version prohibits it totally, as in the
Shuru4t, while Ibn al-Wa4sit3s version prohibits it in Muslim neighbourhoods only. However, in other parts the nusha4tiruhum version uses linguistic formulae that are different from both the Shuru4t and Abu4 Yu4suf.
Above all, the nusha4tiruhum version carries a unique condition, which
does not appear in any of the above treaty formulas:half of the dhimm
dwellings are to be confiscated (wa-ishtarata alayhim an yusha4tiruhum
mana4zilahum wa-yanzila f ha al-muslimu4n) and, in Ibn al-Wa4sit3s version, the southern wall of any confiscated church is to be turned into
a mosque. The confiscation of the southern walls of the churches for
conversion into mosques also appears in Ibn Asa4kirs account of the
conditions imposed on Damascus, a few lines above and in exactly the
same wording, adding that this wall is cleaner and purer (liannaha4 anzaf
wa-athar).142 The stipulation regarding the confiscation of half the homes
and/or churches is known from the individual treaties,143 but not from
other general treaty versions, and is also echoed in some traditions cited
by Ibn Asa4kir elsewhere.144 This stipulation makes the nusha4tiruhum
treaty the most radical and intolerant in its attitude towards the dhimm
population.
conclusion
It appears that around 800 CE there were already several developed
versions which conveyed a general formula of the ahd of the Muslims
with the dhimm s. All the versions but that of al-Sha4fi3, who seems to
have compiled such a version himself, claimed to draw on ancient and
86
respectable sources, exhibiting a reputable isna4d for the document. The

purpose of these formulae was to provide a comprehensive document that
would define the terms applying to dhimm s living under Muslim rule in
general, and in places inhabited by Muslims in particular. Its purpose
was to solve the obscurity concerning the definition of a Muslim misr,
the rules that applied in a city defined as such, and the sensitive problem
of the individual uhu4d possessed by the cities conquered sulhan, in one
fixed formula.
Several different versions, representing different approaches, were
drawn up and circulated. Abu4 Yu4sufs version is the most tolerant,
and in fact serves as a basis for his declared position that dhimm
rights, which were agreed upon at the time of the conquest, should be
respected. Limitations and restrictions imposed on the dhimms should
be minimal, and should apply only should Christians interfere directly
with Muslim life and religion. Al-Sha4fi3s version is somewhat less
tolerant, imposing more restrictions on the dhimms than Abu4 Yu4suf.
Yet he believed that these restrictions should apply only in cases of
MuslimChristian coexistence, and that all previous sulh commitments
should be respected. The restrictions form only a small and comparatively minor section in his version of the treaty, and the document bears
a mutual character enumerating the rights and the obligation of both
sides, and emphasizes the protection of Muslim society from any situation that is injurious to it.
In contrast, the classic text of the Shuru4t is completely unilateral. The
obligation of an ama4n on the part of the Muslims can only be surmised
from the request, while the dhimm s take upon themselves a long and
detailed list of restrictions. The differentiation between Muslim and nonMuslim cities is often obscured, most restrictions being applicable to all
dhimm s wherever they are; nor does there seem to be any obligation to
the original sulh treaties, as in the case of al-Sha4fi3. A large part of the
restrictions that appear in the Shuru4t no longer function as protectors of
Muslim society, but rather as a means of emphasizing Muslim dominance
and sovereignty, as opposed to dhimm subordination and subservience.
The nusha4tiruhum version represents an even more extreme attitude
than that expressed in Shuru4t Umar, condoning the confiscation of half
the dhimm s property.
It seems that at the end of the eighth and beginning of the ninth centuries there was no consensus regarding a general sulh version which would
take the place of the original surrender agreements, and that the issue was
being debated quite intensively.
87
If indeed Shuru4t Umar was already drawn up and circulated by Abu4

Yu4suf and al-Sha4fi3s time, it does not seem to have had a prominent position among the known versions at the time. It seems that it was not until
the middle of the ninth century during the reign of al-Mutawakkil
that it came forth and acquired priority,145 gradually pushing aside other
versions, which presented more liberal and tolerant approaches, and
finally obtained exclusivity.
3
The Date and the Ideology of the Ghiya4r Code
This chapter will try and tackle the question of the date and circumstances in which the element of the ghiya4r (lit. distinguishing signs), that
is, the restrictions regarding the distinctive appearance and behaviour of
non-Muslims, which form the central component of Shuru4t Umar, was
developed.1 I should like to emphasize that I refer here to the dating of a
structured code or set of rules rather than this or that particular regulation, which may in some cases have existed from the beginning of Muslim
rule. There is no doubt that initially we are looking at a process rather
than at a moment in history. Still, there must have been a point in time at
which such a code of dress and appearance was first issued.
In this chapter I will attempt to demonstrate that, despite the inclination
to cast doubt on their veracity, the Muslim sources are correct in attributing the first code regarding the attire and behaviour of non-Muslims in
Muslim society, to the caliph Umar b. Abd al-Azz,
3 and that this code
was indeed part of a planned and deliberate policy which was a result of
his ideology regarding the ascendancy of Islam over the other religions.
It should, however, be emphasized that while there are noted ideological differences between the various codes that were in a state of formation around the year 800 CE (see Chapter 2), all of these sources i.e.
Abu4 Yu4suf,2 al-Sha4fi3, and Shuru4t Umar include a section regarding the
ghiya4r wherein the variations are both minimal and technical.3 They all
insist that the non-Muslims dress and mount should differ from that of
the Muslims, the versions varying slightly from one another in the appurtenances mentioned. It thus seems that, in contrast to matters mentioned
above, there were no essential disagreements or arguments regarding the
issue of the ghiya4r.
88
89
The sources all point in one direction:they all attribute the creation of
the ghiya4r to the caliph Umar b. Abd al-Az3z (r. 99101/71720). Umar
II was, according to tradition, the author of an edict containing a list of
regulations imposed upon the dhimm s.4 Yet one can very easily perhaps
somewhat too easily dismiss this traditional claim as part of the myth
in which Umar b. Abd al-Az3zs figure is shrouded.5 There is no need
to develop here his image in later generations as al-khal fa al-a4dil, an
exemplar of the Muslim virtues of piety, equity and humility.6 Agreatgrandson of the caliph Umar I b. al-Khatta4b on his mothers side, he
was the only Umayyad caliph who was respected, and even revered at
times. Being a symbol of Muslim righteousness, he was believed to be
the first mujaddid (renewer), in a tradition that originated most probably
following the death of al-Sha4fi3.7 Hawting suggests that Umar IIs pious
image was a means of attaining continuity from the Ra4shidu4n through
the Umayyads to the Abba4sids, and for rejecting Sh3ite claims.8 It is thus
tempting to claim that the traditional attribution of the creation of a code
of rules to Umar b. Abd al-Az3z is part of the myth that surrounded the
actual person.
However, the sources seem to insist unanimously and coherently that
Umar b. Abd al-Az3z published an edict that contained a set of regulations regarding the dhimm s. Antoine Fattal, who tackled this question
fifty years ago, reached the conclusion that Umar II had indeed issued
this code of rules, adding that the homonymy was no doubt the cause
of the confusion that led later to the attribution of these rules which
were included in Shuru4t Umar to Umar I.9 I will try to provide further
support for this claim.
the sources regarding umar b. abd al-az3zs edict

The sources citing Umar b. Abd al-Az3zs edict, including Ibn Abd
al-Hakam, Abu4 Yu4suf, Abd al-Razza4q and Abu4 Ubayd, Ibn Asa4kir,10 and
Ibn Zabr,11 are quite consistent regarding the rules included in this edict,
and in fact exhibit a strong linguistic similarity in some parts.
The edict seems to have consisted of two parts:one prohibiting nonMuslims, men and women, from using saddles; the other regulating attire
and appearance, including the obligation to cut the forelock, to wear a
(leather) girdle, and not to wear shoes with straps, a taylasa4n (headgear,
usually a cowl worn over the turban),12 a qaba4 (a luxurious robe), or
a turban (ima4ma or isb).13 In additional matters there are variations
between the versions of Ibn Abd al-Hakam, Ibn Zabr, and Ibn Asa4kir
90
on one hand, and Abu4 Yu4suf on the other, who seems to adduce here a
slightly different edict. These, however, are minor, and in any case not
contradictory.
Abu4 Yu4suf14 and Ibn Abd al-Hakam cite almost word for word the section regarding the prohibition on using a saddle (sirj or riha4la the first
used when referring to men, while the latter when referring to women),
while Abd al-Razza4q employs the same terms. The non-Muslims were
allowed to use only the ika4f (pack-saddle). According to the edict cited by
Ibn Abd al-Hakam they were barred from riding with their legs spread
out, and could only ride with both legs to one side, as women did:
And take care that a Christian shall not ride on a saddle but that they
should ride on pack-saddles, and that a woman of their women shall not
ride saddled, and that she should ride on a pack-saddle, and that they
should not sit on the riding-animals with their legs apart, and that they
should place their legs on one side. Write them a firm letter regarding
this and satisfy me regarding this [matter].15
Regarding clothing, Ibn Abd al-Hakam (with no isna4d),16 the Shuru4t

al-nasa4ra of Ibn Zabr (870940 CE), citing a had th going back to Sufya4n
al-Thawr3 from Talha b. Mussaraf from Masru4k b. Abd al-Rahma4n b.
Ghanam,17 and another version with an isna4d going back to al-Hakam
b. Umar al-Ruayn3 who heard from Umar b. Abd al-Az3z,18 and Ibn
Asa4kir with versions and isna4ds identical to Ibn Zabr19 all cite exactly the
same section of the edict:
Umar b. Abd al-Az3z wrote to the cities of al-Sha4m that a Christian
must always have his forelocks trimmed (mafru4q al-na4siya),20 he should
not wear a robe (qaba4), he should always wear a leather girdle (zunna4r
min jalad), he should not wear a Persian mantle (taylasa4n), nor trousers
with anklets (sara4wl dha4t khadama) nor shoes with straps (adhaba),21
he should not ride on a saddle, and if arms are found in their homes they
should be confiscated.22
Ibn Qayyim al-Jawziyya cites Abu4 Yu4sufs version of this section almost
word for word, with slight omissions and additions.23
As for the term used for girdle:while all versions of this had th use the
term zunna4r, Ibn Abd al-Hakam, Abu4 Yu4suf, Abd al-Razza4q and Abu4
Ubayd all use the term mintaqa when citing the edict, rather than zunna4r.
This seems to indicate that the edict was written before the term mintaqa
was exclusively allocated to the Muslims and zunna4r to non-Muslims,
and the distinct differentiation between these two types of girdles was
91
not yet made, as shown in Chapter 5.24 In the paragraph preceding the
edict Abu4 Yu4suf elects to use the term zunna4r rather than mintaqa, used
in the edict itself. Abu4 Ubayd found it helpful to explain:they should
wear mana4tiq, that means zana4n r. This lends additional support to an
early date for the edict.
The items of clothing mentioned in the edict are of Iranian origin and,
as will be shown below, were in use in caliphal circles in Syria by the first
half of the eighth century.25
Additionally, Abu4 Yusuf
4
and Abd al-Razzaq
4 also include the prohibition on exhibiting the cross in public, and Abd al-Razzaq
4 adds the prohibition on beating the naqu
4 s4 , which is mentioned by Michael the Syrian,
Chronicon ad 1234 and Bar Hebraeus as well.26 Yet the heart of the edict
seems to be well preserved. In fact, Ibn Zabr and Ibn Asakir
4
who both
adduce the same version of the edict note:I did not see this addition in
what has been passed down to us of Shurut4 Umar; rather, I found it related
from Umar b. Abd al-Azz.
3 27 It is interesting to note that, unlike many
other cases regarding regulations governing the non-Muslims, there is no
confusion between Umar b. al-Khatt a b
4 and Umar b. Abd al-Azz3 here.
Eastern Christian tradition also supports the claim of the Muslim
sources.28 Both Michael the Syrian and the Chronicon ad 1234 describe
in detail Umar b. Abd al-Az3zs treatment of the Christians. Although
these sources are quite late, they both derive their information from the
common source, usually believed to be Theophilus of Edessa, which
crystallized before 750.29 Acknowledging his pious reputation, they both
go on to say that he had treated the Christians in an especially harsh
manner and that he was the initiator of a series of regulations against the
Christians. The Chronicon ad 1234 says:Nevertheless he was very hostile to the Christians, more than the kings before him.30 Michael claims
that Umars attitude towards the Christians was due both to his desire to
validate Muslim law and to the wounding defeat suffered by the Muslims
in their attack on Constantinople, which filled him with zeal and left him
greatly opposed to the Christians. Among other things31 the Chronicon
ad 1234, Michael, and Bar Hebraeus mention the prohibition on raising
their voices in prayer, on striking the na4qu4s, and on riding on a saddle.
Michael, citing Dionysius of Tell-Mahre4, adds the prohibition on serving
in high office and on wearing the qaba4 (the latter is not mentioned in
Chronicon ad 1234), while Bar Hebraeus notes that they were prohibited
from wearing the dress of soldiers.
In addition to the Muslim and Christian citations and mentions of
the edict, there is also a Muslim tradition that links these regulations
92
to Umar b. Abd al-Az3z.32 According to this tradition some people

from Banu4 Taghlib who were Christian Arabs appeared before Umar b.
Abd al-Az3z wearing turbans (ama4im), in the garb of the Arabs
(ka-hayat al-arab). When he understood that they were Christians he
at once ordered their forelocks to be cut off, their turbans to be thrown
away, and pieces of their cloaks to be cut off and used as girdles. He
also ordered that they should not use saddles, only pack-saddles, and
that they should ride with both their legs to one side. Admittedly, this
tradition is very late, although its isna4d, if accepted, actually goes back
to the days of Umar II himself. This story is echoed in another tradition
regarding the meeting of Umar II with the Tanu4kh, who were also a
Christian Arab tribe.33 There is a later version of this as well.34 There was
obviously a topos regarding the encounter of Umar II with the Christian
Arabs, who posed a special challenge to the Muslims. Both versions go
along with the edict attributed to Umar II in claiming that riding on
horseback and wearing turbans, ka-hayat al-arab, was an exclusive
Muslim prerogative, since Arab was expected to be synonymous with
Muslim. Tensions between Muslim and Christian Arabs were in the air
from early days of Islam and, unlike other Christians, Christian Arabs
were greatly pressured to convert to Islam, as is testified by Anastasius
of Sinai (d. c. 700).35 It is quite plausible therefore that the Muslims
insisted that Arabs who did not convert to Islam were not to be allowed
to appear in the garb of ruling Muslims, despite their being Arabs,36 and
were expected to abide by the edict of the ghiya4r, a concept that was on
its way to being well established at that time, if indeed the edict cited is
genuine.
the ideology behind the edict

The link to Umar b. Abd al-Az3z is supported by more than just circumstantial evidence such as its unanimous attribution to Umar by the
sources, and their consistency regarding its contents. It does, in fact, seem
to be supported by a consistent ideology which was zealously proclaimed
by Umar b. Abd al-Az3z himself. Moreover, the edict is directly attached
to the proclamation of this ideology.
This is well demonstrated in two of the rasa4il that appear in Ibn Abd
al-Hakams s ra of Umar b. Abd al-Az3z. Ibn Abd al-Hakams s ra
includes a collection of edicts drawn up by Umar b. Abd al-Az3z.37 The
question that arises of course is whether these can indeed be considered
authentic.
93
In his article on the formation and diffusion of Umar IIs image Antoine
Borrut comes to the conclusion that although a large part of Umayyad
history was successively reconstructed throughout the ninth and tenth
centuries, many elements relating to Umars personality and policy do
indeed go back to the first half of the eighth. Among many other things,
he refers specifically to the information adduced in Ibn Abd al-Hakams
s ra, where Ma4lik b. Anas is mentioned in the isna4d as first among a group
of transmitters.38 Putting together the fact that Ibn Abd al-Hakam had
an important role in the diffusion of Ma4lik3 opinions in Egypt, and that
Umar II is widely mentioned in Ma4liks Muwatta, among other things as
governor of Medina, Borrut comes to the conclusion that it is indeed very
likely that Ibn Abd al-Hakam received his information from Medinese
Ma4lik3 sources which were well informed about Umar II, and therefore
that the information in the s ra may be considered trustworthy.39
Regarding the edicts themselves:it has already been noted long ago
by H. A. R. Gibb that the famous fiscal edict adduced in the s ra carries
every indication of genuineness in its content and linguistic style.40
Additional support for the authenticity of the edicts may be gained
from the fact that the s ra includes not only documents of an ideological
or religious nature that accentuate Umars pious character and devoutness, but many edicts relating to mundane matters, which do not serve
solely to exalt Umar and to sing his praises. Such, for example, are the
edict to umara4 al-ajna4d;41 the edict to the khawa4rij;42 a document given
by Umar to Mansu4r b. Gha4lib which contains instructions regarding his
military expedition, including among other things instructions regarding
the treatment of his soldiers, rules concerning the behaviour of the army
in sulh cities, and his choice of spies;43 a letter to the governor;44 another
letter to the khawa4rij;45 and an edict regarding maximum loads of pack
animals.46 Although some of the documents appear in the s ra by Ibn
al-Jawz3,47 Ibn Abd al-Hakams s ra is abundant in documents not cited
by Ibn al-Jawz3.
Michael Cook, who cast serious doubts on some of the early theological rasa4il, among them also on the anti-Qadarite Epistle of Umar b. Abd
al-Az3z,48 also seems to believe that Umars fiscal edict and his long and
pious accession epistle preceded by the wasiyya bi-l-taqwa (injunction
regarding the fear of God), to be discussed below,49 as well as other
epistles in Ibn Abd al-Hakams s ra preceded in this manner, are products of Umars time.50 The accession epistle is not only preceded by the
wasiyya bi-l-taqwa, but also includes a long survey of the early history
of Islam, called the mission topos by Cook, both typical elements of the
94
early religious epistles.51 Crone and Zimmermann point out that the Sitz
imLeben of these epistles could well be the Friday sermon (khutba).52
Another indication pointing to an early date for the documents in Ibn
Abd al-Hakams s ra is the frequent citation of Qura4nic passages and
the rarity of had th.53 Given these considerations, there is no real reason,
therefore, to doubt the authenticity of the documents cited.
Two edicts in Abd al-Hakams s ra support the idea that the publication of the edict regarding the non-Muslims was part of Umars promotion of the ideology of the exaltation of Islam and the degradation of the
non-Muslims. The concept of the chosen people, which was central in
Jewish and Christian theology, was adopted by Muhammad; it is well
represented in the Qura4n, and is even more prominent in later tradition.
Its adoption is naturally coupled with the degradation of unbelievers.54
One edict is the pious accession epistle. Here Umar emphasizes the
fact that before they became Muslims, the Arabs were a degraded and
contemptible people while the entire worlds goods were given to others,
until God decided to honour them with his Book and his Prophet.55 It
was only then that God exchanged this [misery] for you with the bounty,
the victory, the security, and the multitude, and plundered from the others
what you could not plunder with your own might if He had left you on
your own.56 This is supported by a citation from Sura 24:55:God has
promised those of you who believe and do righteous deeds that He will
surely make you successors in the land, even as He made those who were
before them successors, and that He will surely establish their religion
for them that He has approved for them, and will give them in exchange,
after their fear, security.57 The Muslims were chosen by God, and thus
entitled to all the worlds blessings.58 According to this view, as a result of
the gift of Islam the followers of Muhammad were exalted to this superior position in the world, while the others were deprived of the bounties
and advantages that they had previously possessed.
The lowly and humiliating position of those who had been deprived
of their previous advantages, according to this theology, needs to be
guarded and preserved by Islam. This idea is emphasized in an edict
cited by Ibn Abd al-Hakam which, according to him, was sent by
Umar b. Abd al-Az3z to his governors. The edict opens with the statement:The mushriku4n are impure since God has made them the army of
Satan.59 The term mushriku4n here clearly denotes non-Muslims in general rather than pagans, as becomes clear in the following sentences.60
It goes on to speak about about ahl al-shirk who had so far aided
the Muslims, levying taxes and serving as officials and administrators
95
under Muslim rule, a situation which was brought to an end by Amr

al-Muminn. The conclusion is that:
Any official or administrator that I am informed of in your district who
is not a Muslim should be dismissed by you and a Muslim should be
placed in his stead. The annihilation of their deeds is the annihilation
of their religions! It is fitting for them to be reduced to the degree of
humility and contempt (dhull wa-al-sagha4r) to which God had reduced
them. Do that and write to me of the action you took.61
It is especially significant that this declaration forms the prelude to the

edict we have cited above regarding the riding restrictions.
The next edict, directly following this, concerns the dress regulations:62
Umar wrote to the provinces (a4fa4q):a Christian must be distinguished
by his [trimmed] forelocks (mafru4q al-na4siya);63 he shall not wear a
qaba4,64 he shall not walk about except with a leather belt (zunna4r min
julu4d),65 nor [shall he wear] a Persian mantle (taylasa4n), nor trousers
with anklets (sara4wl dha4t khadama) nor shoes with straps (adhaba);
arms shall not be found in his home.
Thus, the ghiya4r edict is a direct consequence of the exaltation of Islam

and the state of humility and degradation that was to be imposed upon
non-Muslims.
The declaration that the non-Muslims need to be reduced to the
degree of humility and contempt (dhull wa-al-sagha4r) is clearly in
concert with the aristocratic ethos adopted by the Muslims from the
Sasanians, according to which the nobility must be dressed in appropriate
attire, while the lower class should wear the dress of baseness or humility
(liba4s al-madhalla) in accordance with its humiliated condition.66
There is no question here of trying to achieve mere technical distinction between Muslims and non-Muslims, meant essentially to distinguish
the Muslims from the non-Muslim majority (rather than degrade the latter), as suggested by Noth,67 but rather a fulfilment or implementation of
the notion expressed in Qura4n 2:61 that all non-believers shall be struck
by humiliation and misery (al-dhilla wa-al-maskana; see also Suras
3:112, 42:45).
The policy executed in this case by Umar II, establishing the superiority of Muslims over the non-Muslims who were still in control in
many vital places, tallies perfectly with his image as a mahd in his time,
following his great ancestor Umar I, who was called al-fa4ru4q (akin to
Syriac pa4ro4qa4:saviour, redeemer).68
96
This same doctrine and structure is repeated 130 years later, in

al-Mutawakkils edict. First he expounds upon the concept of the
Muslims as the chosen people, exalted through their faith and religion
provided that they fulfil the commandments of Islam. He then goes on to
say:The Muslims through Gods favor by which He has elected them,
and the superiority He gave them by the religion He chose for them, are
distinguished from members of other religions by their righteous laws,
their fine and upright statutes, and their evident proof. Only then does he
proceed to supply the detailed instructions regarding the ghiya4r, a direct
consequence of the distinction made previously.69
It should be noted that this concept, usually represented by the hendiadys al-dhull wa-al-sagha4r, used in the edict, is reiterated in the tafs r
and had th literature wherein the non-Muslim is justly doomed to a life
of misery and humiliation.70
Who in reality were the non-Muslims who were required to follow
these new regulations? As has been noted already by Fattal, it is quite
likely that first and foremost Umar II wanted to prevent the non-Muslims
from looking like Muslim soldiers, as indeed stated by Bar Hebraeus.71
Since there were indeed a significant number of newly converted dhimm s
who, as demonstrated by Crone, had joined the Muslim army just in
order to be given the chance to become a part of Muslim society, it was
especially important to be able to differentiate between these and other
dhimm s who pretended to pass as such.72 However, this edict was no
doubt relevant to all non-Muslims living in the garrison cities, as well as
in the major cities of the ajna4d, and included women as well as men. The
most likely candidates were the dihqa4ns (local notables) and the kutta4b
(scribes), who were to be dismissed from office according to the edict.
However, dismissed or not, they could still be walking around clad in
smart Sasanian clothes, exuding status and rank. This had to be mended.
The aim of Umar IIs edict was to create a state in which only Muslims
could appear in dress and paraphernalia signifying social superiority.
the date of the adoption of iranian dress codes

As will be shown in Chapter 5, the concept of displaying social hierarchy via codes of dress and appearance, which forms the basis of the
ghiya4r, was inspired by the Sasanian ethos, where each class had its own
dress code. The question that arises immediately following this conclusion is whether these Sasanian ideas had infiltrated the Muslim elites by
the days of Umar b. Abd al-Az3z, when the capital was still Damascus,
97
before it was transferred to Baghdad. The likely assumption would be,

of course, that the adoption of such codes would occur only after the
Abba4sid revolution.
There is, however, plenty of evidence to support the early infiltration
of Sasanian elements into Muslim dress, art, symbols, and attributes in
areas that had been previously controlled by the Byzantines.73 There was
extensive adoption of Sasanian elements such as dress, paraphernalia,
and status symbols among the Muslim elite during the end of the seventh
century and the first half of the eighth. This erases any doubts that may be
raised regarding the possibility that Umar b. Abd al-Az3z implemented
the Sasanian dress code in order to create a new social order.
Indeed, H. A. R. Gibb has already noted that although the Muslim
world turned its back for good on Byzantium and its culture after the
shameful defeat at Constantinople, this process was indeed well on its
way even before this historic milestone. At this time, ideologically, however, the center of Muslim culture and thought was already located in
Iraq, and the imperial background and determinants of the Arabs of Iraq
were not Byzantine, but Persian, and hostile to Byzantium.74
In conclusion, an examination of the sources supports the traditional
claim that it was Umar b. Abd al-Az3z who institutionalized the use
of ghiya4r; this served as a means of establishing the exaltation of Islam
and the humiliation and degradation of the non-Muslims. During the
eighth century the use of ghiya4r in order to achieve this end was further entrenched and embedded in Muslim society. This can be concluded
from several sources. Thus, Abu4 Yu4suf, in his prologue to this edict, gives
detailed instructions regarding the ghiya4r that are to be imposed upon
the non-Muslims, which are much more detailed than those appearing
in Umar b. Abd al-Az3zs edict, which he himself cites. He describes in
detail the zunna4r, the variegated qala4nis, and their distinct saddles; he
mentions the prohibition on wearing turbans (ama4im), and on wearing
the hair long, and not cutting the forelocks. He then goes on to cite a letter written by Umar b. Abd al-Az3z to one of his governors, reprimanding him for being too lax, and allowing the non-Muslims to dress and
appear otherwise. This, Umar says, is actually a sign of the weakness and
impotence of the governors.75 It is therefore evident that the principle of
ghiya4r, which was instituted in the days of Umar b. Abd al-Az3z, became
entrenched in Muslim society despite the fact that its implementation had
been inconsistent at certain times in certain places.
Al-Sha4fi3, writing around 800, in his model pact with non-Muslim subjects, presents the issue of the ghiya4r in a succinct yet clear manner:You
98
shall wear the zunna4r over all your garments, your cloaks and the rest,
so that the girdles are not hidden. You shall differentiate yourselves by
your saddles and your mounts (suru4jikum wa-raku4bikum), and you shall
distinguish your and their qalansuwas by a mark which you shall place
on your qalansuwas (bi-ilm tajalu4nahu bi-qala4nisikum). You shall not
occupy the middle of the road or the seats in the market, obstructing
Muslims.76 It seems that by al-Sha4fi3s time the regulations regarding the
ghiya4r were quite clear and well known to all, and he felt no need to go
into superfluous details or explanations.
While the gradual adoption of Sasanian elements of dress and appearance seems to have begun soon after the conquest, the establishment of
the model according to which Muslims formed the upper class while
the non-Muslims were the inferior class took place under Umar b. Abd
al-Az3z. From then on this code became accepted, and took root in
Muslim society despite lapses and periods of laxity in its enforcement.
By the last quarter of the eighth century it seems to have been conceived
among the ruling Muslim aristocracy as the rule and norm, to be respected
and implemented. This code, which seems first to have been laid down at
the beginning of the eighth century, was later incorporated into all three
known alternative versions of a general document pertaining to the nonMuslims, among them the one that had triumphed:Shuru4t Umar.
4
The Enforcement of Shuru4t Umar
In the previous chapters we tracked the development of formal relations

between the Muslim conquerors and the communities of the conquered,
from the surrender agreements to the formation of a comprehensive
legal document containing rules and regulations regarding non-Muslims
around 800 CE. The question that remains is: were Shuru4t Umar
enforced, and if so, to what extent?
The idea that these regulations may have remained on paper, or at
least were only sporadically and rarely enforced, is not far-fetched. On
the contrary, it is common knowledge that laws often fail to be implemented, and scholars concerned with the treatment of ahl al-dhimma
have mostly held that in the early centuries restrictions enjoined upon
dhimm s were irregular and sporadic and, when issued, often not
enforced. Two famous episodes the restrictions imposed by the caliph
al-Mutawakkil (r. 84761), and those of al-Ha4kim (r. 9961020) are
seen as exceptions in the sense that they were imposed, but typical in
the sense that they did not remain in force. Thus Fattal is of the opinion that les dits de Mutawakkil tombrent vite en dsutude. Muqtadir
(908932) en Irak, et Muhammad al-Ikhsh3d (934) en Egypte, essayrent
en vain de les faire revivre.1 Goitein states that practice must have differed widely from theory, and that the bizarre edict on the attire of
Christians and Jews promulgated by the caliph al-Ha4kim in a spasmodic
fit of religious zeal (or political expediency) proves only that no such
discrimination had been customary before. He goes on to say that on
other occasions, when the authorities did enforce restrictions concerning dhimm attire, this was done only in order to extort money from the
dhimm s. He does, however, concede that towards the end of the twelfth
99
100
century the wearing of distinctive marks by non-Muslims was already

generally accepted and the stern warning was addressed only to a few
transgressors, presumably of the upper class.2 Gil believes that even the
restrictions of al-Mutawakkil were not enforced, at least not systematically:These restrictions were imposed by the rulers in Palestine as well,
but to the best of my knowledge there is nothing in the sources indicating
their particular application in Palestine. However there is no doubt that
they were theoretically imposed in Palestine. Except that here, as elsewhere in the caliphate they were not rigorously observed; this is the only
explanation for the fact that they had to be renewed from time to time.3
It would indeed seem that one cannot speak of a consistent policy of
enforcement during the first century of Islam, but the situation changed
considerably from the second, and more particularly from the third
onwards, when a single set of rules was not only promulgated, but its
enforcement was attempted by many Muslim rulers.
As demonstrated in the previous chapters, the process of canonization of a comprehensive legal document containing rules and regulations
regarding the non-Muslims which began with the formulation of the
ghiya4r came to its completion in the well-known Shuru4t Umar. Before its
canonization, the Shuru4t was one of several competing documents. The
canonization of the text accepted finally as Shuru4t Umar involved therefore the victory of its supporters over the promoters of alternative views.
This victory is attested in the edict that was promulgated and strictly
applied by al-Mutawakkil, which reflects in most of its clauses the regulations listed in Shuru4t Umar.4 From that time onwards, in fact, this code
of rules was applied by several rulers, and was the norm by the second
half of the ninth century.
restrictions upon the DHIMMI 3s prior

to al-mutawakkil
Dionysius of Tell-Mahre4, the Jacobite Patriarch in the first half of the
ninth century, reports that as early as the days of the caliph Uthma4n
(r.64456) the governor of Damascus ordered the extirpation of crosses
and forbade the public exhibition of the cross during Christian festivals and supplications.5 It is reported that Abd al-Malik b. Marwa4n
(r. 685705) ordered the slaughter of all the pigs in the caliphate and
the removal of all crosses.6 His decree regarding the Arabization of the
diwa4ns seems to have also entailed an attempt to curtail the employment
of non-Muslims in the administration.7 Some time during the end of the
101
seventh century or the beginning of the eighth, rulers in Egypt started

demanding payment of jizya from monks, who had till then been exempt
from this payment.8
As noted above, the first caliph to have issued an edict containing a
set of regulations regarding the dhimm s was Umar II b. Abd al-Az3z
(above, Chapter 3). The main part of this corpus of regulations focused
on the ghiya4r. In addition, there was an order regarding the dismissal of
non-Muslims from public service. This may have been a clause in this
same edict, or may have been published separately. This is in fact the
earliest evidence in existence regarding the enforcement of the principle
of khila4f or ghiya4r.
Several sources mention that Umar b. Abd al-Az3z also promulgated
orders regarding the exhibition of crosses in public9 and the use of the
na4qu4s.10 Michael the Syrian also mentions the prohibition on raising
voices in prayer.11 There are conflicting reports in Muslim sources concerning his attitude towards dhimm prayer-houses. Some sources report
that he issued an order prohibiting the destruction of churches and the
building of new ones.12 On the other hand, Abd al-Razza4q al-Sana4n3
insists in more than one place that he ordered the destruction of ancient
as well as newly built churches in amsa4r al-muslim n, or those that were
taken anwatan.13 Nevertheless, there is information concerning the
building of several new churches in the days of Umar b. Abd al-Az3z
(r. 71720).14 There is also conflicting evidence related to Umar b. Abd
al-Az3z regarding the question of the right to bequeath to the church.15
The case for accepting these traditions as historical was set out in the
previous chapter. Here it may be added that the tradition adduced by
Abu4 Yu4suf regarding the ghiya4r emphasizes the fact that this was meant
to be strictly enforced. It was transmitted by Abd al-Rahma4n b. Tha4bit b.
Thu4ba4n, a well-known Damascene transmitter of his own day, an ascetic
(za4hid),16 who heard it from his father:17
Umar b. Abd al-Az3z wrote to one of his governors: Regarding the
matter at hand:you shall not permit a cross to be manifested, that is
not smashed or effaced; a Jew or a Christian shall not ride on a saddle
(sirj), but shall ride on a pack-saddle (ika4f); their women shall not ride
on leather saddles (riha4la), they shall ride on a pack-saddle (ika4f). Order
this expressly, and prevent those who are under your authority [from
letting] a Christian wear a qaba4, a silk garment, or a turban (asb).
I have been told that many of the Christians under your authority have
returned to wearing turbans (ama4im), have given up wearing the
girdles (mana4tiq)18 on their waists, and have begun to wear their hair
long and to neglect cutting it [i.e. their forelocks:M.L.R.]. I swear that
102

if anyone under your authority does so, this attests to your weakness,
inability, and flattery, and when they go back to this [i.e. their former
costumes and habits:M.L.R.] they know what you are. Look out for
everything which I have prohibited and prevent it from being carried
out. Goodbye.
The fact that Umar b. Abd al-Az3z protests that the governors have
allowed matters to deteriorate could be taken to mean that the rules predated his reign, but he may equally well be referring to an edict that he
himself had published in the past.
It seems therefore, that the idea of ghiya4r struck deep roots in the
Muslim domain as early as the second/eighth century, and by the end of
the eighth century CE, at the time when Abu4 Yu4suf was writing his Kita4b
al-khara4j, had become an accepted concept which, although not always
rigorously enforced, was nevertheless considered an official code of dress
and appearance to which non-Muslims were required to adhere.19 The
gradual progression of the enforcement of the concept of ghiya4r specifically, and other restrictions in general (see below), may thus be traced to
the eighth century CE and onwards.
Between the time of Umar b. Abd al-Az3z and that of al-Mutawakkil
there are several mentions of restrictions and regulations, most of which
reappear later in the Shuru4t. Thus, several restrictions were prescribed
by the caliph al-Mansu4r (r. 75475) at the beginning of the Abba4sid
period. These include prohibitions on the employment of Christians in
public office, on holding vigils for liturgical purposes, and on teaching
in Arabic.20 Al-Mansu4r also removed the crosses from the tops of the
churches,21 ordered that the palms of the dhimm s be marked (a claim
not attested elsewhere as far as I know),22 and imposed the jizya on the
monks, who had been exempt from it up to that time.23 The order to
remove the crosses existed prior to Umar b. Abd al-Az3z, and dates
to the seventh century,24 while the first systematic attempt to oust nonMuslims from public office was carried out by Abd al-Malik,25 and the
second by Umar b. Abd al-Az3z.26 Similar prohibitions on raising voices
in prayer (in this case vigils) and on teaching in Arabic would reappear
later in the Shuru4t.
Such restrictive policies were applied about the same time in Egypt.
Sa4lih b. Al3 (governor of Egypt 1334/7501; 1367/7535) prohibited
public theological debates between Christians and Muslims, the exhibition of crosses in public, and the building of new churches.27
Ha4ru4n al-Rash3d ordered that churches in the frontier areas be razed
to the ground, an order which no doubt had to do with the security
103
situation along the border. As already noted above, particular instructions concerning dhimm appearance, a clear development of Umar b.
Abd al-Az3zs edict, were articulated by al-Rash3ds legal adviser, Abu4
Yu4suf Yaqu4b (d. 789). Al-Tabar3 mentions that al-Rash3d ordered that
the dhimm s in Baghdad change their appearance in order to differ from
the Muslims.28 However, Abu4 Yu4sufs edict is a general one and applies
to all non-Muslims; there is thus good reason to presume that at the very
least it was enforced wherever there was a significant Muslim presence.29
Between Ha4ru4n al-Rash3ds time and that of al-Mutawakkil there is a
report concerning the caliph al-Wa4thiq (r. 8427), who prohibited the use
of the na4qu4s in churches.30
The above evidence seems to indicate that although there was no single document, nor a single consistent, accepted, and comprehensive set of
regulations regarding non-Muslims, many of the regulations themselves
were starting to take shape and were enforced at least, but not only, in the
above-mentioned cases which were recorded by historians and preserved.
Shuru4t Umar (and the competing documents) constituted therefore an
attempt to give one formalized and uniform expression to a host of variegated regulations which were applied sporadically under different rulers.
the restrictions issued by al-mutawakkil

The first established set of restrictions known to us is, of course, that
which was promulgated by al-Mutawakkil (r. 84761). For the first time
the caliph issued an organized set of restrictions to be applied to the
dhimm s.
This decree included the requirement to wear yellow as a distinguishing colour (see details below); to wear the zunna4r; to ride on saddles with
wooden stirrups affixed to two pommels at the rear; to attach two buttons to the qalansuwa, and to attach yellow patches, front and back, on
the clothes of slaves and the lower-class dhimm s; to destroy new prayerhouses (biyaahum al-muhdatha); to seize a tenth of their residences; to
nail wooden images of devils (shaya4t n) to their doors; to prohibit the
employment of non-Muslims in the government; to prohibit their children from studying in Muslim schools or to be taught by Muslims; to
prohibit public processions including on Palm Sunday, and to level graves
that resemble those of Muslims.31 In contrast to the opinion that this was
an isolated case, I will proceed to demonstrate here that these restrictions
in general, and the ghiya4r in particular, were issued again and again in the
following period. In fact, once clearly established, this set seems to have
104
become a paradigm for later caliphs and rulers to follow. Yet, although
the principle was now unambiguous, and its enforcement attempted by
many a ruler, it nevertheless needed to be incessantly reiterated, since the
non-Muslims continuously attempted to ignore these humiliating restrictions, and seem to have taken every opportunity to evade these distinguishing marks and to adopt the respectable attire and paraphernalia of
the Muslims.
The sources mention several caliphs who issued similar edicts.
Al-Muqtadir (r. 90832) is reported to have issued a set of regulations
concerning the employment of dhimm s in public service, the distinctive honey-coloured attire, and other ghiya4r.32 Al-Maqr3z3 reports that
Jawhar, the waz r of the Fa4timid caliph al-Muizz (r. 95375), imposed
the regulations of the ghiya4r upon the dhimm s.33 Most famous of all is,
no doubt, al-Ha4kim, who, it should be admitted, went to much greater
lengths, not only inflicting the restrictions upon the dhimm s ruthlessly
and mercilessly, but also demolishing all synagogues and churches, and
confiscating property.34 The question is, of course, whether these were
indeed enforced, or whether they were merely a dead letter.
Actual Enforcement of Issued Restrictions
The decrees issued by al-Mutawakkil are well known, and there is no doubt
that this caliph did indeed issue a detailed decree regarding the dhimms.
Yet information concerning the enforcement of the decree is partial and
insufficient. What evidence do we have for its enforcement then?
On the Muslim side, al-Tabar3 reports that in Muharram 239 (12
June11 July 853 CE) al-Mutawakkil ordered that dhimm s affix two
yellow sleeves (dhira4ayn) to their outer cloaks. This does not exactly follow the edict as cited by al-Tabar3 himself earlier, according to which the
more affluent of the non-Muslims, including the merchants and the secretaries, were to wear yellow hoods (taya4lisa), yellow turbans (ama4im),
and their women yellow mantles (iza4r), while those of their humble followers beneath these in station, whose circumstances prevents them from
wearing hoods should affix two yellow patches to their cloaks, front and
back.35 Moreover, in the preceding description of the edict the yellow
patches serve as the distinguishing sign for the slaves (mama4l k),36 while
nothing is said about sleeves. In Safar (12 July9 August) of the same
year he ordered that they restrict their mounts to mules and donkeys, and
avoid riding and pack horses.37 Ibn al-Jawz3 recounts that in 236/8501,
following the general edict issued in Shawwa4l 235/850,38 Christians were
105
ousted from public office; they were also discharged from the wila4ya4t,
and were in general no longer to be employed in anything related to
the affairs of Muslims.39 In agreement with al-Tabar3 he relates that in
Muharram 239/853 the order that non-Muslim men should wear honeycoloured patches on their gowns and overcoats, and that women should
wear honey-coloured veils, was enforced; in Safar that year the dhimm s
were prohibited from riding horses and were restricted to using donkeys
and mules.40 In 240/8545 it was announced in public that the children
of the dhimm s were to be taught Syriac or Hebrew, and were forbidden
to learn Arabic.41
Severus b. al-Muqaffa reports that al-Mutawakkil ordered all churches
to be demolished, a claim which is not corroborated elsewhere; forbade
dhimm s to wear white, and ordered that they should wear only dyed
garments so that they might be distinguished among the Muslims (this
evidence relates to the issue of differentiation through colour, though it
does not name a specific colour, but does not mention any of the other
items of the ghiya4r); commanded that frightful pictures should be made
on wooden boards and be nailed over the doors of the Christians; and
that dhimm s should not serve in the government of the sulta4n. Severus
also notes that he forced many to convert to Islam, a fact not mentioned
elsewhere.42 The Jacobite chronicler Gregory Bar Hebraeus mentions the
new requirements concerning appearance, the prohibition on exhibiting
crosses in processions on Palm Sunday, the destruction of new churches,
and the appropriation of partial areas of large churches. He also notes
that similar restrictions were imposed upon the Jews.43 These two last
sources are general and brief, and it is difficult therefore to corroborate
them with the specific details of al-Tabar3.
A chronicle written in Samaria at the time of the events by members of
the local Samaritan community provides substantial new evidence on this
issue.44 This is a continuation of the Samaritan chronicle of Abu4 l-Fath.
It appears in a unique manuscript found in the Bibliothque Nationale
and was known to Vilmar, the first editor of Abu4 l-Faths chronicle. Due
mainly to linguistic considerations, Vilmar chose not to publish this manuscript, which continued up to the time of the caliph al-Ra4d3 (r. 934
40), and so this part of the chronicle has until now been disregarded, in
spite of its importance as a well-informed source based on an eyewitness
account. According to this chronicle al-Wa4thiq45
was succeeded by his brother Jafar [al-Mutawakkil], who afflicted the
world with every kind [of affliction];46 he ordered at first that people
106

should wear distinguishing clothes, except for the black and the blue,
which he reserved for his faith. He ordered that there should be no
scribe or public official (a4mil) except from his faith, and that there
should be no one in charge of a fort or holding any kind of position
except from his faith. There were Christians whom he cast out, and he
appointed all the officials from his faith. He ordered that no one should
wear a garment with an embroidered edge (tira4z)47 except the members
of his faith (ahl millatihi), and no one should ride a horse [except the
members of his faith]. He commanded that every dhimm should wear a
distinguishing sign in front and back, and that he should not sit in front
on a velvet-like sofa,48 that no one except the members of his own faith
should have iron stirrups49 the rest [would have] wooden ones. He
ordered that every grave resembling the graves of the members of his
faith be destroyed,50 and the grave of the ras Nethanel51 was destroyed.
Before that occurred, he ordered that every dhimm should afix to his
door a wooden idol bearing the label of idol (wathan)52 (240).
The Samaritans who resided in Nablus, may God remember them
favourably, having presented the governor (wa4l) of Nablus with something [i.e. a gift], asked him to grant them a delay so they could go down
to Ramla, and he agreed to that. [Now] in Da4ju4n53 there was a man
possessed of dignity and power, (242) whose word was accepted by the
ruler (sulta4n), by the name of Abu4 Yu4suf b. Dha4s3, may his memory be
forever blessed. He called on the governor and petitioned him, and he
[the governor] told him that it was not possible to annul the order of the
king but [said]:Choose for yourself an image which is not offensive.
He chose the image of a candelabrum that we make; it was put in an
envelope, and he stamped it and sent it to the governor of Nablus. [The
governor] commanded that a Samaritan should only make [an image]
like that which Yu4suf b. Adha4s3 made [that is] a candelabrum. They
rejoiced greatly in this and profusely thanked God, may He be praised
and exalted. As for those [Samaritans] who were in [the province of]
Jordan this [concession] was not granted to them, and an image was
made [by them] like the other peoples according to the law.
In his days it was decreed that a man shall not raise his voice in prayer,
and shall not raise voice in ,54 a funeral shall not be seen, and a
dhimm3 shall not lift his face to a Muslims55 face in order to speak or
respond to him.
Unlike the other non-Muslim sources cited, the Continuatio gives us a

nearly complete list of al-Mutawakkils restrictions. Many of the restrictions overlap: thus the Continuatio mentions the requirements of the
ghiya4r, and specifically the patches in front and rear; the prohibition on
the use of iron stirrups; the prohibition on riding horses (which is mentioned by al-Tabar3,56 but not in the actual edict as cited by him); the
prohibition on holding public office; the prohibition on holding public
funeral processions; the order to level all dhimm graves so that they
107
do not resemble Muslim ones; the order to affix idols to the doorposts.
Missing from the description are the destruction of new prayer-houses,
the prohibition on studying in Muslim schools, and the seizing of a tenth
of their residences. It may be that the latter were not mentioned because
they may not have been applied or relevant in Samaria, or there is always
a possibility that they were overlooked by the writer.
On the other hand, the Continuatio includes many additions not
found elsewhere. Although the items included in the ghiya4r (such as
zunna4r, yellow colour, shoes, headgear) are not specified and seem to
be self-understood by the writer, the chronicle mentions that black and
blue were reserved for the Muslims only, as was the embroidered edge of
the cloak the tira4z. In addition, there are restrictions mentioned which
are a part of the Shuru4t, including the prohibition on raising the voice in
prayer, on sitting in the respectable seats (=we shall rise from our seats
when they wish to sit), and the requirement to talk to Muslims with eyes
cast down (=we shall show respect toward the Muslim).57
The report given here concerning the restrictions imposed upon the
dhimm s during the days of al-Mutawakkil is very detailed and of considerable significance. The Samaritan text confirms unequivocally that not
only were dhimm s in provinces such as Palestine and Jordan familiar in
detail with these decrees, but that the restrictions were indeed enforced
quite strictly. This is well in line with al-Mutawakkils letter cited by
al-Tabar3 and in Shuru4t al-nasa4ra,58 which demands that what they do
shall be inspected to ensure that the orders of the Commander of the
Faithful are carried out by their clear compliance. The inspector should
be able to spot compliance readily, it being immediately apparent You
shall instruct the officers concerning the orders of the Commander of the
Faithful, and do so in such a way that they are motivated to carry out
their examinations as commissioned.59
This, incidentally, is well in line with the edict of Umar b. Abd al-Az3z
to his governor which ends with the directive that he shall watch out for
anything that I have prohibited and stop those who commit it.60
This enforcement of the orders is well demonstrated by the practical
consequences related in the Samaritan chronicle such as the story of the
image that the dhimm s had to attach to their doorposts. The Samaritans
of Palestine went to great trouble to evade this order, which in their eyes
was equivalent to idol worship; nonetheless, they had to settle for a
compromise the use of an image of a candelabrum. The Samaritans of
Jund al-Urdunn were not even granted this concession by their governor,
and had to abide by the original decree. Another example of the strict
execution of the decrees is the levelling of the grave of the Samaritan
108
head Netanel because it resembled a Muslim grave. It may be presumed

that these decrees were imposed with equal severity upon all the other
dhimm s in the junds of Fila4st3n and Urdunn as well. Moreover, it can
be deduced safely that this would have been the situation all over the
caliphate. It seems that it can no longer be claimed that al-Mutawakkils
regulations were not fully enforced.
the long-lasting enforcement of

al-mutawakkils restrictions
It nevertheless remains to be asked whether this policy had any real effective long-lasting influence after al-Mutawakkils days. It has been noted
above that there are succinct references to similar restrictions imposed
by al-Muqtadir, by al-Ikhsh3d,61 and by al-Muizz, before we come to
al-Ha4kims notorious decrees. Yet these were not regarded seriously by
scholars, as already emphasized. Fattal, for example, adopted Ar3b b.
Sad al-Qurtub3s62 evaluation that al-Muqtadirs restrictions concerning
the prohibition on employing dhimm s in government did not last, and
applied this evaluation to the whole set of prohibitions. This cannot actually be deduced, as al-Qurtub3 refers specifically to this one decree which
was not adhered to in Cordova, while Ibn Taghr3 Bird3, who refers not
only to this prohibition, but also to the elements of ghiya4r, does not state
that these prohibitions were not carried out in practice.63 Thus, although
Fattal rightly emphasizes the fact that the prohibition on employment
was not adhered to, this relates to a specific and very problematic issue. It
is well known that the prohibition concerning the service of dhimm s in
public office was the most difficult to enforce. This was due not only to
the fact that Christians and Jews had so much experience and knowledge
in the field of administration and management that they had become
almost irreplaceable, but probably also to the reluctance of the rulers to
replace these loyal and efficient officials who, in contrast to their Muslim
counterparts, posed no threat to their rule. This is well documented by
Ibn Qayyim al-Jawziyya, Ibn al-Naqqa4sh, and al-Qalqashand3, who focus
almost obsessively on this issue.64
Here, too, the Samaritan chronicle supplies us with material evidence
that proves wrong the accepted opinion that al-Mutawakkils decrees
were but a short-lived episode characteristic of their initiator, and that
attempts made by other rulers to impose such regulations were not
put into effect, and ended in total failure. It gives us new information
which shows that al-Mutawakkils decrees were in fact a turning-point.
109
Although his restrictions had not automatically stayed in force, they were
renewed in an amazingly short time.
The information concerns Ahmad b. Tu4lu4n, the founder of the Tu4lu4nid
dynasty, who ruled Palestine between 878 and 884:65
He [i.e. Ibn Tu4lu4n] oppressed the people in every way. In the second
year a governor (wa4l) came to [rule over] the people on his behalf and
oppressed [them] in every way; he ordered that the dhimms should
wear distinguishing signs (ghiya4r), engraved [lit. made] idols (awtha4n)
on their doors, [ordered that] a dhimm should not raise his head in the
presence of a Muslim [lit. goy] and that he should not raise his voice in
prayer; and that he should not blow the horn; he also destroyed a synagogue of the Jews. All the religious communities were in fear of him, lest
he extend [his] hand to their houses of worship so as to put them to his
own use. He prohibited the drinking of wine66 and oppressed [them] in
every possible manner.
Although more succinct than its predecessor, this set of regulations is

almost identical to that of al-Mutawakkil, the distinctions in appearance
of the dhimm s being encapsulated under the title ghiya4r. There are two
regulations missing: one, concerning the funerals, probably just due to
carelessness of the author since its parallels, such as blowing the horn
or raising the voice in prayer, are present; the other the order concerning the levelling of the graves probably because after al-Mutawakkils
actions Samaritan graves no longer resembled the Muslim graves, but
were built without tall tombstones in the first place. An additional
decree appearing here is the prohibition on drinking fermented drinks,
which was not listed among al-Mutawakkils restrictions. This prohibition, rooted in the edict of Umar b. Abd al-Az3z, and prohibiting the
Muslims from drinking wine,67 may have resulted in a prohibition on
the possession of wine in Muslim cities, as is claimed by Theophanes.68
In al-Sha4fi3s Kita4b al-umm there is a prohibition on selling Muslims fermented drinks,69 while in Shuru4t Umar the prohibition seems to be on
its sale altogether.70 Thus, it may well be that by Ahmad b. Tu4lu4ns time
this ban on wine was already being imposed not only upon Muslims
and non-Muslims in the amsa4r, but even upon non-Muslims living in the
more neglected agricultural periphery of the caliphate.
There seems to be no other evidence regarding Ibn Tu4lu4ns restrictions.
In fact, his image in the sources is a positive one,71 and were it not for
this one source there would have been no reason to suspect this. The text,
written by members of the local community, clearly shows, however, that
these measures were imposed upon the population and strictly enforced.
110
It can be safely assumed that they were enforced in the same manner in
all provinces under Ibn Tu4lu4ns jurisdiction. Ibn Tu4lu4ns restrictions are
therefore a case in point which demonstrates that one cannot deduce
from the silence of the sources that the restrictions were not applied.72
In fact, although it is not part of the current discussion, it should be
noted that the chronicle gives a detailed description of Ibn Tu4lu4ns rule,
especially in Palestine, the author complaining not only about the restrictions but about the oppressive behaviour of his emissaries in general.73
The new evidence of the Continuatio of the Samaritan Chronicle,
provided by dhimm s living in a provincial area of the caliphate, demonstrates that not only were al-Mutawakkils decrees not forgotten,
but that they were strictly enforced by another Muslim ruler shortly
after al-Mutawakkils death. This leads us in a rather different direction than the one taken until now. The additional references we have
to al-Muqtadir (r. 90832), al-Ikhsh3d (r. 934), al-Muizz (r. 95375),74
al-Ha4kim (9961020), al-Mustansir (r. 1086) in Egypt,75 and al-Muqtad3
(r. 1091),76 and the Seljuk sultan Mahmu4d in Baghdad in 112177 who
also enforced similar decrees should thus be regarded carefully, rather
than being waved away and discarded nonchalantly. While the information about the first three is minimal, the evidence in the rest of the cases
is detailed and impressive. Following the descriptions of the enforcement of al-Mutawakkil and Ibn Tu4lu4ns decrees, there is no reason to
suspect the more succinct references. There is in fact no need to continue and review the history of the enforcement, as this has been done
quite thoroughly by Tritton and Fattal.78 It seems therefore that from
al-Mutawakkils days onwards, the regulations published by him were
to become a norm which the caliphs and other rulers within their orbit
strove to impose and enforce.
Occasional mentions demonstrate that some of these restrictions were
considered an accepted norm by the end of the ninth and during the
tenth centuries. Thus, Elias of Nisibis reports that in the year 271/8845
the people of Baghdad rioted against the Christians because they rode
on horses.79 Al-Muqaddas3, in the second half of the tenth century, notes
that in Sh3ra4z you will not see Magians with ghiya4r and the wearer of
the taylasa4n has no grandeur. I have seen wearers of the taylasa4n drunk,
and even beggars and Christians wearing it.80 This last piece of evidence carries a double message: al-Muqaddas3 obviously considered
this scandalous behaviour: Magians should definitely wear ghiya4r, and
Christians should not be wearing the taylasa4n; in Sh3ra4z, however, this is
not the accepted norm. It can be deduced here that what was already an
111
unquestioned norm in Egypt, Syria, and Mesopotamia was not quite so

in the small provincial town of Sh3ra4z, which at that time was only beginning to gain importance.
Another issue brought to light by the evidence is that although the
term ghiya4r and the principal demands included in it such as the zunna4r,
the patches, and the prohibition on riding horses were well understood,
there were nevertheless provincial variations, as well as changes concerning details such as colours that were forbidden to the dhimm s an issue
not mentioned in al-Mutawakkils edict or the Shuru4t (white in Egypt,
black and blue in Palestine; shoes with or without straps (above), or the
requirement to wear two different shoes;81 the demand to hang a medallion with the word dhimm around the neck,82 a bell, or a figure in the
form of a cross for a Christian or a calf for a Jew.83 These variations
support the verisimilitude of independent enforcement of the ghiya4r in
various regions.
However, although enforcement of the ghiya4r thus seems to have been
more significant than has been traditionally assumed, this does not mean
that from al-Mutawakkils days onwards these regulations were an integral and non-negotiable part of dhimm life. It is quite understandable
that dhimm s felt more restricted and humiliated than they had been
before, and that they fought against their new situations, testing and trying the determination of each ruler to enforce the restrictions now and
again. It may also be presumed that some rulers were indeed more lenient
than others, especially when it suited their internal or external political
ends. There are ample examples of allowances and concessions concerning the building of prayer-houses, the employment of dhimm s in government bureaux84 etc. after al-Mutawakkils days.85
Nevertheless, though the rules were often bent in favour of a more lenient
policy, and these regulations were often enough disregarded and evaded,
they were never annulled, and could be imposed or enforced strictly at any
moment. At the rulers will, their enforcement could also be retracted. This
is well exemplified by the behaviour of both al-Ha kim
4
and Sala4
h al-Dn,
3
86
who first imposed these regulations and then retracted them.
The evidence thus shows that from al-Mutawakkils time onwards
these regulations had become the rule in the lands controlled by the
caliphs in al-Sha4m, in Iraq, and in Egypt and its dependencies. What happened in other parts of the empire is not clear. Al-Muqaddas3s reference
to Sh3ra4z above may indicate that in areas that were further out the process occurred at a much slower pace. The Maghrib and Spain may also
have adopted these restrictions late in the day.87
112
The set of regulations which began with the ideology promoted and
applied by Umar b. Abd al-Az3z, and struck deep roots from the second
half of the ninth century onwards, continued in fact to expand, becoming
more and more elaborate and more strictly enforced with time, as can be
seen later on in the Mamlu4k period, when Ibn Qayyim al-Jawziyya wrote
his opus magnum, Ahka4m ahl al-dhimma.
5
The Provenance of the Modes
of Subordination of Non-Muslims
The Muslim conquest entailed the subjugation of large populations

throughout the conquered territories. As we have seen in the first chapter,the Muslims were at first generally content with the submission of the
local inhabitants and their agreement to pay tax, and in return allowed
them in most cases to continue their lives as before.1 However, as time
went by the growing presence of Muslims in cities that had previously
been occupied solely by non-Muslims necessitated certain changes and
adaptations in those cities where Muslims and non-Muslims lived side
by side. This led, as we have seen, to the drawing up of Shuru4t Umar,
a universal document which gradually replaced the former diversified
treaties.
While the question of the enforcement of the Shuru4t has been widely
discussed, the question of its provenance has been neglected. The aim of
the present chapter is to try and trace the origins of the various components of Shuru4t Umar. The principal question is therefore:what were the
sources of the various elements of this document that defined the position
of non-Muslims in this newly emerging society? My assumption is that
an examination of the origin of the notions, concepts, and terms of which
this document is made up will lead to a better understanding of its social
significance and goals.
The Pact of Umar was probably drawn up somewhere towards the
year 800 CE.2 In the period following the conquest, throughout the seventh and eighth centuries, the Muslims underwent a long process of adaptation to their new status as rulers. This process was accompanied by a
See translation of Shuru4t Umar in Appendix I.
113
114
profound process of acculturation which dictated the structure and nature

of the new society that was now being shaped. At the time of the forging
of the Shurut4 this process was well under way, as is clear in the Shurut4
themselves. As we shall presently attempt to demonstrate, many of the elements of the Shurut4 originated in notions and concepts that were adopted
by the Muslims from the cultures of the conquered societies and integrated
into Muslim culture and legislation. Thus, among many other things, the
Muslims adopted certain social concepts that, as shall be shown, were
to have great significance regarding the place that was allocated to nonMuslims in the new social order that was being created.
There has not been much discussion of the social significance of
Shuru4t Umar. Most probably as a result of the views expressed by
the classical jurists, it seems to have been taken for granted that the
Muslims wanted to impose their rule and subjugate the non-Muslims.
Among modern scholars, only Albrecht Noth has raised the question
of the purpose of the Shuru4t.3 It was he who asked whether its purpose was, in fact, to humiliate the non-Muslims. After an exhaustive
analysis of the document, he concluded that its intention was not to
humiliate, but rather to differentiate between Muslims and non-Muslims. He reasoned that the status of the Muslim conquerors as a small
minority among the conquered population caused a need for a means
of differentiation between the two groups. Ghiya4r,4 or distinguishing
signs, is indeed the term used in Muslim literature with reference to
the demand made upon non-Muslims regarding the need to distinguish
their appearance from that of Muslims. The issue of the cultural origins
of the ghiya4r, a central element in Shuru4t Umar, and its significance,
will be extensively discussed in this chapter.
What then were the origins or the sources of inspiration for creating such a document? As shown, the initial surrender agreements made
between the Muslims and the conquered populations were designed
after a well-known model that had existed for many centuries before
the Muslim conquest. Were there also parallel documents which imposed
restrictions on minorities or other distinct groups in the societies of the
Near East before the Muslim conquest?
the status of minorities in ancient

near eastern societies
In an attempt to discover the origins of the notions behind the Shuru4t,
my initial research focused on previous legislation defining the status of
The Provenance of the Modes of Subordination of Non-Muslims 115

minorities in ancient Near Eastern societies.5 A general survey shows that
all aliens in ancient Near Eastern and Classical societies had to acquire
the fundamental status of resident alien, which granted them their basic
protection. Thus we have the Akkadian uba4rum, the biblical ger,6 the
Athenian metoikos, the Roman peregrinus, or ordo libertinorum,7 and
the Arabian ja4r,8 mawla4 (client), and hal f (guest ally).9 Resident aliens
were subject to many restrictions. Thus, in the ancient Near East there
seem to have been strict limitations on the ownership of land,10 and resident aliens were not entitled to the abolition of debts or to the abolition
of servitude as a result of debts.11 In general, a resident alien was considered of inferior status and, according to a common proverb, a resident
alien in another city is a slave.12
More is known about the status of the metoikoi13 in the Classical
world. They were prohibited from marrying a citizen (epigamia) under
penalty to both sides; they were prohibited from owning land, and even
from allowing their cattle to graze on public land; they were also barred
from holding public office, or from giving public speeches and taking
part in various contests. They had to pay a special tax and to serve in the
army as hoplites, but were barred from serving in the cavalry, which often
participated in important civil events such as festivals and processions, as
well as in athletic games.14
Similarly, in the Roman world resident aliens were prohibited from
marrying Roman citizens until the time of Augustus (conubium), from
holding public office, and from serving as sacerdotes or in the Senate.
They were judged according to the Law of Nations (ius gentium), which
pertained to both citizens and non-citizens and was regarded as the
ius naturale, and from the end of the third century BCE resident aliens
had their own praetor (praetor peregrinus).15 However, the Constitutio
Antoniana, published by Caracalla in 211 CE, accorded citizenship to all
inhabitants of the empire, thus abolishing any differences and restrictions
that had existed previously.
It must be noted that while in general there were various regulations and
restrictions on individual aliens in the ancient Near East, groups of the same
ethnic and cultural background, or engaged in the same occupation, might
be awarded a different status, or at least special rights. The Old Assyrian
trading colonies in Anatolia were granted extra-territorial status.16 In the
Graeco-Roman and Hellenistic worlds there were cases where groups of
resident aliens, especially those involved in international commerce, were
allowed to have their own courts and to solve legal issues according to
their own law and customs, under the jurisdiction of foreign magistrates.17
116
In addition, as noted above, surrender agreements in the ancient world

often included a clause allowing the conquered to live according to their
ancestral customs (kata patrious nomous), continuing their lives to a great
extent as before, under the hegemony of the new ruler.18
the status of minorities in the byzantine

and sasanian empires
Especially relevant to the issue under discussion here is the status of
minority groups under RomanByzantine and Sasanian rule.
Jews in Byzantine Society
The most ancient and distinctive minority group living under Roman
Byzantine rule, and one about which we are comparatively well informed,
is the Jewish community.
The principle of a communitys right to live according to its ancestral
customs stood at the basis of the autonomous rule of Jewish communities in the Hellenistic world.19 The authority of ancient custom in Roman
tradition (the mos maiorum, ex consuetudine, vetus mos et consuetudo)20
was, in its turn, the principle that guided the Roman empire in regard
to the Jewish communities when it continued to preserve the rights that
were given to them initially in 63 BCE following the conquest of Syria
and Palestine by the Romans.
On the basis of this, Jewish communities under Roman rule were given
wide-ranging autonomy. This included judicial autonomy headed by a
Patriarch; the right to collect taxes; the right to excommunicate members
of the community; the exemption from duties entailing the profanation
of the Sabbath and holidays; the special privilege of circumcising their
children (an act that was generally prohibited throughout the empire, as
it was considered mutilation); and more.21 In fact, this autonomous status
at times conferred upon the Jews privileges and exemptions in preference
over other citizens of the empire.22
Byzantine law, representing the Christian Roman empire, added a new
dimension. Although Judaism remained what Tertullian called a religio
licita,23 new laws were gradually added, aimed at restricting the Jewish
communitys autonomy and rights, and segregating them from Christian
society. Thus, in a growing number of issues, Jews, as well as Samaritans,
pagans, and so-called heretics, were discriminated against by law, due to
their religious identity:

1. Marriages between Christians and Jews were completely prohibited.24 This clearly originated on religious grounds, and
was based on canonical prohibitions.25 Thus, while the ancient
prohibition of intermarriage between citizens and aliens26 was no
doubt already much eroded (as by that time, needless to mention,
all the inhabitants of the provinces were citizens of the empire),
its place was now taken by the prohibition on interreligious
marriages.27
2. A law passed in 415 prohibited the building of new synagogues,
and ordered the destruction of abandoned ones, thus preventing
the expansion of the Jewish presence.28 Existing synagogues were
to remain in their present state.29 It was permitted, however, to
make repairs to prevent immediate ruin.30
3. Ownership by Jews of Christian, non-Jewish slaves, and slaves who
had converted to Christianity was prohibited, and their manumission was ordered under penalty.31
4. In 415 the authority of the Jewish Patriarch was restricted:he was
prohibited from serving as judge in legal cases between Jews and
Christians, and allowed to sit only in civil cases.32
5. A total ban was issued on conversion to Judaism, under heavy
penalties.33 Furthermore, Jews were warned against attempting to
dissuade their co-religionists from converting to Christianity.34
6. Jews were banned from holding public office, and even from the
law profession, although it seems that these prohibitions were not
efficiently enforced.35
7. Non-Christians were prohibited from giving testimony against
Christians unless asked to do so by the state.36
8. Jews could not disinherit their children if they converted to
Christianity.37 Christian relatives were in fact to be given precedence over non-Christians.38
9. Jews were warned against contempt of Christianity and its cult.39
It could thus be said that in certain areas Jews, Samaritans, and other
non-Christians were treated as second-class citizens, and their status
was similar in several aspects to that previously accorded to resident
aliens (e.g. prohibition of mixed marriages, restrictions regarding the
ownership of property, the holding of public offices and functions, and
the ownership of slaves). Nevertheless, the Christian emperors still had
a clear obligation to protect the Jewish community, and the Jewish
community had the right to live according to its ancient laws and
118
customs. Until 425 the community had its own Patriarch, who carried
the dignified title Illustris and was protected by law from insults and
threats.40 There is no direct evidence that the Byzantine authorities
had any part in the annulment of this office. After the cessation of the
Patriarchate, Jewish institutions continued to function as before.41 Jews
and Jewish synagogues, which were under growing threat, were protected by law from any harm or damage.42 Jews were not to be forced
to attend court on the Sabbath, due to vetus mos and consuetudo.43 The
conversion of Jews to Christianity, which was generally encouraged,
was prohibited if prompted by debts or other penalties decided by the
Jewish courts.44 There was a prohibition on outsiders fixing the prices
of Jewish merchandise.45
Was this protection a consequence of the fact that Jews were still citizens of the empire, or that their inferior status, similar to that of resident
aliens, now enjoined their protection by the authorities?46 It may have
been a bit of both. The law regarding the protection of synagogues from
vandalism expressly notes that no one shall be destroyed for being a
Jew and that even if someone is entangled in his crimes, the vigour of
the courts and the protection of public law (vigor iurisque publici tutela)
appear to have been instituted in our midst for that very reason, that no
one shall have the power to permit himself to take vengeance.47
Clearly, under Byzantine rule being part of the Jewish community
had turned from an advantage to a disadvantage, its members having
become second-class citizens. Nevertheless, it can be surmised from the
laws themselves that Jews still held public office and served as lawyers as
late as Justinians days, and probably later, and thus maintained, at least
partially, their former social status.48
Non-Zoroastrians in Sasanian Society
Like Byzantine society, Sasanian Persia was an empire with a declared
official religion. Society was divided into Zoroastrians and nonZoroastrians. Additionally, the ideal society is presented in Sasanian
literature, and in later sources preserving material from the Sasanian
period, as a rigid class-based society. Adapting an ancient Avestan concept, these sources depict a society whose citizens belonged to four different classes: (a) the priests; (b) the warriors; (c) the scribes; and (d)
the peasants, artisans, and tradesmen. Class membership was inherited.49
As in the Near Eastern societies of ancient times, non-citizens resident

aliens, temporary visitors, people who were banished from the community, illegitimate children, and slaves were devoid of any civic rights. 50
What was the status of non-Zoroastrians under Sasanian rule? We are
not as well informed about it, since, unlike the Romans, the Sasanians
apparently did not compile legal codes; or if they did, they have not
survived.51
There is, however, one Sasanian law-book, The Book of a Thousand
Judgments, which provides us with some information. According to this
book, a Zoroastrian who had left his religion was cast out from his family
and his community, and lost all his family and community privileges. In
spite of the prominence of the state religion in the Sasanian empire, he
did not, however, lose his basic rights:he was still a member of the civic
community,52 and contractual obligations as well as personal property
remained valid after his conversion.53 In addition, a non-Zoroastrian
was not allowed to own a Zoroastrian slave, and a non-Zoroastrian
slave who converted to Zoroastrianism had to be manumitted, as did a
slave who converted to Christianity in the Byzantine empire.54 However,
according to Sasanian law the slave had to compensate his master for
the economic loss involved.55 If a Zoroastrian was sold into slavery to a
non-Zoroastrian, the transaction was considered a theft, they were both
branded, and the buyer was not reimbursed.56
Unlike the law in Byzantium, Zoroastrian law permitted marriage to a
non-Zoroastrian, but the children were not considered part of the family
lineage and therefore were not regarded as heirs of their father. They were
also exempted on this account from the familys commitments and debts.
The same applied to a child who had converted from Zoroastrianism to
another religion.
As far as can be determined, apart from the issues of marriage and
slave ownership, there does not seem to have been a consistent policy
regarding the non-Zoroastrian communities living under Sasanian
rule. Both Jews and Christians seem to have had their ups and downs,
depending on the specific policy and outlook of the current king. Jews
and Christians were persecuted in different periods;57 especially famous
were the persecutions of Yazdgird II (r. 43857), who persecuted both
Christians and Jews, forbidding the Jews from observing the Sabbath.58
Although Khusro Anu4sh3rwa4n was more tolerant, he imposed a special
tax on Christians and Jews.59
On the other hand, Sha4pu4r III, Bahra4m IV, and above all Yazdgird I
(r. 399421) adopted a favourable policy towards the non-Zoroastrians,
120
mostly on political grounds.60 Yazdgird, who had signed a treaty with the
Byzantine emperor, Theodosius I, allowed the rebuilding of the destroyed
churches, freed Christian prisoners, and actually validated the decisions
of the church council that convened in Ctesiphon in 410. The bishops
were convened at the Royal Port and were promised freedom of cult, the
right to build churches, and secured the authority of the Catholicos over
the community.61 Similar commitments were made in an appendix to the
ByzantinePersian treaty of 561, cited by Menander Protector, regarding
the rights of the Nestorian Christians living under Persian rule.62
Inconsistency was such that at other times non-Zoroastrians could be
part of the upper strata, and could be accorded special honours. Thus,
Yazdgird I married Shoshinduxt, the daughter of the Jewish Exilarch
(resh galutha), and one of the Exilarchs received from one of the kings
a kamar, a belt, as a sign of high distinction, usually indicating a vassal
lord relationship.63
Both Jews and Christians served as scribes and officials in the Sasanian
government. There were Christians who served in the upper echelons of
the Sasanian army, bearing titles such as aspa4r and marzba4n. These were
actually members of the local nobility who had become Christian. This
posed a big problem to the Sasanian rulers, who tried to convince them
to revert to Zoroastrianism time and again. The problem was especially
acute regarding Armenia, where significant parts of the population had
become Christian.64
In general, there do not seem to have been many rules on religious
identity. As stated by Widengren, the position of the religious minorities is impossible to understand if we do not take into account the fact
that there existed no articulated legal principle regulating their position
except the religious law, which would have led to the execution of all
non-Zoroastrians.65
In sum, in the Byzantine empire a series of laws restricting the nonChristians developed following the Christianization of the empire. These
laws multiplied and changed with the proliferation of Christianity
throughout the empire and the consolidation of its position as the exclusive religion of the empire. Nevertheless, the laws mentioned regulated
specific issues and were interspersed among other laws. At no stage was
there a separate comprehensive document including a set of regulations
pertaining specifically to non-Christians. This is even truer in the case of
the Sasanians, where laws pertaining to non-Zoroastrians were few, while
Sasanian policy towards them was volatile, and shifted often between
tolerance and intolerance.
muslim and pre-muslim modes

of subordination compared
Byzantine and Sasanian Precedents to Muslim Law
Regarding Non-Muslims
If we review the status of non-Muslims under Muslim rule, we could
point out several significant restrictions that correspond with Byzantine
law, and to some extent even with Sasanian law, though the latter differ in some respects. As shall be presently demonstrated, only some of
these corresponding restrictions are incorporated in Shuru4t Umar, while
others have become part of Muslim law and shar a.
A methodological comment is in place here:while the Byzantine laws
regarding the Jews go back mostly to Theodosians Code in the first half
of the fifth century, they were later included in Justinians Code, and
appear also in epitomes, collections of Justinians Novellae, canonical
collections, and nomocannons which were compiled throughout the sixth
and beginning of the seventh centuries.66 Moreover, Justinians Code continued to be in use in the East until the second half of the ninth century,
when a new code, the Basilika, took its place.67 The Basilika, it must be
added, was in its turn based on the Corpus Iuris Civilis.68 As has been
demonstrated by Jokisch, Byzantine law was known and employed in the
caliphate at the end of the eighth century CE, and thus could well have
influenced the process of defining the status of non-Muslims in Muslim
society which was taking place at that time.69
The following issues in Muslim law of non-Muslims correspond with
either or both Byzantine and Zoroastrian law:
1. The general concept that existed in both empires, of dividing citizens according to their religious identity Christians and nonChristians, or Zoroastrians and non-Zoroastrians was a basic
tenet of Muslim rule. This is at variance with other preceding
concepts, where the division was according to tribal bonds or
citizen versus resident alien. Furthermore, Byzantine rule distinguished between Jews, on one hand, and pagans and heretics, on
the other. Thus pagans were forbidden to hold feasts or carry out
any cult. Their altars everywhere were to be destroyed, and
their temples vindicated for public use.70 Pagans and heretics were
penalized for their beliefs and cult by capital punishment, confiscation of property, and exile.71 This is totally in concert with the
Muslim concept, which differentiates between ahl al-kita4b and the
122
mushriku4n. Just as the Jews were considered adherents of a religio licita and allowed to keep their synagogues and their cult, so
were the ahl al-kita4b. This was true neither for the pagans nor for
the mushriku4n, who were persecuted until they converted. It thus
may be that the special status given to ahl al-kita4b in the Muslim
caliphate was inspired initially by the prior status of the Jews in the
Byzantine empire.
2. As we have seen above, the fate of non-Muslim prayer-houses
was the focus of heated discussion throughout the eighth century,
and opinions ranged from the obligation to protect all prayerhouses in places conquered through an agreement to the claim that
they may or even should be destroyed in any place settled by
Muslims.72 This discussion seems in effect to have come to an end73
with the acceptance of Shuru4t Umar, which stated that we shall
not build, in our cities or in their neighbourhood, new monasteries,
churches, convents, or monks cells, nor shall we repair, by day or
by night, such of them as fall in ruin (ma khuriba min kana4isina)
or are situated in the quarters of the Muslims. The law regarding the status of synagogues in the Shuru4t is almost identical to
the analogous Byzantine law. In both cases, prayer-houses are to
be protected. However, new ones are not to be built, and abandoned ones not to be reconstructed.74 Although not mentioned
explicitly in the Shuru4t, the text implies, as does Byzantine law,
that repair work was permitted in order to prevent immediate ruin,
with the exception of those that were already ruined, and presumably had fallen into disuse.75 Additionally, some of the versions of
the Shuru4t note that Christians were not allowed to congregate in
churches that were located in the midst of Muslim quarters (wa-la
naqsidu al-ijtima4 f ma4 ka4na minha4 f khitat al-muslim n),76 or to
repair those churches that were in Muslim quarters (wa-la nujaddid ma khuriba min kana4isina wa-la ma ka4na minha4 f khitat
al-muslim n).77 This seems to be an extension of the law preventing the reconstruction of abandoned churches. In fact, the message was that churches located where the population had become
mainly Muslim were as good as abandoned.
3. Like Byzantine and Zoroastrian law, Muslim law prohibited nonMuslims from owning Muslim slaves. Thus a Christian was compelled to sell his Muslim slave, or a slave who had converted
to Islam, to a Muslim.78 The Shuru4t states that Christians were
not allowed to hold slaves who had been taken by the Muslims

(la4 nattakhidhu min al-raq q ma4 jarat alayhi siha4m
al-muslim n).79 This seems to be again an extension, stating that
in fact no prisoner taken by Muslims could be sold to a nonMuslim. As captives were the main source of slaves,80 this would
mean that, at least in theory, it may have been quite difficult for a
non-Muslim to purchase a slave.
Nonetheless, the Geniza documents indicate that the Jews were
big in the slave trade. Jewish traders purchased their slaves in
distant places such as Sudan, Nubia, India, and Europe (Greek
speakers as well as Franks) and often converted them to Judaism,
in spite of the Muslim prohibition on converting anyone to any
religion other than Islam.81
4. The prohibition on holding public office, which Byzantine law
imposed on all non-Christians, was adopted by the Muslim authorities as well.82 Precisely the same sentiments on the question of
unbelievers in public office that are reflected in the Byzantine and
Sasanian prohibitions83 are expressed in Muslim literature through
4 warns repeatedly
the ages.84 Interestingly, although the Quran
against the hypocrisy and disloyalty of the unbelievers, and advises
against taking them as awliya4 ,85 and despite the fact that there was
general consensus regarding the unacceptability of unbelievers in
public office, such a prohibition is not mentioned in the Shurut4 . This
may be due to the fact that rather than being part of the Shurut4 , a
document to which the non-Muslims commited themselves in order
to receive the status of the Protected People, here is a law that
needed to be put into effect by the rulers, given that employing a
non-Muslim was their exclusive privilege.86 This prohibition was
indeed imposed by the caliph al-Mutawakkil (84661CE).87
5. Both Byzantine and Islamic law punish apostasy with death, and
deem the preventing of an unbeliever from joining the true religion
to be a cardinal offence.88
6. The attitude to intermarriage in Islamic law seems to deviate somewhat from both Byzantine and Zoroastrian law. Byzantine law
prohibited such marriages unequivocally, while Zoroastrian law
permitted marriage to a non-Zoroastrian woman while denying
the children of such offspring the status of heirs. Based on Q 5:5,
Islam allows marriage between a Muslim man and a non-Muslim
woman from ahl al-kita4b, while forbidding marriage between a
non-Muslim man and a Muslim woman (Q 2:221). Some jurists
explain that the latter would be a form of enslavement of a Muslim
124
by a non-Muslim, yet it is self-evident that while in the former

case the children are raised as Muslims, in the latter they remain
dhimm s, a situation which is of course undesirable.89
7. The Muslim law regarding inheritance by non-Muslims is similar
to Byzantine law in that dhimm s cannot inherit from Muslims, as
already noted by Fattal,90 and to Zoroastrian law in that a nonMuslim woman may not inherit from her husband.91
8. Muslim law, like Byzantine law, considers a non-Muslims testimony inadmissible against Muslims.92
9. As Byzantine law (above) prohibits Jews from mocking and insulting Jesus and Christianity,93 so Muslim law condemns the defamation (sabb; shatm) of Islam and the Prophet.94 This issue is discussed
by Abu4 Ubayd,95 and is the first issue treated in al-Sha4fi3s alternative document, which was to be signed with the Christians seeking
an ama4n.96 The Shuru4t seems to relate to this indirectly in several
clauses, prohibiting the display of polytheism, the selling of wine,
and the rearing of pigs in Muslim neighbourhoods.97
Theoretically, it may be assumed of course that some of these similarities between Byzantine law and Muslim law regarding the minorities are
sociologically inherent to the situation. However, in two issues the similarity is striking. The first the division of society into Muslims and nonMuslims, non-Muslims being themselves divided into mushriku4n and
ahl al-kita4b is completely in concert with the distinction in Byzantine
society between Christians and non-Christians; non-Christians themselves divided into Jews, on one hand, and pagans and heretics, on the
other. The second is found in the law regarding the prayer-houses, stating
in both cases that existing structures might be kept, new ones might not
be built, and repair work could be allowed only in those that were still
active. If we are to add to these two the accumulation of similarities in the
other issues, including the holding of public office, the matter of slaves,
and those of inheritance, defamation, and partially intermarriage, it
seems more than likely that Muslim law regarding non-Muslim communities was influenced by the position of non-Christians in general, and
Jews in particular, in the Byzantine empire.
This is well in line with Patricia Crones view, as expressed in her
Roman, Provincial and Islamic Law, that the Muslims adopted laws
that prevailed in Near Eastern societies, adapting them to their needs;
they were not necessarily taken directly from the Byzantine law code,
but rather followed the prevailing local or provincial laws and customs,

absorbing elements from Roman, Jewish, Hellenistic, Sasanian law, etc.98
Moreover, Jokischs thesis regarding the adoption of the actual Roman
law code by the Muslims via its later versions such as the Digestsumma99
towards the end of the eighth century reinforces the view that the Muslims
were well acquainted with Byzantine law and did not hesitate to adopt it
where they found it adequate for their needs.
At the time of the arrival of the Muslim conquerors, societies of the
Byzantine East distinguished clearly between its Christian and nonChristian members, and applied a series of regulations and restrictions
that discriminated against non-Christians. The conquerors could well
have been made aware of these both via the Latin or Greek versions
of the law, as well as through contact with members of the conquered
societies.
Rules Regarding Non-Muslims that Cannot Be Traced Back
to Byzantine or Zoroastrian Law
We have seen above that there are significant similarities between
Byzantine and Muslim law regarding the position of members of other
accepted religious communities.100 In the following section I would like to
examine a large number of Muslim rules regarding non-Muslims which
have no apparent parallels in Byzantine or Zoroastrian law. As will be
seen, these rules appear in various Muslim sources; yet given the prominence of the Shuru4t document in Muslim society, and the special place
accorded to these unprecedented clauses in the document itself, I believe
it is best to examine this issue on the basis of the Shuru4t.
It should be noted that there is one striking difference between the
Shuru4t and pre-Islamic law regarding minorities which needs to be
emphasized. This is the fact that the Muslims chose to draw up a special
document presenting the regulations and restrictions by which the nonMuslims were bound in return for their ama4n. Documents regarding the
status of minorities or subordinate groups are not attested before the
coming of Islam. Thus, despite the similarities discussed above, we are
seemingly faced with a unique phenomenon.
As noted, Shuru4t Umar contains many clauses regarding the religious
minorities whose origins cannot be found in the Byzantine or Zoroastrian
laws. We are thus faced with the question of the origin and purpose of
these clauses.
The origins of some clauses are more easily explained; thus, clause2,
regarding the right of passage of all Muslims, the duty to supply them
126
with board and lodging for three days, and the commitment not to give
shelter to spies or to conspire against Muslims, is clearly a remnant of
the early agreements, as has already been noted by Noth.101 The authors
of the Shuru4t, just like Abu4 Yu4suf, based the general sulh agreement on
the early conquest agreements, extracting from them both the format of
a surrender agreement and the clauses that were still suitable for their
needs. Other clauses clause 9, prohibiting the selling of wine or the
rearing of pigs in the vicinity of Muslims;102 clause 3, the prohibition on
teaching the Qura4n to our children; or the clause on displaying crosses;
or that concerning polytheism, which does not appear in al-Turtu4sh3s
version103 are likely to be an adoption and extension of the Byzantine
prohibitions on contempt of the local religion and cult,104 although they
may well have developed independently, being natural and intuitive measures aimed at protecting the ruling religion and asserting its supremacy.
We are still left with a significant number of clauses which are not
accounted for:
1. clause 5: The obligation to show respect to the Muslims and
give them priority in seating, and not calling on Muslims in their
homes.105
2. clause 6:The requirement not to resemble the Muslims in dress,
hairstyle, speech, or kunyas. Listed specifically here are the qalansuwa, the turban (ima4ma), footwear, and the parting of the
hair.106
3. clause 7:The prohibition on the use of saddles, girding swords or
carrying arms.
4. clause 8: The prohibition on possession of seals engraved in
Arabic.
5. clause 10: The requirement to clip the [hair at the] front of the
head presumably to cut the hair across the forehead short, i.e. to
wear a fringe.
6. clause 11: The requirement to dress the same way wherever we
may be and to bind the zunna4r around our waists.
7. clause 12: The prohibitions on conducting religious processions,
especially on Palm Sunday and at Easter (wa-la4 nakhruju ba4u4than
wa-la4 shaa4n n), on using the clappers loudly, on the raising of
voices in church services in the presence of Muslims, or during
funerals, or the showing of lights in Muslim roads or markets.
What are the sources and the significance of all these prohibitions and
demands imposed by the Muslims on non-Muslims? As noted above,

most of these have easily been explained away as products of the demand
for ghiya4r, or the need for a distinctive appearance, based on the ideal of
la4 tashabbahu4,107 an ideal that sprang up in the nascent Muslim community, which had to differentiate and define itself clearly versus the other
long-established religious communities of Jews, Christians, and polytheists. Noth tried to claim, indeed, that all of these demands were a result
of this need for differentiation.108 Yet all researchers have agreed that the
document was finally formulated and canonized at a time when the rule
of Islam was already well established. If differentiation of the unformed
Muslim community was the issue then it would have been much more
fitting for such prohibitions and requirements to be applied during the
period following the conquest! However, we do not have any evidence
that such demands were imposed at the time. Even according to Muslim
tradition itself, it is only in the days of Umar b. Abd al-Az3z that demands
for distinctive dress began.109 In addition, the ideal of la4 tashabbahu4 actually requires the Muslims not the non-Muslims to take action to differentiate themselves. Examples are the requirement that Muslims dress
differently, greet each other differently, or refrain from kissing tombs, as
Jews and Christians do.110 Ibn Qayyim al-Jawziyya himself states that
until the conquest only believers could be commanded to adopt certain
distinctive customs, and that it was only after the conquest that Umar b.
al-Khatta4b, to whom the document is attributed, could order ahl al-kita4b
to distinguish themselves from the Muslims.111 As already noted, however, this did not in fact happen following the conquest, but, it seems,
only almost a century afterwards, as shown above (Chapter 3).
In addition, some of these clauses cannot be defined as an expression
of la4 tashabbahu4. Examples are clause 5, the obligation to show respect
to Muslims and give them priority in seating, the prohibition on calling
on Muslims in their homes; or clause 12, which includes prohibitions on
conducting religious processions, on using the clappers loudly, on raising
the voice in church services or in the presence of Muslims, on raising the
voice in funerals, and on showing lights in Muslim roads or markets.
These clauses, central to the Shuru4t, should probably be seen as the
expression of a new perception developed by the Muslims in the course
of the eighth century, regarding the place of non-Muslims in Islamicate
society. As shall be demonstrated, the roots of this new perception were
deeply entrenched in the Iranian social ethos and mores.112 My claim is
that the building-blocks used by the Muslims to create this new social
order actually came from the dismantling of the old. That is, the Muslims
adopted the Sasanian aristocratic ethos regarding the social structure
128
and its status symbols, and used them in order to appropriate it, and
thereby establish their own supremacy over the non-Muslims.113 The
non-Muslims, or dhimm s, were now treated as the lower class had been
in Sasanian society, thus becoming an inferior class of resident aliens in
the new social order that was being promoted. This was reflected in the
Muslim appropriation of ancient Iranian status symbols, which were
thenceforth forbidden to the non-Muslims. The Shuru4t document thus
needs to be seen within the comprehensive context of this new Muslim
perception of a social order that was being formed, and not just as a set
of rules pertaining to the non-Muslims which was invented ex nihilo by
the Muslim rulers following the conquest.
Before I go on to demonstrate this I should like to make a methodological comment:there is no doubt that the Muslims adopted many notions,
concepts, and costumes from both the Byzantine and Sasanian worlds. In
the first part of this chapter I referred to regulations and laws that seem
to have originated in the Byzantine realm. In the following pages I will
be referring mostly to concepts that seem to me to have originated in the
Iranian court culture and social mores. Still, it must be taken into account
that some of these customs and practices were prevalent in the Byzantine
realm as well. This is not surprising, as the Hellenistic kingdoms had
been, from the time of Philip of Macedon,114 and much more so from the
reign of Alexander the Great, well aware of the Persian court, and sought
in fact to emulate many of its practices. Alexander adopted articles of
Persian royal dress the tiara, the girdle made use of the golden throne,
adopted Persian court ceremony, and even attempted to introduce the
proskynesis (prostration), an ancient act of homage to the Persian kings,
among his Macedonian and Greek subjects,115 an attempt that aroused
much controversy in the Greek world and was vehemently rejected.116
Although there is disagreement over whether Alexander proclaimed himself Great King (Shahunshah), there is no doubt that he adopted all of
these components as part of his kingship of Asia. Kingship, half Greek
and half Oriental, dominated the eastern Mediterranean for the three
centuries after Alexander.117 This process was further enhanced later on
by the changes in imperial status and the adoption of the Hellenistic royal
practices introduced from the end of the third century CE onwards by the
emperor Diocletian and the following emperors, when all the remaining
external manifestations of the Roman civitas ideal had been disposed of.
By the sixth century Byzantine aristocratic court dress, manners,
and costume were just as lavish and luxurious as those of the Persian
nobility,118 and ceremony and ritual just as pompous.119 Elaborate silk

robes and other paraphernalia were adopted in honorific investiture ceremonies such as church accession, ambassadorial exchange, and bureaucracy promotion.120 Justinian issued an edict according to which no one
but the emperor was allowed to decorate his bridles and saddles or his
belts with pearls, emeralds, or hyacinths.121 It is nevertheless clear that
the majority of these affectations originated in the Iranian world. In
fact, scholars have suggested that Byzantine emperors admired Sasanian
pomp and emulated it and that the mosaics of Ravenna were a deliberate
attempt to imitate Sasanian imperial iconography.122
Nevertheless, it seems to have been the Sasanian aristocratic social
ethos that played the main role in the creation of dhimm status. For
one thing, the principles of a rigidly structured hierarchical class society,
which stand at the basis of Shuru4t Umar (as will presently be shown), are
typical of the Sasanian rather than the Byzantine social ethos. Byzantine
society was in fact, both in theory and in practice, a much more mobile
society, as noted by Haldon, who says that Late Roman Society was
neither rigidly hierarchical nor inflexible. It was possible to move from
relatively humble status to a position of considerable wealth and power,
particularly through service in the army or another state service.123 In
Byzantium the unique appearance and the special appurtenances represented the office, the formal position, or the function, rather than a fixed
and inherent social class that one formally belonged to, as was the case in
Sasanian society theoretically, even if not always practically.
This is not to say, of course, that Islamicate society would replicate
Sasanian society, nor would it adopt its social ethos in full, yet it did
espouse many of its social concepts.124 Although less hierarchical than
Sasanian society, Islamicate society would be made up of two main sectors:that of the ruling Muslims, which would in its turn be divided into
kha4ssa (the distinguished people or upper classes), aya4n (the notables),
and a4mma (common people); and that of the non-Muslims, which
was socially inferior to the Muslim sector. One should, however, take
into account that in both Sasanian and Islamicate society there existed
an evident gap and continuous tension between social theory and
everyday reality.
In addition, many of the clauses of the Shuru4t to be discussed here are
clearly of Sasanian origin, and bear no Byzantine parallel. This goes for
both items of apparel and various social mores.
Another claim in support of Iranian rather than Byzantine influence is
that while Byzantine presence greatly diminished following the conquest,
Sasanian presence did not. Byzantine formal military and civil elements
130
shrank dramatically in the wake of the Arab conquest. Although certain

elements of Byzantine administration and rule continued to be applied
throughout the seventh century in the territories that had previously
been part of the Byzantine empire, large parts of the upper strata and
certainly the formal administration left the conquered territories and
emigrated to Byzantium.125 The only remaining representative element
of the Byzantine formal hierarchy was the clergy. Thus Byzantine society actually lost the main constituent of its upper classes, including the
backbone of its administration. The result was that there probably was
not a substantial group of Byzantine bureaucrats, officials, generals, and
others present, who could serve as a model. True, the Byzantine system
of administration was retained to a large extent throughout the first century after the conquest, but its bureaucrats were no doubt low-grade
local administrators rather than those who served in the upper echelons
of the Byzantine administration. Sasanian society, on the other hand,
apart from the highest echelon of its nobility, retained most of its social
components, including its administrators and scribes; these went on to
serve as the transmitters of the social values and concepts of Sasanian
society.126 Significant parts of the Sasanian elites thus continued to display all the symbols and appurtenances of the social status into which
they were born.
the sasanian origins of the social position

of non-muslims in islamicate society
The Rejection and Adoption of Royal Sasanian Manners
and Status Symbols
A brief review of the process of adoption of the Sasanian social ethos and
symbols of authority is in order here, as this serves as the key to unravelling the meaning and significance of the central part of the Shuru4t.
The conquest of the East by the Muslims was followed by a process of
slow and gradual transformation of Muslim society, from an Arab tribal
society to a highly urbanized and cultivated civilization which adopted
many elements characteristic of the local cultures. This is documented in
traditions that both mirror this process and question or reject the values
that stand at its basis.
One of the first and most prominent elements was the adoption of new
items of apparel which designated the status of their wearers. Muslim
literature is replete with traditions regarding the prohibition on wearing

silk127 and luxurious garments. Many of these traditions are attributed
to Umar b. al-Khatta4b. Although it is highly doubtful that the following
incidents took place as described, they nevertheless demonstrate that the
conquest had brought about a significant change in the dress and apparel
of the Arabs, a change that aroused great controversy among them.
Umar b. al-Khatta4b is presented in both Muslim and non-Muslim historiography in the guise of a Bedouin, dressed in shabby clothes, riding
on a camel, carrying with him only his most essential needs. Upon his
arrival in Jerusalem, he refused offers from Muslims and Christians to
improve his apparel or to exchange his camel for a valuable horse.128
According to the Christian sources he finally agreed to accept the clothes
offered him by Sophronius only temporarily, until his own shabby garments were washed. He is described in one of the sources as weaving
baskets and mats out of palm leaves for his living!129
According to these traditions Umar vehemently rejected anything to
do with the luxury, finesse, and other elements of comfort and extravagance that abounded around him in the territories that were being conquered, thus unmistakably representing the Bedouin ideal. This is well
exemplified in a famous story, also told in several versions, the earliest
of which appears in al-Bala4dhur3:130 When Umar went to Syria he saw
Mua4wiya accompanied by a retinue, such as was typical of Sasanian
royalty. Umar disapproved of this, reprimanded him, and said:This is
the Khusro of the Arabs! (la-hadha kisra4 al-arab). Mua4wiyas response
to this was Since we are in the land of our enemy, and are being watched
by many of the enemies spies, we need to exhibit our strength. Umar
replied that he did not know whether this answer was a sound strategy or a clever deceit. Thus, although according to this tradition Umar
disapproved of it, he could understand the logic of adopting the ways
of the aka4sira (Sasanian royal manners). A similar incident is adduced
by al-Tabar3 in which Umar met Yaz3d b. Ab3 Sufya4n, Abu4 Ubayda and
Kha4lid b. al-Ja4biya. They appeared riding on horses and clad in brocade
and silk. Umar pelted them with stones, saying:How quickly were you
turned away from your senses! Is it me that you are coming to meet in
this attire? If you did this at the head of two hundred men, I would
have replaced you with others. Their reply was:O Commander of the
Faithful, these are coats (yala4miq) and we have our weapons with us.131
This story is repeated in a more elaborate manner by al-Qalqashand3
and by Ibn Khaldu4n. Al-Qalqashand3 adds that when Mua4wiya became
caliph he constantly displayed his sovereignty. The caliphs that followed
him went even further, till the caliphate became like a kingdom surpassing
132
those of the Khusros and the Caesars.132 In his chapter The characteristic
emblems of royal and government authority Ibn Khaldu4n says:
At the beginning of Islam, the Muslims wanted to avoid the coarseness
of royal authority and do without royal customs. They also despised
pomp, which has nothing whatever to do with the truth. The caliphate
then came to be royal authority, and the Muslims learned to esteem
the splendor and luxury of this world. Persian and Byzantine clients,
subjects of the preceding (pre-Islamic) dynasties, mixed with them and
showed them their ways of ostentation and luxury.133
Although these traditions evidently should not be taken at face value,

they clearly reflect awareness of the fact that Muslim rulers abandoned
Arabian tribal customs and values in favour of those of the Sasanians and
the Byzantines, and it is clear that this transformation was controversial.
The adoption of the manners of the aka4sira by Arab caliphs and commanders is conceived as a negative process that would culminate in the
Arabs losing their own cultural and political tradition. The terms used
most often to describe this process aka4sira, kisrawiya, kisra4 al-arab,
and kisraw al-fil indicate that the Muslims themselves perceived this
process as the adoption of aristocratic Sasanian manners.
The tension around the adoption of royal manners, i.e. imitating the
ostentatious and lavish dress, paraphernalia, and manners of the Sasanian
nobility, may also be found in traditions expressing controversy regarding the appearance of the Prophet. Thus, there are conflicting traditions
regarding the Prophets attitude towards such issues as the use of prestigious textiles such as silk, satin, and brocade clothes, silver vessels, gold
seal rings, and hairstyles.134
In later times, however, the term seems to have lost its negative connotation. Thus, the term kisraw al-fil can denote a true royal style of life.
Abu4 Dulaf al-Ijl3 (d. 840), a prominent figure in the court of al-Mamu4n
and al-Mutasim, described himself thus.135
It is quite evident that these early stories of the rejection of Sasanian
royal manners or status symbols from the beginning of the Umayyad
period in fact point to their early adoption. This espousal encompassed
many areas.136 First and foremost was the adoption of the local vestimentary system by the ruling classes among the Muslims, as is clearly shown
in Yedida Kalfon-Stillmans book Arab Dress.137 Thus, while according
to the Qura4n the Prophet had rejected luxury and extravagance (isra4f),
such as silk clothes (Q 7:26), his followers in the major cities during the
Umayyad period were already wearing silk, brocade, satin, etc.

There is plenty of evidence supporting the early infiltration of Persian
elements into Muslim dress and art, as well as symbols and attributes,
even in areas that had previously been controlled by the Byzantines. In
fact, some such elements of Sasanian culture penetrated Byzantium and
Arabia even before the conquest.138 Persian items of apparel were widely
used by the Umayyad caliphs, beginning probably from the time of Yaz3d
I (r. 6803).139 The Umayyads adopted typical Persian dress items such
as the Persian trousers the sirwa4l and the qalansuwa.140 Sulayma4n b.
Abd al-Malik (r. 71517) and his courtiers dressed luxuriously, wearing
among other garments a sirwa4l and a qalansuwa taw la made of variegated silk.141
The caliphal statue (possibly representing the caliph Hisha4m, who was
known for dressing in a luxurious manner) at Khirbat al-Mafjar (second
quarter of the eighth century) is also presented with Persian royal dress and
attributes.142 The figure of the caliph is wearing an ankle-length Sasanianstyle qaba4 with pearl border, the coat cinched with an ornamented
belt, and sirwa4l can be seen below the hem of the coats flared skirt.143
This caliphal figure is holding a dagger in the manner of the Sasanian
nobles. A stone qalansuwa taw la, a Sasanian royal attribute, hung in
the audience hall of Khirbat al-Mafjar, most probably over the throne.144
The art and architecture of these desert retreats Khirbat al-Mafjar, Qasr
al-Hayr al-Gharb3, and Qusayr Amra are full of Sasanian influences,
and especially of Sasanian royal symbols such as winged horses, and the
simurgh as well as various artistic elements.145 In addition, the winged
crown found on Umayyad coins pre-dating Abd al-Malik as well as in
the Dome of the Rock is modelled after the shahs crown.146 There is
no doubt, therefore, that there was extensive adoption of Sasanian status symbols, including items of apparel and other attributes, among the
Muslim elite as early as the end of the seventh century and the first half
of the eighth. Kalfon-Stillman believes, however, that the Umayyads used
these only in private, while formally the appearance of the sayyid was
still preserved.147 She notes that Persian cultural influences became more
pronounced under the Abba4sids [at that time] Persian garments such
as sirwa4l, jawrab, and the qalansuwa taw la became widely popular.148
It should be noted that some prestigious articles of clothing adopted
by the Arabs, such as the khila, or robe of honour, and the tira4z, embroidered margins or embroidered robes, as well as the mintaqa, the special
bejewelled sword belt worn by officials (see below), were worn by both
Byzantine and Persian aristocracy. However, the custom of bestowing a
robe of honour was a very old and significant Iranic tradition, as were the
134
Illustration 2. The figure of the caliph holding a dagger and wearing a

Sasanian-style qaba4, an ornamented belt, and sirwa4l; from the bath at Khirbat
al-Mafjar, second quarter of the eighth century (photograph by Clara Amit,
courtesy of Israel Antiquities Authority).

special belts. Taking into account the multitude of Iranian items of dress,
it is thus very likely that these two were also most probably passed on
through Sasanian rather than Byzantine cultural mediation.
The appropriation of Persian status symbols, however, went much
further than just the adoption of prestigious items of apparel. In fact,
with the rise of the Abba4sids, the caliphs seem to have reconstructed,
to a great extent, the Persian court, its ceremonies, and its manners.149
This is expressly noted in adab literature.150 They adopted the insignia
of sovereignty such as the throne (kurs , sar r), the curtain (sitr), the
royal seal (khatam), the banners (alwiya), robes of honour (khila).151
Other paraphernalia such as turbans, boots, belts and swords (which will
be referred to later on) were an important part of this. Colours were
also a very important element. The caliph wore a black turban and a
black official robe, and red boots the latter being a symbol of royalty
during the Sasanian period.152 In addition, black, the dynastic colour of
the Abba4sids, was limited only to the caliphal family and the various
court officials.
Like the Persian court, the Abba4sid court was extremely formal and
hierarchic, and the position of each member in the audience hall, his
dress, and the way he was addressed, were set according to his rank and
dignity. There was a distinct difference between ceremonial dress and
ordinary dress.153 The titles, ceremonies, and protocols of this court are
minutely described in adab literature such as that of Pseudo-al-Ja4hiz
(776869),154 Ibn Abd Rabbih (860940) in al-Iqd al-far d,155 and Hila4l
al-Sa4b3 (9691056), in his Rusu4m da4r al-khila4fa.156 The centrality of
these issues in Muslim society led in fact to an abundance of compositions related to adab and zarf.
It may therefore be concluded that the Muslim elites and caliphs had
by the end of the seventh century started to adopt many items of apparel
and paraphernalia used previously by the Sasanian elites in order to
exhibit their prestigious social status.
This process gained speed and importance with the rise of the
Abba4sids, when the adoption of royal Persian dress and manners moved
from the private to the public sphere and was encorporated into official
ceremonial events of the caliphal court. As noted by Rose, Sasanian royal
iconography continued for many decades after the end of the dynasty,
influencing courtly fashion in Islamic Iran as well as in Byzantium and
central Asia.157 As we shall see, this process was directly connected to the
central place that the element of ghiya4r acquired in the Shuru4t.
136

The Dib3rs and the Dihqa4ns as Transmitters
of Cultural and Social Concepts
The adoption of this formal hierarchical system and the accompanying

status symbols came about first and foremost through the influence of
two groups within the Sasanian society: the scribes (Persian dab rs;
Pahlavi dib rs; Arabic kutta4b); and the lesser nobility (dihqa4ns).158
According to the formal division of Sasanian society into four estates,
which was promoted during Khusro Anu4sh3rwa4ns reign,159 the dib rs
(scribes and secretaries) belonged to the third estate of Sasanian society, which comprised physicians, scribes, and astrologers, and came
after that of the priests and the warriors, and before the fourth and last
estate, the peasants.160 The dib rs were not only the official correspondents, accountants, and judicial secretaries of the king, but also important political figures:they held posts such as secretary of the army (ka4tib
al-jund), and secretary of finance (ka4tib al-khara4j), served at times as
ministers of defence, and were put in charge of land reforms and special
investigations.161
The dihqa4ns also belonged to the third estate, which most probably emerged as a social class following the land reforms of Khusro
Anu4sh3rwa4n (r. 53179).162 They represented the lower echelons of the
nobility, and were smaller land owners than the great nobles; they were
responsible for the management of local affairs, and their main duty was
to collect taxes for the government.
Both the dibrs and the dihqa4ns played important roles in the nascent
Islamic state. While large parts of the upper strata of Sasanian nobility
lost their status and privileges, the lower nobility, including the dibrs
and the dihqa4ns, adapted and, to a large extent, kept its former status.
The dibrs and the dihqa4ns managed to secure recognition of their privileges and positions, and in some cases converted and gradually made
their way into the upper strata of Muslim society.163 After the conquest,
the Muslims relied on the dibrs in the former Sasanian empire, as well
as on the local administrators in the conquered Byzantine territories,
to maintain the administration in its different forms in the conquered
areas. Their dependence on these administrative classes is well attested
in the sources.164 The dihqa4ns too fulfilled an important role in the comparatively undisrupted management of their districts, cities, and villages.
The fact that this process was also taking place in Iraq, where there was
a multitude of Christians and Christian centres, makes it doubly important when it comes to the issues we are about to expound on. These

groups thus served as the backbone of the Islamic administrative system;
through it they introduced the Muslim conquerors to various aspects
of their culture and society, and influenced them immensely during
the formative period of Islamic society.165 It is first and foremost via
these two groups that the developing Muslim society adopted many
of the concepts regarding social order and status that stemmed from
Sasanian society.
the ideology of the iranian class system

According to the prevailing theory, ancient Iranian society was divided
according to the Avestan concept into three classes or estates:166 the
priests; the warriors, or charioteers; and the farmers.167 Later, these
estates underwent some changes both in structure and terminology. Thus,
the priests became magians (magu-); the warriors were replaced by
the new noble estate a4za4ta4n, which also included the cavalry; and the
last estate, more variegated, was now called ram (flock). The Sasanian
reform, made not later than the beginning of the fifth century, included
four estates, and reverted to the Avestan terminology of priests and
judges, warriors, and farmers. Following the warriors in the hierarchy, a
new estate:the scribes (dipira4n), was added.168
Although it is unclear how strictly this ideal was upheld in practice,
this estate membership was seen as inherited from one generation to the
next, and theoretically there was to be no mobility from one class to
another.169 This is well exemplified by Firdaus3s famous story about a
shoemaker who was willing to lend the king a large sum so that his son
could become a scribe (dib r), but was rejected.170 The refusal was due to
the belief, voiced also by al-Thaa4lib3, that this might lead to a situation
where someone of inferior origin might humiliate the nobles.171 This is
in fact exactly the same reasoning as would be found in Muslim sources
regarding the prohibition on employing non-Muslims in official governmental positions.172
The establishment of these classes was attributed to mythological
founders such as Zoroaster (thus the Avesta and the Bundahishn) and
Jamsheed (thus Firdaus3 and al-Thaa4lib3). mile Benveniste demonstrates
that in fact these two traditions are based on one ancient tradition which
recounts that Jamsheed (Yama) founded the three classes, and these
were later given other names by Zoroaster. According to Benveniste, the
Avesta is here reflecting an ancient Mazdean pre-Zoroastrian tradition.173
The Sasanian social reform which took place around 500 CE actually
138
adopted this tradition in its ancient form, reverting once again to the
Avestan nomenclature, while adding the new class of the scribes (dipira4n).
Another class that seems to have been added to this scheme only in
the Sasanian reform was that of the dihqa4ns, which, as noted above,
represented the lower nobility. It should be noted that although dihqa4ns
are attested only in the late Sasanian period and most probably emerged
as a social class as a result of Khusro Anu4sh3rwa4ns land reforms, there
was nevertheless a strong tradition attributing the creation of the
dihqa4n class to We4kart or Waygird, brother of Hu4shang, the legendary
Iranian king, and also one attributing it to Manu4shihr, the mythical
ancestor of the Iranians.174 These myths attest to the fact that once the
social status of the dihqa4ns was established, it came to be conceived as
a primeval status in an established hierarchy, which could not be altered
or penetrated.
The ideal of the Iranian estates is clearly conveyed in several documents
originating in the Sasanian period. The first is Ahd Ardash r b. Ba4bak, a
document originating in the late Sasanian period, which was found along
with three other documents of Sasanian origin, in MS Kprl 1608.175
The document describes the division of society into four classes:(a) the
warriors; (b) the priests, ascetics, and guardians of the fire-temples; (c)the
scribes, astrologers, and physicians; and (d) the servants, tradesmen, and
peasants. It emphasizes the importance of this division, comparing its
preservation to the care of the body itself. Nothing can afflict the position of sovereignty more than the transfer of one of these classes to the
rank of another because the transfer of people from their rank will soon
bring about the removal of the king from his throne either by his removal
from power or even by his being killed.176 It goes on to explain that any
change in the social order would only bring about envy and anger and
wreak social and political havoc.177
The same idea is expounded in another document found in Kprl
1608, identified by Grignaschi as the Kita4b al-ta4j f srat Anu4shrwa4n
cited by Ibn al-Nad3m,178 which was written originally in Pahlavi in the
late Sasanian period and translated into Arabic, according to tradition,
by Ibn al-Muqaffa. This treatise describes the manner in which Khusro
Anu4sh3rwa4n organized the ceremony of the address from the throne during the feast of Nauro4z.179 It begins by saying that on the day of Nauro4z,
the king of the Persian kings would hold a public audience, distribute
gifts and robes, and set up tables. The people would sit according to their
rank, differentiated by the clothes worn for that day, and be silent.180
Another document, central for our understanding of the Shuru4t, is
the Letter of Tansar to the king of Tabarista4n, familiar in its Persian

translation from the early thirteenth century; this version was based
on Ibn al-Muqaffas Arabic translation (eighth century) of the Pahlavi
text.181 The letter, which was allegedly written in the time of the first king
of the Sasanian dynasty, Ardash3r I, but which according to some scholars
may have been written or edited towards the end of Sasanian rule,182 says:
He [i.e. the king] has established a visible and general distinction (tamyz
ta4hir wa-a4mm) between men of noble birth and common people (ahl-i
daraja4t wa-a4mma) with regard to horses and clothes, houses and gardens, women and servants. Furthermore, he has set differences (tafa4wut)
among the nobles themselves with regard to entrance and drinkingplaces, sitting and standing places, clothes, ornaments and houses,
according to the dignity of each mans rank. That they may look after
their own households and know the privileges and places appropriate
to themselves. So no commoner may share sources of enjoyment of life
with the nobles, and alliance and marriage between the two groups is
forbidden.183
Several pages later the question of class divisions and distinctions,

especially in dress and appurtenances, is underscored again, even more
emphatically:
This he has made a binding law, his purpose being to make clear the
divisions and distinctions among the people (wad kard wa qasd-i
awsa4t-i taqdr dar miya4n-i khala4 iq ba4dd a4ward), that the appurtenances proper to each class may be plainly seen. The nobles are
distinguished from the artisans and tradespeople by their dress and
horses and trappings of pomp, and their women likewise by silken garments; also by their lofty dwellings, their trousers, headgear, hunting
and whatever else is customary for the noble. As for the soldiers, or
fighting men, he has often conferred positions of honour and favours
of all kinds upon that group, because they are forever sacrificing their
own lives and possessions and followers for the welfare of those who
labour It is fitting that the working people should salute them and
bow before them and that the fighting men in turn should show reverence to the nobles, and that they should have regard one for another
according to the loftiness of their rank, and that they should maintain
their dignity.184
If there is still doubt that there was an element of subjugation involved

in these external differences, a proverb cited later on in the letter dissolves
this:
If you are one of the rulers of men, rule the noble with kindness and
generosity, but rule the base with contumely (dhull), for on contumely
they will reform. The base (lia4m) are to be subdued.185
140
An a4y n attributed to Ardash r, providing the code of behavior for the

nobility,186 emphasizes once again the importance of the class division,
and especially the dress code that differentiates between these different
classes:187
We have decided that there will be a special attire for the audience of the
kings that will not be departed from; and a special attire for feast days
which will not be departed from; and a special attire for the meeting
of young men and men which will not be departed from; and for the
reception of [people asking] favours an attire that will not be departed
from. We have imposed this on the noble because of his richness and
honour, and on the humble because of his poverty and baseness. As for
the domestic servant and the slaves we have imposed upon them the
dress that may be tucked-up during times of rest and travel. Finally,
regarding the peasants and cultivators we have imposed upon them
the dress of baseness (liba4s al-madhalla).
It is evident from these texts that it is not only that the rich nobles were
distinguished naturally by their extravagant and luxuriant clothes, but
that these clothes were actually imposed upon them, and were suited to
the function they were attending. As for the servants and the farmers,
not only could they not afford the rich, luxurious clothes of the nobles,
they were actually prohibited, according to this aristocratic ethos, from
wearing them and must be dressed in the garb that was imposed on them,
a garb that indicated their low and base station in society. Thus, according to this ethos, social status should be permanent and could not be
changed, and must be exhibited clearly and distinctly through ones attire
and paraphernalia.
The Iranian Class System:Between Ethos and Reality
It is now accepted that Arthur Christensens view, which accepted
this aristocratic ethos as a reflection of Sasanian society, needs to be
revised.188 Rather, it appears that this conception was promoted by the
Sasanians in order to serve their own interests against certain elements
in society that rejected this ethos and the social order it represented. This
may be observed in the Letter of Tansar itself, which implies that the
correct social order had deteriorated and that the nobles in particular
had been negligent in guarding their honour and the splendor of their
position.189 At a time when their rule was threatened on one hand by
the Parthian noble families who had rebelled several times against the
king, and even managed to depose him at times, and on the other by the

Mazdakite movement, which caused an upheaval among the lower strata
of society,190 the ideal of a closely structured social hierarchy characterized by immobility served as an important means of stabilizing the different social groups and restraining their ambitions.
Nevertheless, this ideal cannot be conceived just as empty Sasanian
propaganda. There is no doubt that Sasanian society was based on a
strong concept of hereditary social hierarchy. This is evident, as Rubin
emphasizes, in Sha4pu4rs inscriptions in Hajjia4ba4d and the Kaba-yiZardusht, which reveal a clear concept of a social hierarchy with rising
ranks of dignity according to the closeness to the king, already under
the early Sasanian kings.191 The connection between social status and
agnatic origin is also attested in the sigillographic evidence; thus the
kola4hs, or special hats worn by the high functionaries, carry, in most
cases, the emblems of the great Parthian noble families, indicating that
such functionaries usually originated in these families.192 Pourshariati
also notes, in connection with the effects of the Mazdakite movement,
that the economic, politico-religious, and finally territorial dimensions
of the agnatic structure of Iranian society, and the strong cohesive bonds
that these established, rendered the fabric of Iranian society far too interconnected for it to be overhauled easily. The agnatic structure especially
applied to the Parthian dynastic families.193 Thus, in spite of the fact that
the great dynastic families challenged the kings authority and power at
times, they themselves were avid adherents of the hierarchical hereditary
social system which awarded special privileges to the upper classes, while
the lower classes could never gain enough power to overhaul this ancient
system. There must of course have been a significant divergence between
ideology and reality, but it was nevertheless this hierarchical ideal that
stood at the base of the Sasanian social system.
The Adoption of Sasanian Aristocratic

Ethos by the Muslims
This hierarchical conception, expressed through a wide array of symbols
and codes, did not survive unchanged in its Sasanian form in Islamic
times. I will not, however, delve here into the discussion regarding the
extent of the influence of the Iranian social system on Muslim society.
Especially instructive and convincing for those wishing to examine it
in more detail is Marlows comprehensive analysis, which supports the
influence of the Iranian quadripartite social hierarchy on the social ideology of Muslim medieval society.194 There were arguments for and against
142
this system among Muslim philosophers and historians, and in practice it

was mitigated by the egalitarian and meritocratic streak in Islam.195 Still,
by and large, Muslim society carried with it significant social and cultural baggage from Sasanian society. In fact, according to Dimitri Gutas,
the Abba4sids actually saw themselves as the inheritors not only of the
Sasanians, but also of the ancient cultures of Mesopotamia and Iran.196
All in all, the Sasanian social conception was to have a meaningful influence on the shaping of the new social order in general, and specifically,
as I will try and demonstrate here, on the status of the non-Muslims
within it. The essence of it is that Shuru4t Umar is in fact a reflection of
the Sasanian social conception, which was passed on by Persian informants who were familiar both with the Sasanian aristocratic ethos and
with the social reality. As shall presently be argued, in the newly formed
Islamicate society the non-Muslims now at least in theory took the
place of the lowest class in society, who represented the base which
should be subdued.
The Status of Mawa4l3 in Early Muslim Society
The Arab conquest brought about the complete breakdown of Sasanian
society. Yet, following the conquest and during the Umayyad period there
were, in spite of the basic egalitarian concepts inherent in Islam, four
different groups in the society under Muslim rule:the Arab Muslims; the
non-Arab Muslims (mawa4l ); the dhimm s; and the slaves. At this point
the mawa4l , despite their conversion to Islam, were socially inferior.197
They were considered vile, while clients of clients were the most miserable persons to walk on earth.198 Many of the attributes of the position
of the mawa4l in Muslim society actually resemble those that are to be
found in the Shuru4t regarding non-Muslims. Thus, mawa4l were prohibited from marrying Arab women, and the life of a mawla4 was worth less
than that of an Arab. They were also, at least formally, considered unsuitable for holding official positions in the government.199 In addition, as
in the case of the non-Muslims, there were explicit status symbols that
differentiated them from the Arab Muslims:they were not to use a kunya
(surname of relationship), but ism (name) only, precisely as is stated in
the Shuru4t;200 they were not to walk alongside Muslims, and in public
gatherings they were allotted the last and humblest seats.201 The same
requirements are made in the Shuru4t:We shall show respect towards the
Muslims, and we shall rise from our seats when they wish to sit and in
al-Sha4fi3s Kita4b al-umm:You shall not occupy the middle of the road or
the seats in the market, obstructing Muslims.202 Many of these manners

of discrimination originated in the Sasanian social conception, as will be
presently shown.
Although the differences between the status of Arab and non-Arab
Muslims in Muslim society were never completely obliterated, they gradually diminished, the above-mentioned regulations disappeared, and some
of the mawa4l many of whom were dib rs (kutta4b) and dihqa4ns were
awarded central government positions, and became respectable members
of Muslim society. It was the non-Muslims, or dhimm s, who were now
to fill theoretically, although not always in practice the position of the
lowest estate in Islamicate society.
The Dhimm3s as a Social Stratum within Islamicate Society
While rules of differentiation and discrimination between the social
classes may not have been strictly set or followed within Muslim society,
they were now rigorously applied, at least in theory, to the non-Muslims
within the Islamicate realm. For the first time, the rules of social discrimination that the Sasanian aristocratic ethos had promoted as a means of
manifesting hereditary social position were based on religious rather than
on agnatic social origin, or even on profession, merit or wealth.
The term used in al-Mutawakkils famous edict to denote the nonMuslims is tabaqa203 that is, stratum indicating that the restrictions
in his edict applied to a social stratum, based in this case on religious
distinction rather than on agnatic origin or social position. The fact that
ahl al-dhimma were conceived in theory as a social stratum corresponding to one of the Sasanian social strata is evident in al-Tabar3s description of the tax system which according to tradition was established by
Umar b. al-Khatta4b. When comparing the Sasanian taxation system to
the Muslim one, he says that the three upper strata in Sasanian society
were exempt from taxation, while the rest, i.e. the fourth stratum, were
taxed according to their ability. These are, says al-Tabar3, the taxes
which Umar had emulated or adhered to (iqtada4 bi) when he conquered
the land of the Persians, and ordered that they should be collected from
Ahl al-Dhimma.204 The non-Muslims are clearly compared here to the
fourth stratum in Sasanian society.
The Concept of Ghiya4r
In the light of this, I would like to suggest that the concept of ghiya4r itself
originated in the Iranian world rather than in the ideal of la4 tashabbahu4
which, as already mentioned above, is a different concept.205
144
The notion of distinguishing marks was an established principle in

Sasanian society. Thus, the Letter of Tansar, cited above, says:He [i.e.
the king] has established a visible and general distinction (tamy z -
ta4hir wa-a4mm) between men of noble birth and common people (ahl-i
daraja4t wa-a4mma) with regard to horses and clothes, houses and gardens, women and servants. Furthermore, he has set differences (tafa4wut)
among the nobles themselves according to the dignity of each mans
rank.206 Later on he says: This he has made a binding law, his purpose
being to make clear the means of distinction among the people (wad kard
wa qasd-i awsa4t-i taqd r dar miya4n-i khala4iq ba4d d a4ward)207 that the
appurtenances proper to each class may be plainly seen.
To the best of my knowledge, there is no record of early usage of the
term ghiya4r in Arabic, before it was designated to mark the differentiating signs between Muslims and non-Muslims. It may be, therefore, that
it is a translation of the Pahlavi term for distinction or differences (possibly juda4g h?), lost to us in the process of the translation of the Letter of
Tansar to Arabic and then again to Persian.208
The Unprecendented Clauses of the Shuru4t as Part
of the New Social Ethos
Those clauses of the Shuru4t that could not be accounted for in other
ways are easily understood once they are viewed in connection with the
Sasanian habit of regulating the clothing of different social strata, including that of the lowest.
In fact, what these clauses forbid to the dhimm s are ancient status
symbols of mostly Iranian origin. Some of them are derived from the
dress customs of the Iranian nobility, others from the code of behaviour
in the presence of the king or other high officials in the royal court or
outside it. Some of these customs are very ancient, and are attested in the
days of the Achaemenid dynasty as one would expect, given that both
the Parthian and the Sasanian perceptions of royal authority were based
on the same Achaemenid prototype.209 Thus, for example, the terminology used to portray the state hierarchy in the Sasanian rock inscriptions
goes back to the time of the Achaemenians,210 as do the court titles of the
Sasanian period.211 But there are rules, such as those relating to the use
of kunyas, the clipping of forelocks, or funeral customs, that derive from
Arab tribal customs. Anything symbolizing authority, government, or
social prestige is denied to the dhimm s. In all cases, the supremacy of the
Muslims is exhibited alongside the servility and humility of the dhimm s,

who now formally took the place of the lowest social stratum. This is
more than anything else a formal de iure declaration of the supremacy of
Islam and its believers over the non-Muslims though de facto, of course,
this social hierarchy was never that simple, and there where many cases
where a rich dhimm , a dhimm doctor, or a dhimm office holder ranked
above a Muslim, often much to the discontent of the latter, as is often
attested in later Muslim literature.
I shall now review the rules one by one.
Clause 5:The Obligation to Show Respect to Muslims and Give
Them Priority in Seating (and on the Road)
The Letter of Tansar brings up the issue of seating arrangements:
Furthermore, he has set differences among the nobles themselves with
regard to entrance and drinking-places, sitting- and standing places.
Seating was indeed a central matter in Iranian society.
In an Armenian book of ranks (ga4hna4mag) which has survived,212 it
is mentioned that at the Armenian court the nobles and dignitaries sat
in the presence of the king according to their rank; the higher the rank,
the more cushions on the seat:They sat on cushions that were placed
progressively higher as they came nearer to the royal cushion.213 The
ga4hna4mag mentions the fixing of the positions of 400 cushions.214 This
custom was part of the ceremony of the Arsacid kings of Iran, which was
continued by the Sasanians. It goes much further back, however; hierarchic seating order was an ancient custom, as Xenophon (431355 BCE)
reports in his Cyropaedia that seating arrangements in the court of Cyrus
were ranked according to the esteem of the king, and that those who
misbehaved were reassigned to seats at the back.215 This special seating
arrangement in the Iranian court is also recorded by al-Masu4d3 and by
Pseudo-al-Ja4hiz who noted that under Ardash3r I (r. 22641 CE) the first
class (tabaqa), made up of the asa4wira4 and the princes sat to the right
of the shah, at a distance of about ten cubits (adhru); the second class,
which included the important marzuba4ns, the tributary kings present at
the court, and the spa4badhs, sat at about ten cubits distance from the
first; and the third, which was composed of clowns, men of leisure, and
jesters, of respectable origin and status, sat at the bottom of the room.216
The hierarchical seating order was closely adopted by the Abba4sids.
Thus, al-Tabar3 reports that in the days of the caliph al-Mutasim (r. 833
42 CE) the caliphal court was arranged in the following manner: first
would be the members of the Abba4sid family; then other respected
families (asha4b al-mara4tib); and then other courtiers and dignitaries
146
(wa-ghayrihum mi-man lahum martaba).217 The cushions, or seats of

honour, continued to be of great significance. Hila4l al-Sa4b3 reports that
when Adu4d al-Dawla was appointed governor of Iraq he received, among
many other things, an unstuffed seat of honour embroidered with gold,
complete with its leather cushions. Upon receiving this Adu4d al-Dawla
asked that the seat of honour be stuffed and carried in the street so that
others can see its splendour and the honour which it brings.218
The issue of the seating comes up, in fact, with regard to non-Muslims,
in the description of the restrictions imposed on the dhimm s by the
caliph al-Mutawakkil in a Samaritan chronicle of that period. The text
reads: He commanded that every dhimm should wear a distinguishing sign front and back, and that he should not sit in front on a velvetlike cushion (martabat khaml).219 The priority granted to the Muslims
regarding seating is reiterated by al-Sha4fi3,220 who says: You shall not
occupy the middle of the road or the seats in the market. Both elements,
the sitting position and the cushion, mentioned here, are clearly drawn
from the Iranian concept of court hierarchy and class distinctions, which
were continued into the Muslim period.221
The requirement not to occupy the middle of the road also appears
in al-Mutawakkils order (an yashmaillu4 f al-tar q)222 and in Muslim
had th literature, stating that Jews and Christians encountered on a road
should be forced to the narrowest part of the way.223 The idea that the
lower classes should clear the road is also attested in the Iranian royal
code of behaviour. According to the latter, the road should be cleared
on the approach of a king or dignitary. Xenophon recounts that before
Cyrus procession arrived rows of soldiers stood on this side of the street
and on that, just as even to this day the Persians stand, where the king is
to pass; and within these lines no one may enter except those who hold
positions of honour (tetime meno4n). And policemen with whips in their
hands were stationed there, who struck anyone who tried to crowd in.224
Clause 6:The Prohibition on Resembling Muslims
We shall not seek to resemble the Muslims by imitating any of their
garments, the qalansuwa, the turban (ima4ma), footwear or the parting of the hair. We shall not speak as they do, nor shall we adopt their
kunyas.225
Muslim tradition was well aware of the Iranian habit of identifying social strata by their special headgear, overgarments, trousers, and
footgear, as well as other paraphernalia.226 Muslim writers report that

Sasanid nobles wore a qalansuwa, the nature of which proclaimed their
status. Thus, al-Tabar3 notes that the highest nobility, belonging to the
seven noble families, wore a qalansuwa embroidered with jewellery
worth 100,000 dirha4ms, while the marzuban
4 s wore ones whose value
227
was 50,000 dirha4ms. Spa4badhs wore diadems twice as expensive as
those of the marzuba4ns. A marzuba4n wore a qaba4, and a bejewelled
belt. Important officials wore belts embroidered with a ewe, which was a
symbol of nobility. A dihqa4n had a crown and bracelets.228 Al-Jahshiya4r3
(d. 331 H) in Kita4b al-wuzara4 reports that it was the custom of the
Persian Emperors that each of the classes (tabaqa) in their service was to
wear an attire which was not to be worn by any other class. When a man
appeared before the Emperor, his trade and his class would be known
by his attire. All the kutta4b in attendance would wear their stipulated
attire.229 Al-Tabar3 recounts that Manu4shihr was the first who set
up dihqa4ns, imposing a dihqa4n over each village, making its inhabitants
his chattels and slaves, clothing them in garments of submission (liba4s
al-madhalla) and ordering them to obey him [the dihqa4n].230 Muslims
were thus acquainted with the Iranian concept that dress codes were not
only used to accord privileges to the upper classes and distinguish them
from their inferior compatriots, but were actually applied to the lower
classes as well. According to this latter statement, the leading principle in
the dress code of the lower class was the manifestation of their humility
or submission. This is especially significant to the subject under discussion here, since it indicates not only that the concept of the ghiya4r was
Iranian in origin, but that its purpose in this case was to exhibit the low
and subjected status of the non-Muslims the lower class before the
Muslims, who were now regarded as the upper classes of society.
In fact, Muslim tradition recounts that not only was there a different
dress code for the different classes, but there were also internal dress
codes within the classes. Thus, al-Masu4d3 reports that the dihqa4ns
branched into five groups according to their rank, and their dress varied
according to the might of their rank.231
Like the Iranians, the Muslims reserved certain items of clothing
for the upper class. Headgear was considered very significant. Ha4ru4n
al-Rash3d was displeased when a certain poet entered his presence
wearing a qalansuwa taw la and soft shoes (khuffa4n dumalika4n).232
The qalansuwa taw la was the heir of the tall rounded hat worn by the
Persian dignitaries from the Achaemenid period onwards.233 It is no wonder, therefore, that the question of headdress is central in the instructions
regarding the attire of non-Muslims.
148
Abu4 Yu4suf recounts234 that Umar b. Abd al-Az3z prohibited both the
turban and other prestigious dress items for dhimm s, among them the
qaba4, the mans gown the Persian kapa4h,235 silk clothes,236 the mana4tiq
(belts, on which more below), and the special hairstyle reaching the shoulders, all of which were status symbols of the privileged upper classes and
apparently now worn by the non-Muslims of his day, inappropriately
of course.237
Another article of clothing that was to be exclusively Muslim was
the tira4z. The Samaritan chronicle, enumerating al-Mutawakkils restrictions, mentions that he ordered that no one should wear tira4z238 i.e.
a garment with an embroidered edge. The tira4z was an unmistakable
symbol of royal power alongside the minting of coins and the seal, in
both the Byzantine and the Sasanian empires, and was perceived as such
among the Muslims too.239 Ibn Khaldu4n, in his chapter The characteristic emblems of royal and government authority, devotes a section to
tira4z. He notes that it originated with the pre-Islamic non-Arab rulers,
and that royal garments are embroidered with such a tira4z in order to
increase the prestige of the ruler or the person of lower rank who wears
such a garment. It is thus evident that having absorbed elements of the
Iranian social ethos, the Muslims now wanted to reserve all the dress
items of the Iranian upper class to themselves, and wished to prevent
members of the new lower class, the non-Muslims, from possessing them.
Another matter particularly significant to our discussion here is the
existence of a social colour code in Iranian society. According to the
Iranian cosmogonic myth as told in the Bundahishn, Ohrmazd wore
white, the garment of the wise, signifying the estate of the priests; the
warriors had a golden dress; while the husbandmen were distinguished
by a dark-blue dress.240 According to another tradition, the warriors wore
red rather than gold.241 There is no need here to dwell on the ancient tradition of the purple, which was exclusive to royalty all over the ancient
Near East as well as in the Classical world.242 As already noted above,243
red shoes were an exclusive privilege of Iranian royalty. Hila4l al-Sa4b3 also
notes that it was undesirable to enter the residence wearing red sandals
or shoes, because red is the color of the caliph and of those who rebel
against his authority.244
This colour code245 was, no doubt, the inspiration of al-Mutawakkils
edict stating that non-Muslims were obliged to distinguish themselves
through colour. According to this edict non-Muslims should wear yellow
(asl ) hoods (taya4lisa), and turbans (ima4ma), the women should wear
yellow mantles (iza4r), and their slaves should have yellow patches onthe

front and the back of their outer garments (on al-Mutawakkils edict
see above, Chapter 4).246 When reporting about al-Mutawakkils edict
the Samaritan chronicle records that that al-Mutawakkil ordered that
the people should wear ghiya4r, except for the black and blue which he
reserved for his faith.247
It is interesting to note that yellow, the colour that was assigned to
the non-Muslims, was considered a bad colour in Muslim tradition. Ibn
al-Washsha4 in his Kita4b zarf al-zurfa4 emphasizes that not only was it a
colour used by women, including songstresses and female servants, but
that it was used by them during times of impurity, such as menstruation and illness.248 Al-Tirmidh3 in Abwa4b al-liba4s cites a had th that the
Prophet prohibited clothes of low quality as well as yellow-coloured
clothes.249 Al-Mutawakkil exchanged the prohibition on dhimm s wearing ama4im for a requirement that they should be honey-colored, an act
that evidently emphasized the significance of the colour distinction.250
The use of colour as a means of distinction among the social classes
was thus an ancient concept, which was adopted and adapted by
the Muslims. It was not invented specifically for the dhimm s, nor was
it used only in their case. The Muslim colour code was less defined than
the Iranian code; it certainly was not in use on a regular basis in order to
signify position or class within Muslim society, yet, as mentioned above,
it could be used as a family or political marker, and of course to designate
the caliphal court, on one hand, and the dhimm s, on the other. In the
case of the non-Muslims, as in Iranian society there was a certain colour
that marked a distinct group, defined this time not according to agnatic
criteria or social position, but according to religious identity. The colour
identified non-Muslims in Islamicate society, and was thus another means
through which they were positioned in its hierarchy.
The last issue mentioned in this clause is the kunya. Patronyms were
generally a respectful form of address, and were a considered a basic
expression of respect and of honour. Abd al-Razza4q cites a tradition
against the use of kunyas for non-Muslims since they should not be
glorified by a kunya.251 Ibn Asa4kir cites another tradition in the name
of Damra b. Hab3b,252 according to which Umar said regarding ahl
al-dhimma:Call them by their names and not by their kunyas, humiliate them but do not wrong them, and if the road brought you and them
together, push them into its narrowest part.253
As noted above, one of the means of humiliating mawa4l was to forbid
them to use kunyas.254 Although the prohibition seems never to have been
effective, as noted by Goldziher,255 it nevertheless indicates that this was
150
a form of social degradation. This stress on humiliation becomes stronger in sources from the Abba4sid period. Hila4l al-Sa4b3 reports that it is
not customary to call anyone by his kunya in the presence of the caliph,
unless the caliph honored him by calling him thus, and entitled him to
that rank.256 Al-Tu4s3 says that when one is grown, people treat him like
a man and call him by his kunya. Later when he has shown merit and
skill in public life the king bestows upon him the honour of a title.257 The
kunya thus serves as the basic title of respect in adult society. Slaves had
no kunyas. The phrase we shall not adopt their kunyas or literally we
shall not be called by their kunyas may be understood in more than one
manner:we will not use kunyas at all; or we will not use Arabic kunyas.
In both cases it is clear that non-Muslims were not meant to draw from
this patronymic element the basic respect it bestowed on its owner.
Note that there is a direct analogy between the behaviour of a nonMuslim in the presence of a Muslim and the behaviour of a Muslim in
the presence of the caliph. Unlike the other elements in this clause, this
does not seem to stem from status symbols of the Iranian or Byzantine
cultures, but rather from Arabian social tribal mores.
Clause 7:The Prohibition on Using Saddles, Girding Swords,
and Carrying Arms
From ancient times,258 horses and the equipment that accompanies horseriding was the exclusive prerogative of the nobility in Iran. In fact, the
name of the warrior class, arte tara4n, means charioteers,259 and the
word asba4r/aswa4r, horseman, actually designated knight or nobleman,260 while the term for foot-soldier came to mean commoner. By
contrast, the Roman equestrian class was considered inferior to the nobility. In Iran, a fine horse was the best gift that could be given; an Iranian
noble was inseparable from his mount, and going on foot was considered undignified. Among the presents given out by Cyrus, according to
Xenophons Cyropaedia, were robes, bracelets, necklaces, and horses
with gold-studded bridles. For as everybody knows, no one over there
is allowed to have such things except those to whom the king had given
them.261 According to Xenophon, all of these items were status symbols
that separated the nobility from the common people. During the Sasanian
period the saddle became more sophisticated and ornamental, and thus a
highly prestigious article.
The Iranian perception of horses and riders was known to Muslim
writers. Al-Jahshiya4r3 reports that in the time of the Sasanians no one
but the king, the kutta4b, and the judges were authorized, to ride on gentle
Illustration 3. A silver plate, with the Sasanian Sha4pu4r II hunting boars, riding his saddled, richly decorated horse in full regal attire including a bejewelled
belt and dagger (from Vereino, Perm, Russia; fourth century AD. Freer Gallery of
Art, Smithsonian Institution, Washington, DC:Purchase, F1934.23).
and steady horses (hama4l j).262 Al-Bala4dhur3 mentions that the dihqa4ns
rode a mount and wear golden rings on their feet, both evident status
symbols.263
Although horse-riding was adopted by the Arabs in the Arabian peninsula in about the fifth century CE,264 such formal decorum was unfamiliar to them. However, following the conquest, and due most probably
to Iranian cultural influence, this prestigious riding culture was fully
adopted by the Muslims and buttressed by Qura4nic passages. Hila4l
al-Sa4b3 recounts an occasion upon which a Byzantine envoy arrived at
al-Muqtadirs court: The soldiers of different ranks and in excellent
attire, were drawn up in two lines and mounted on animals with saddles
152
of gold and silver; and near them were the reserve horses in similar elegance, displaying many types of arms and equipment.265 Horses with
splendid saddles were also given out as gifts, as they had been in former
times.266 The privilege of riding a horse into the palace was granted only
to the vizier, and that only up to a certain point.267 When the Ghaznavid
sultans appointed important officials the masters of ceremony would call
out:Let the horse of the am r such and such be brought forth, thus publicizing and glorifying the appointment.268 One cannot escape the similarity to the event described in the book of Esther, which took place around
1,400 years earlier:when King Ahasuerus wanted to promote Mordechai
he bestowed on him his robe, and one of his horses; Haman, while leading the horse, called out:Thus shall be done to the man whom the king
wishes to honour (Esther 6:9). Dismounting, or walking behind a persons horse on foot, was a sign of great respect for the person in question.
Riding a mule instead of a horse was considered humiliating.
In the light of all this, it is quite clear that the insistence that dhimm s
should not use saddles, only pack-saddles,269 that they should use wooden
instead of iron stirrups,270 or not ride horses at all, only mules,271 is not
simply a way of distinguishing physically between Muslims and nonMuslims, but an unequivocal statement that non-Muslims, however rich
or respectable, could not employ the prestigious insignia of the nobility
or upper class.
This is also the issue behind the prohibition on girding swords or
carrying arms. As early as in Achaemenid Iran, girding a sword or a dagger in ones belt was a distinctive sign of nobility. Xenophon recounts how
the Persian nobles arrived at court on their horses, with their spears.272
Daggers or swords were an indispensable part of the attire of Iranian
nobles throughout time.273 In the caliphs court too, riding a horse and
wearing a sword was a status symbol, employed by all those who surrounded the caliph.274 The special sword, al-sayf al-kha4ss, was considered, in fact, one of the insignia of sovereignty, and was therefore often
studded with gold and jewels.275
Rather than being an issue of security, therefore, wearing a sword or
dagger should be considered a symbol of social status in general, and at
the royal court in particular.
Clause 8:The Prohibition on Having Seals Engraved in Arabic
The seal or signet-ring khatam, an Aramaic word was an ancient
symbol of authority.276 A seal gave validity to a document, and the
possession of another persons seal represented the delegation of that

persons authority. Seals were widely used in the ancient Near East from
very early times. When the pharaoh appointed Joseph over the whole
land of Egypt, he gave Joseph his signet, dressed him in splendid robes,
put a gold necklace around his neck, and made him ride in the second
chariot (Gen. 41:42). When Ahasuerus wanted to delegate to Esther his
authority to retract his decree to annihilate the Jews in his kingdom, he
said:Write as you please in the name of the king, and sign it with the
kings seal, since a decree written in the name of the king and sealed with
the kings signet-ring cannot be retracted (Esther 8:8). The importance
of the seal is well attested by Menander Protector in his description of
the signing of the 561 treaty between the Byzantines and the Persians
(see Chapter 1).
Seals were also adopted by the Muslims. It is interesting to note that
the earliest seals in Arabic date to the period immediately after the conquest.277 In fact, Muslim tradition recounts that the Prophet first adopted
a seal when he wanted to write to the Byzantines and they would not
accept his letters unless they were signed with a seal.278 Bearing the royal
seal was a sign of royal authority. Ibn Khaldu4n says that the [use of
the] seal, which is the written signature or engraving used for closing
and tying letters, was peculiar to the ministry of correspondence (d wa4n
al-rasa4il). In the Abba4sid dynasty, it belonged to the waz3r.279
Having a seal in Arabic was therefore another emblem of authority,
like the horse, the saddle, the girding of the sword, and various articles of
clothing and other paraphernalia, and for this reason it could not allowed
to a non-Muslim.
Clause 10:The Requirement to Clip the Front of the Hair
This clause represents the requirement to cut the hair across the forehead
short, and needs to be attached to clause 6 regarding the hairstyle. As
noted above,280 Muslims used to wear their hair down to their shoulders and apparently parted in the middle, a hairstyle attributed to the
Prophet.281 This hairstyle also appears in representations of Iranian nobility and was, it seems, an inseparable part of their dress code.282
Cut forelocks mark the status of the dhimm s as freed captives. The
custom of cutting them is attested in early Arabic sources. Arabian tribesmen who took captives and set them free cut their forelocks first, thus
displaying their generosity and their expectation of reward.283 The nonMuslims were conquered people who owed their freedom to their Muslim
captors, and were stigmatized by their cut forelocks. A dhimm is therefore sometimes called muqassas (one whose forelock has been cut off).284
154
Clause 11:The Requirement to Dress the Same Way Wherever

We May Be and to Bind the Zunna4r around Our Waists
This is adduced in detail by Abu4 Yu4suf, who says:It should be decreed
that none of them should resemble the Muslims in his dress, in his mount,
and in his appearance and that they will be made to wear the zunna4rs
around their waists, like a coarse cord which each of them should tie
around his waist.285
In al-Mutawakkils official decree, the command is that all those
of this class (tabaqa) who wear the mintaqa should wear zana4n r and
kasa4t j instead of the the mana4tiq that were [till now] around their
waists.286 Three different terms are used here for a belt:mintaqa, which
the dhimm s are forbidden to wear, and zunna4r and the kasat j, which
they are now obliged to wear. What is the purpose of this decree?
The zunna4r (pl. zana4n r), the girdle or waist-belt, was considered the
most distinctive sign of dhimm status. This, as well as additional special
elements of clothing such as the variegated hats (qala4nis mudarraba), and
doubled shoelaces (shira4k muthniya), were first mentioned by Abu4 Yu4suf
in the last quarter of the eighth century.287 As already noted in Chapter3,
Umar IIs edict does not use the term zunna4r, but mentions that they
have abandoned the belts, using the term mintaqa rather than zunna4r.288
This means that while the idea of using a belt to distinguish the nonMuslims from the Muslims was already in use in his time, the terminology was not yet established.
As Tritton said, the word zunna4r is derived from Greek zo4narion, the
diminutive form of zo4ne 289 belt in Greek. The word was adopted probably, however, via the Syriac zu4na4ra4.290 Tritton noted that monks wore
such girdles around their waists, and held that this was why the zunna4r
became the identifying mark of the Christians, and eventually of all the
non-Muslims. If so, the zunna4r could be conceived as a mere distinguishing element which had nothing to do with social status:it was simply an
article of clothing used by a particular group of Christians which was
arbitrarily adopted as the emblem of all non-Muslims.
Religiously affiliated belts, however, are attested not only among
Christians, but among Jews in Babylon and among the Zoroastrians as
well. Among the Zoroastrian believers a belt called kust g was an obligatory item of dress. Its purpose was to separate symbolically between the
upper part that of thought and speech and the lower part, where
the baser bodily functions take place.291 According to chapter 38 of the
Da4desta4n de n g, cited by Shaked, the kust g is a sign of human servitude to God (Dd 38:2) The merit of humility in the presence of the

deity is highly extolled (Dd 38:5).292 Passage 11 says that it is also not
fitting to walk about ungirdled without a mark of servitude to the lord
(Dd 38:11).293
Babylonian Jews who lived under Sasanian rule also wore belts that
had religious significance, and were used in a similar manner to that of
the kust g, although there is no evidence that they were actually obligatory, as the kust g was. This belt was called hamya4na4, an Aramaic word
borrowed from Persian.294 Babylonian Jews also adopted the ritual of
the kust g practised by the Persians. Thus, according to the custom in
Babylon (but not in Eretz Israel), one starts eating after he has undone
his belt.295 The discussion that follows demonstrates that Jews observed
the girding of the hamya4na4 before praying.296 This was accompanied by a
special blessing. Thus, in the Babylonian Talmud (Bavli), Berachot 60b it
is said:When someone ties his girdle, he should say:Blessed be He who
girds Israel with valiance.
Until now we have seen that religiously affiliated belts were customary not only among Christians, but among all three confessions during
the Sasanian period. In all three this was a widely used accessory which
exhibited the affinity of the believer to his God. These belts (including the
hamya4na4, which most probably resembled the kust g outwardly) could
indeed serve as characteristic items which conveniently distinguished the
non-Muslims in Syria, Iraq, and Persia.
While Umars edict indicates that the requirement of the zunna4r was
established at the beginning of the eighth century, the terminology distinguishing between the two developed later that century, when the Muslims
adopted the terminology used by the non-Muslims for their own belts,
and designated the Arab word mintaqa solely for the belts of the Muslims.
According to V. Loukonine and V. Ivanov, Persian Art, London
2003, item no. 38, the round buckle and two hasps of this gold bejewelled Iranian belt-buckle are in the polychrome style characteristic of
the fourthearly fifth centuries The engraved cornelian gem on the
buckle depicts a rider at a hunt. What is the mintaqa that the nonMuslims were prohibited from wearing?
Widengren claims that in ancient Iran there were two different kinds
of belts. The first was the kust g or the hamya4na4. The second was the
kamar or kamar-band. In fact, both the zunna4r/kust g and the kamar
began as signs of servitude:the first servitude to God, as noted above; and
the latter servitude to ones master. In fact, the word band in kamar-band
comes from bond, fetter, and is actually the belt of the bandak, Middle
Persian for (loyal) servant.297 The kamar (and the earrings) was the
156
Illustration 4. Belt buckle. Gold, cornelian and almandine; decorated with

granulation. 5.2 x 3.7 cm Iran. secondthird century. Inv. No. Z_436/1,2,3 (The
State Hermitage Museum, St Petersburg. Photograph The State Hermitage
Museum/Photo by Valdimir Terebenin, Leonard Kheifets, Yuri Molodkovets).
distinct symbol of the binding of the nobleman to his king. All bandaka4
who entered the kings presence had to wear their belts, and all stood by
him with their hands on their belts.298 It was thus the sword belt worn by
warriors of all levels; kamars studded with pearls and gems were given by
Parthian and Sasanian kings as a mark of distinction.299 In the Ka4ba-yiZardusht inscription, line 4,300 the priest Karte4r declares that he received
such a kamar from the king. The Arukh a Gaonic Jewish source notes
that under the Persian kings Jewish Exilarchs used to wear kamars.301
The clearest distinction between the hamya4na4 and the kamar appears
in the Babylonian Talmud, Shabbat 49b, where it is asked whether one
could on Shabbat wear a kamar on top of the hamya4na4, the answer being
that one cannot wear two hamya4na4s! The kamar is described here as a
gilded ornamental belt, made either of fabric or of metal, and is likened
to a kings belt.302
As noted above, such belts were worn with daggers by the Iranian royalty and nobility. An Iranian noble would not be caught without his belt
in public. Dihqa4ns, being part of the nobility, wore such belts. On one
occasion, when the Muslims wanted to humiliate dihqa4ns who refused to
convert to Islam, they tore their clothes and bound their belts (mana4tiq)
around their necks.303
A similar belt, called the zone in Greek, cingulum in Latin, was found
in the Byzantine empire. Initially part of Roman military costume, it
spread during the reigns of Diocletian and Constantine to become part
of the dress of every official. The fashion became so widespread, in fact,
that the state tried to restrict its use by civilians. It was at that point,
it seems, that the monks, following the widespread trend, adopted the
use of the girdle, viewing it as a symbol of purity, temperance, and
manliness.304 But it remained a symbol of public office. When Justinian
prohibited the entry of non-Christians into public service, he said:nor

shall he put on an official belt (zone ), neither civil nor military, nor
belong to any office.305
The kamar or cingulum/zone , which served as a mark of status and
distinction both in the Iranian and Byzantine empires, is, no doubt, the
belt that would be called mintaqa in Arabic. In the Muslim empire, the
mintaqa worn with a sword was a typical part of the outfit of soldiers,
the officers mana4tiq being studded with gems, just as the Iranian ones
had been.306 The belt and the sword were part of the official attire. Hila4l
al-Sa4b3 recounts that on procession days, the chief chamberlain (ha4jib)
comes fully attired in black robe and black turban, wearing sword and
belt.307 The caliph himself, according to Hila4l, was girded with the sword
of the Prophet.308 Such decorated belts were considered not only a prestigious gift, but an emblem of authority. Thus al-Mutasim presented
al-Afsh3n with a crown (tawwaja), and girded him with two jeweled
belts,309 and installed Ashna4s on a throne, awarding him a crown and a
ceremonial girdle (wa-tawwajahu wa-washshahahu).310
It is thus evident that the Muslims adopted this ancient distinction
between the official belts of the nobility, which marked the high position of their owners, and the sacred girdles, which designated confessional identity and were worn by people of all strata of society among
Zoroastrians and Jews, as well as by monks, who were both numerous
and highly regarded in Christian society. The mana4tiq, which were the
privilege of the Sasanian upper classes, became the privilege of the new
upper class:the Muslims. The girdles, which had been worn in the past by
the believers of various confessions of their own will, had become obligatory, and served to distinguish the non-Muslims from the Muslims, who
wore no such girdles.
The imposition of the zunna4r was therefore accompanied by the prohibition on wearing a mintaqa, a prohibition which was no doubt relevant to many non-Muslims who had previously worn them to mark their
distinction, and wanted to go on doing so. The zunna4r (kust g/hamya4na4),
the sacred girdle, common among believers of various confessions, had
now become a manifestation of the new social order, a distinctive mark
of the newly created lower class.
Clause 12:The Prohibitions on Processions, Clappers, Raising
of the Voice, and Showing Lights
This clause is made up of several parts
(1) The prohibition on conducting religious processions:We shall not
display (or in other versions take out) our crosses or our books
158
in the roads and markets of the Muslims we shall not go out

on our Easter or on Palm Sunday (la4 nakhruju ba4u4than wa-la4
shaa4n n).
(2) The prohibition on using clappers (nawa4q s) loudly.
(3) The prohibition on the raising of voices in church services or in the
presence of Muslims.
(4) The prohibition on the raising of voices in funerals, the prohibition on showing lights in Muslim roads or markets.
(1) The prohibition on conducting processions and that on showing
lights both need to be seen in the general context of the mawkib the
procession or retinue of the ruler. Such processions had been part of
the ceremonial protocol in the Achaemenid court. Xenophon recounts
the passing of Cyrus procession:the king rode in a chariot wearing his
tiara, dressed in purple and followed by a magnificent procession, while
soldiers and cavalry dismounted and stood along the road, their hands
thrust through their sleeves to signify their submission, just as they do
even to this day when the king sees them.311
As mentioned above, this custom, although criticized, was adopted by
the Umayyads, along with the strict protocol observed when accompanying the ruler, and especially the concept of the tart b, the arrangement
according to hierarchical order, at the court and in such processions.312
Christian religious processions, and especially those of Palm Sunday
and Easter, were and still are highly formal and ceremonial in their
nature. In these processions items such as holy icons, banners, and holy
books are carried in public in an exhibitive manner. According to Abu4
Yu4suf, in his chapter On the prayer houses and crosses, these processions included not only banners313 (probably embroidered with icons
or saints), but arms as well. Abu4 Yu4suf has them agree to a surrender
agreement that:
They will not take out their banners (ra4ya4t)314 on their holidays, that
they will not wear arms on their holidays, nor keep them in their houses.
If they do any of these things, they will be punished and it [the peace
agreement] will be taken from them. These were the conditions of the
agreement (sulh); but they said to Abu4 Ubayda:Set a day in the year
for us on which we can take out our crosses without banners, namely
the day of our great holiday.315
Banners and flags were considered a distinctive sign of authority under

Sasanian and Muslim rule,316 and Ibn Khaldu4n lists them among the royal

insignia, along with other elements of the a4la.317 Military commanders received a white liwa4 (standard, banner) on the occasion of their
appointment.318 The liwa4 was, in fact, an old symbol of royalty going
back to Parthian and Sasanian times.319 Hila4l al-Sa4b3 mentions among
the expenditures of the caliph the costs for carrying the standards during
the Two Feasts (i.e. I!d al-Fitr and I!d al-Adha4).320
As a distinct sign of royal authority, banners and arms were to be
excluded from Christian processions, even according to the liberal
Abu4 Yu4suf, who, unlike the adherents of Shuru4t Umar, supported the
endorsement of the Easter procession.
A magnificent procession organized by someone who was not authorized to do so could well be considered an infringement upon royal
authority.321 Hila4l al-Sa4b3 tells a story about a qa4d who happened to
pass by a grand procession of one Na4zu4k on his way to see the waz r.
Having been delayed by the procession, he was late and had to apologize to the waz r and explain, although he feared that Na4zu4k would be
reprimanded for allowing himself such splendour and pomp.322 Luckily,
Na4zu4ks behaviour was excused by the waz r, who said that in this magnificent procession Na4zu4k had in fact graced the state and Islam, and
spited the nose of the blasphemers and the deviators.323
Ceremonial public processions were therefore considered a manifestation of sovereignty, and thus could not be a privilege of the subordinate
non-Muslims.
The prohibition on showing lights in Muslim roads or markets should
be interpreted in the context of the procession. Na4zu4ks procession is
described as follows:He was accompanied by more than 500 attendants
carrying processional candles, and by a larger number of naphta-candle
carriers.324 Showing lights in Muslim roads was conceived as a characteristic element of processions, and thus a manifestation of royal authority.
(2) The prohibition on using the clappers loudly: beating drums, as
well as blowing trumpets or horns, were also royal prerogatives. In the
Sasanid army the general (spa4badh) was the only one allowed to enter the
camp to the blowing of the trumpets.325 In the Abba4sid court this was
done on occasions such as the emergence of the caliph from his court, or
when a procession accompanying one of the dignitaries left the court;326
in addition, it was used as a call to prayer. Hila4l al-Sa4b3 reports that
it was not customary in the past to beat the drums for prayer in the
capital except for the caliph. It was later permitted for the crown princes
and the commanders of the armies during the three prayers the early
160

morning and the two evening prayers if they were traveling or were far
from the caliphal presence When Muizz al-Dawla came to power he
wished to have the drum beaten at his door in the City of Peace. He was
then living in the palace of Munis, close to the Residence. He sought
the permission of al-Mut3 li-Allah to do so. Although they hardly
disagreed, the caliph did not grant him permission because it was not
customary.327
Later, special permission to beat the drums for prayers was given to certain dignitaries in specific circumstances.328 Drums, trumpets, and other
instruments were also played on religious festivals, and other feasts,
as well as on the rulers birthday. The beating of kettle-drums in honour of the ruler and on the occasion of appointments to high positions
was an ancient practice.329 Ibn Khaldu4n notes in fact that the Muslims
refrained from beating the drums and blowing the trumpets at the beginning of Islam. They wanted to avoid the coarseness of royal authority
and do without royal customs The caliphate then came to be royal
authority, and the Muslims learned to esteem the splendor and luxury of
thisworld.330
The fact that beating the drums, especially for prayer, was a prerogative of the caliph clearly meant that the beating of the na4qu4s loudly and
in public was an infringement on royal authority. This sheds light on the
instruction to beat it quietly inside. In the same vein the Samaritans were
prohibited from using the horn (bu4k) during their ritual in the reign of
Ibn Tu4lu4n, according to the Samaritan Chronicle.331
(3) The prohibition on raising the voice in church services (prayer
and or readings) in the presence of Muslims:332 in Iranian culture, raising
ones voice was considered an expression of coarseness and vulgarity.
Xenophon reports in the Cyropaedia that Cyrus wanted to set an example of refined and honourable manners. By setting such an example
Cyrus secured at court great correctness of conduct on the part of his
subordinates, who gave precedence to their superiors And among
them you would never have detected anyone raising his voice in anger
or giving vent to his delight in boisterous laughter; but in seeing them
you would have judged that they were in truth making a noble life their
aim.333 Al-Masu4d3 reports that when the shah came to sit with his courtiers (nudama4), they were ordered to keep silent and moderate their gestures.334 Pseudo-al-Ja4hiz instructs that in the presence of the king a man
should not rush in his speech, nor point with his hand, nor raise his voice
(la4 yarfau sawtahu), move his head, move in his seat in any manner, or
meet the gaze of anyone present but the king.335

Hila4l al-Sa4b3 notes that when the waz r speaks or converses in the
presence of the caliph his voice must be low, and he should not raise it
except as necessary to prevent repetition or clarification.336 During an
audience of the caliph, it is also the rule for the people not to speak, and
that neither sound nor clamour should be heard from them.337 Al-Ghaza4l3
mentions that one should not raise ones voice in the mosque.338
The prohibition in the Shuru4t on raising the voice pertains to the
prayers of non-Muslim in the presence of Muslims, presumably present
in the precincts of the church. It is therefore evident that the behaviour
required was that of the Muslims in the presence of the caliph and
inthe mosque. The Muslims did not want to be offended by the noise
and clamour that the non-Muslims raised during their rituals. This is
true also for the following clause, prohibiting the commotion created
during funerals.
It is interesting to note that the Samaritan chronicle testifies that under
the restrictions of al-Mutawakkil it was decreed that a man shall not
raise his voice in prayer, and shall not raise voice in ,339 a funeral shall
not be seen, and a dhimm shall not lift his face to a Muslims face in
order to speak or respond to him.340 This is repeated again later, under
Ibn Tu4lu4n: A dhimm should not raise his head in the presence of a
Muslim, and he should not raise his voice in prayer.341 Thus, raising the
voice in prayer, looking straight at a Muslims face while speaking to him,
or raising ones voice in his presence was considered insolent and disobedient behaviour, just as it was in the presence of the caliph.
The Muslim here is in fact equated with the caliph. In other words,
for a non-Muslim every Muslim is a caliph or a king, and should be
treated with appropriate reverence and respect. This clearly goes back
to the Iranian hierarchical concept according to which it is fitting that
the working people should salute them [the warriors] and bow before
them, and that the fighting men in turn should show reverence to the
nobles, and that they should have regard one for another according to
the loftiness of their rank, and that they should maintain their dignity.342
Thus, just as the Muslim should not raise his voice in front of the caliph,
so the non-Muslim should not raise his voice in front of the Muslim.
(4) Prohibition on the raising of voices in funerals: rules prohibiting noise, lamenting, wailing, songstresses, slapping oneself and other
customs such as plucking the hair or rending the clothes, which were
common in the ja4hiliyya, were forbidden in Islam in theory. Muslim
shar a regarding funerals drew its inspiration in this case perhaps from
the rejection of ja4hil customs. However, although forbidden in Muslim
162
fiqh literature, with the exception of the Zoroastrians, lamentations and

wailing continued to be a central element of the funerary ceremonies in
the ancient Near East.343
Since this was not really accepted or abided by, it seems that the requirement that non-Muslims have quiet funeral processions was a result of the
general principle that rituals and customs of the dhimm s should not
offend Muslim feelings or contaminate the public sphere. While Muslim
rituals were a public issue, those of the dhimm s were tolerated only if
they were performed quietly, behind closed doors. Just as non-Muslims
had to make way for the Muslims, give up their seats for them, and speak
quietly, so they had to keep their public rituals as invisible as possible so
as not to offend the Muslims or threaten their superiority.
conclusion
In his article on la4 tashabbahu4 Kister, using the example of the change
in attitude to wearing shoes while praying, noted that the Muslims were
actually going through a fundamental change in which customs initially
frowned upon as an imitation of the unbelievers were actually adopted
as the only form of behaviour.344 This was in essence what was happening in the process described here. Shuru4t Umar reflects a process in
which Muslim society was redefining itself versus the conquered societies.
During this process, various elements from the ethos and codes of the
conquered were adopted by the Muslims. These adopted codes were then
used to dispossess the non-Muslims of their former place in society, thus
creating a new situation in which the Muslims held the superior position
of rulers in Islamicate society while the non-Muslims were the ruled and
subjected. Though they were protected and had certain rights, their position was considered inferior to that of all Muslims, a fact which had to be
outwardly manifested in their appearance and social behaviour.
Shuru4t Umar was the document that defined this new hierarchy. It
may be that the impetus for creating such a universal legal document was
the process of codification of Muslim law which was based on Byzantine
law, and was taking place in the second half of the eighth century CE, as
has been demonstrated by Jokisch. Byzantine law is indeed reflected in
Shuru4t Umar, and provides precedents for the clauses regarding synagogue building, slave ownership, apostasy and prevention from joining
Islam. Other Muslim laws regarding non-Muslims not found in Shuru4t
Umar, such as those relating to the prohibition on holding public office,
questions of inheritance, testimony, and the defamation of Islam also
originated in Byzantine law regarding non-Christians.

However, Byzantine law supplies the sources for only few of the
clauses of Shuru4t Umar. What forms the heart of Shuru4t Umar are the
clauses that come under the definition of ghiya4r or distinguishing marks,
the roots of which are to be found in the social ideology that prevailed
in Sasanian society before the conquest, an ideology which promoted
explicit distinction and immobility among the four classes of society.
Having adopted this ideology, the Muslims took the place of the upper
classes in Sasanian society while the non-Muslims were allotted the position of the lower class (tabaqa).
The new Muslim regulations, based on the Sasanian adaptation of the
ancient Iranian codes, provided that non-Muslims not only had to dress
the part of the lower class, but also had to play it. They had to clear the
road for Muslims, get up for them, speak to them quietly with downcast
eyes, just as Muslims did when addressing the caliph. This attitude is
well exemplified in Kita4b al-ta4j, attributed to al-Ja4hiz, where there is a
story of a competition between Jar3r b. al-Khatafa4 and al-Akhtal for the
title of Abd al-Maliks poet. When the caliph decreed that the Christian
al-Akhtal had won, he commanded that al-Akhtal get up and mount
upon him. The Muslim poet refused to submit to this, and the reaction
of the crowd was:A han f [non-Muslim] cannot mount upon a Muslim,
and cannot have an advantage over him (la4 yarkabu al-han fu al-muslima
wa-la4 yazharu alayhi).345
In practice, of course, this was much more difficult. Non-Muslims, as
is well known, continued to be employed in high positions in the court,
serving mainly as scribes and physicians, and often acquired the privileges that went along with these positions.
A famous case is that of Bukhtishu4 b. Jurj3s, physician of Ha4ru4n
al-Rash3d, who received from the caliph every Muharram a load of new
robes, gold and silver, tira4z, and other paraphernalia all signifying his
high social position, despite his being a Christian.346 Another is that of
the Jewish personal physician in the Maghrib at the time of al-Ha4kim,
who was taken for a Muslim because he was dressed as a courtier.347
Naturally, non-Muslims who had previously held respected positions
were reluctant to accept the inferior status that was allocated to them
in the new social hierarchy. Nevertheless, as has been demonstrated in
the previous chapter, this social ideology was gradually enforced more
systematically from the ninth century onwards, despite many exceptions
that were continuously made in numerous situations and circumstances.
Conclusion
The definition of the social and legal status of non-Muslims in Muslim

society was a product of a long and complex process which began during
Muhammads lifetime and found expression in his various, inconsistent,
statements regarding ahl al-kita4b to be found in the Qura4n, as well as in
the early surrender agreements that were signed in his lifetime. It was,
however, during the conquests made from 634 CE onwards, when city
after city fell into the hands of the Muslim conquerors, that they began to
adopt a more consistent policy based on the sulhan/anwatan principle
the sulhan, in its turn, being based on the ama4n, or promise of security
given in exchange for the payment of tax by the conquered inhabitants.
This basic code was applied, according to Muslim historiographic tradition, with some additional conditions and commitments that were applied
at times and varied from one place to another. Muslim sources dating
from the end of the eighth century CE onwards often adduce reports
regarding these surrender agreements, and at times include versions of
the agreements themselves.
As demonstrated in the first chapter, such codes were in fact part of
an ancient diplomatic tradition, customary in the ancient Near East from
the middle of the second millennium BCE. Written treaties in general, and
written surrender agreements in particular, are to be found in the ancient
Near East, as well as in the Graeco-Roman world, and in the Byzantine
empire, its successor. Such agreements were specific and detailed, and
included besides major issues such as personal security and the safeguarding of possessions, prayer-houses, ritual and customs in return for
the payment of tax such minute details as the manner and duration
of the evacuation of a certain territory, the number of hostages taken
164
Conclusion
165
in order to assure fulfilment of all the clauses, etc. There was a common international diplomatic terminology for use in the drafting of these
documents.
The accumulated evidence points to an established procedure of surrender: representatives of the surrendering inhabitants would emerge
from the city or stronghold and offer to capitulate if certain conditions
were fulfilled. The general would consider these terms, and negotiations
would take place. After the parties had come to an agreement, the documents would be drawn up, signed, witnessed, and sealed. Copies of these
documents would be kept for reference by both parties.
There is thus not only no reason to doubt the reports regarding the
specific surrender agreements adduced by the Muslim sources, but there
is sound ground to support their acceptance as authentic documents
which reflect an established procedure of surrender. In particular, there
is no reason to mistrust the detailed agreements reported by the sources
during the Muslim conquest, since these followed a long and established
tradition with which the inhabitants of the surrendering settlements were
thoroughly familiar, and in fact had most probably employed just two
decades earlier, during the Persian conquest.
The surrender agreements were suitable to both parties at the time
of the conquest: they prevented further war and bloodshed for both
sides, allowing the local inhabitants to carry on with their lives, and the
conquerors to continue their swift takeover, guaranteeing their rule and
income, keeping the conquered territories intact and functioning, and
demanding only minimal effort on their part.
It was only some time later that this situation was found to be unsuitable. When the Muslims settled down in cities and areas previously
inhabited only by non-Muslims, some of the stipulations previously
agreed upon were found to be offensive and inappropriate. Somewhere
around the end of the first century and the beginning of the second
century of Islam things had changed to such an extent that a need to
redefine the status of the non-Muslims arose. The individual surrender agreements not only provided the conquered peoples with mostly
liberal conditions that often infringed upon Muslim superiority, especially where public space was involved, but also formed an inconsistent
and sometimes contradictory body of documents containing varying
conditions and regulations concerning the non-Muslims, a situation
which could not be tolerated for long. One set of rules needed to be
formed and applied to all non-Muslims. The redefinition of the status
of non-Muslims involved a significant discussion of such questions as
166
the continuing validity of the surrender agreements, on the one hand,

and, on the other, the right of the Muslims to impose restrictions upon
the non-Muslims living under their rule, such as for example on their
right to conduct public processions on holidays, to display crosses, to
sell pigs or wine, or to beat the na4qu4s in call for prayer. This discussion, which took place in the course of the early ninth century CE, is
reflected in the juristic sources; especially significant are three general
documents which attempt to redefine the status of all non-Muslims
under Muslim rule without differentiation or variation: the first is
found in Abu4 Yu4sufs Kita4b al-khara4j; the second in al-Sha4fi3s Kita4b
al-umm; and the third is the well-known Shuru4t Umar. Although all
three have traditionally borne this title, in fact only one is the canonic
Shuru4t Umar. True, there is substantial common ground between the
three; they all emphasize the importance of the ghiya4r and accept the
ideology of the superiority of Islam which it is meant to convey. But
the first two texts represent in certain cases different, and more lenient,
attitudes to non-Muslims.
As regards the origins of Shuru4t Umar an exhaustive search for a
similar legal document regarding the status of a minority or subordinate
group in earlier or contemporaneous Mediterranean or Near Eastern
societies produced no result. In this respect, therefore, Shuru4t Umar
seems to be a unique document. Nevertheless, there were separate rules
and regulations in various legal traditions pertaining to certain minorities or groups in society such as the metoikoi in Athens, the gerim in
Israel, and the Jews and Samaritans under Byzantine rule. Sasanian policy
towards non-Zoroastrians seems, on the other hand, to have been often
less regulated. A few of the clauses in Shurut Umar such as the right to
keep synagogues and churches in their original state but not to renovate
them, the prohibition on the conversion of Muslims to other religions, or
the taking of slaves allotted to Muslims, have parallels in Byzantine law.
Clauses regarding the selling of wine and pigs, the display of crosses or
polytheism have to do with the defamation of Islam, and may well have
been inspired by similar clauses regarding the defamation of Christianity
in Byzantine law.
In addition, there are several additional Muslim laws regarding nonMuslims that have close parallels in Byzantine laws, but which do not
appear in the Shuru4t, most probably because these laws were a Muslim
rather than non-Muslim responsibility. These include issues such as the
holding of public office, the testimony of non-Muslims, the holding of
slaves, and the right of inheritance between Muslims and non-Muslims.
Conclusion
167
The bulk of the clauses in Shuru4t Umar, however, pertaining to the

presence, dress, appearance, and behaviour of dhimm s in the Muslim
public sphere cannot be traced back to such laws. They include such
issues as the obligation to show respect to Muslims and give them priority in seating, the requirement not to resemble Muslims in dress, hairstyle, speech, or kunyas; the prohibition on using saddles and on bearing
swords or carrying arms; the prohibition on having seals engraved in
Arabic; the demand to clip the front of the hair, to dress in a certain manner and to bind the zunna4r; the prohibitions on conducting religious processions, especially on Palm Sunday and at Easter, on using the clappers
loudly; on the raising of the voice in church services or in the presence of
Muslims, in funerals, or on holding processions accompanied by processional candles or torches in Muslim roads or markets.
This central part of the Shuru4t was, in fact, an expression of a new
social order that evolved in the course of the eighth century, a social
order whose roots were deeply entrenched in Persian social ethos and
mores. The status of the non-Muslims in Muslim society was now defined
according to the social ideology that had existed in Sasanian society
before the conquest. According to this ideology, society was divided
into four hierarchical classes with no mobility between them. Articles of
appearance and modes of behaviour distinguished clearly between these
classes, and prevented any confusion or vagueness. Thus, ancient Persian
symbols of status and sovereignty such as elements of dress and paraphernalia, riding-beasts and riding-gear, hairstyle, titles, respectable seats,
and any other manifestations of authority and status were forbidden to
non-Muslims, and were appropriated exclusively by the Muslims. NonMuslims, in contrast, had to wear distinctive clothes which suited their
inferior position in society; just as in Sasanian society the lower class had
to dress in a manner reflecting their low social station. The concept of
ghiya4r or distinguishing marks was in fact an established principle in
Persian society, where a visible and general distinction had to be made
between men of noble birth and common people with regard to horses,
clothes, ornaments, houses and gardens, women and servants, drinkingplaces, sitting- and standing-places.
The Muslims had therefore adopted concepts, values, and status symbols from Sasanian society, and used them as a means of establishing their
own superiority. Shuru4t Umar thus reflects the completion of a process
in which Muslim society was redefining itself versus the conquered societies. In this new order, the Muslims took the place of the upper classes
in Sasanian society the priests, the warriors, and the scribes while the
168
non-Muslims were allotted the place of the lowest caste in society:the

peasants.
It should be remarked here that although it also possessed some of the
relevant status symbols at the time, Byzantine society was not distinctly
hierarchical, and dress and appearance played a much less central role
than in Sasanian society. There was comparative social mobility, and status symbols did not emanate from the class one was born into, but rather
from the position one managed to acquire. They were thus part of the
office rather than the social class.
The process of adoption of this Sasanian ideology began with Umar
b. Abd al-Az3z, who took the initial steps and issued the first edict listing
a set of demands regarding the appearance of non-Muslims in public, the
ghiya4r, as parcel of his adoption and implementation of the ideology of
the Chosen People. According to this ideology, as a result of the gift of
Islam the followers of Muhammad were exalted to this superior position
in the world, while the others were to be deprived of the bounties and
advantages they had previously possessed, and were to be degraded and
humiliated. Umars ingenuity is manifest in the idea that he appropriated
the Sasanian social dress and behaviour code not to distinguish between
strata within Muslim society, but rather to highlight the exclusion of the
non-Muslims. Unlike many other rules, this social code became unanimously accepted, and is the one issue that is in total agreement between
Abu4 Yu4suf, al-Sha4fi3, and the Shuru4t.
The first series of regulations in the spirit of the Shuru4t was the set
of restrictions issued by the caliph al-Mutawakkil. New evidence based
on a Samaritan chronicle from the early Muslim period, in addition to
well-known evidence in other sources, show that these regulations were
renewed and enforced with some frequency between the reigns of alMutawakkil and al-Ha4kim, at least in lands controlled by the caliph and
the rulers of Egypt. Although the attempt to impose the Shuru4t was only
a partial success in some cases, the attempts show that it was increasingly
coming to be seen as a set of authoritative rules which had to be observed
in practice.
On a more general level, the conclusion that may be drawn from this
study is that the Muslims did not dictate the terms of the relationship unilaterally and arbitrarily, but that it was the product of a constant process
of exchange and sometimes of negotiation between the conquerors
and the conquered. The Muslims were deeply aware of the local cultures.
They chose to espouse some of their concepts, customs, and institutions,
adapting them to their own needs, while rejecting others. This occurred
Conclusion
169
consciously and intentionally in some cases, and unconsciously and inadvertently in others. It must be stressed that even before the conquest the
Muslims were not unfamiliar with these cultures; this process naturally
became much more intensive with the progress of their conquest and consequent settlement in the conquered lands. The emergence of the regulation of the non-Muslims in this regard is just one case in point which
reveals that Islamic culture and society were forged within the milieu of
conquest and that the various populations which came under Muslim
domination had played a major role in their formation.
Appendix I
Al-Turtu4sh3s Version of Shuru4t Umar
The translation of Shuru4t Umar adduced here is based on Bernard

Lewiss translation of al-Turtu4sh3s version in Sira4j al-mulu4k (Lewis,
Islam, pp. 21719). The list of restrictions in the text is divided into
clauses by the author for reference only, and is not found in the original. Select additions are given in parentheses:IM=Ibn al-Murajja4, Fada4il
bayt al-maqdis; IQ=Ibn Qayyim al-Jawziyya, Ahka4m ahl al-dhimma. For
citations of the main versions of the Shuru4t, see the introduction, n. 3.
shuru4t umar
We heard from Abd al-Rahma4n b. Ghanam [d. 78/697] as follows:When
Umar b. al-Khatta4b, may God be pleased with him, accorded a peace to
the Christians of Syria, we wrote to him as follows:
In the name of God, the Merciful and Compassionate. This is a letter to
the servant of God Umar, Commander of the Faithful, from the Christians
of such-and-such a city. When you came against us we asked you for safeconduct (aman
4 ) for ourselves, our descendants, our property, and the people
of our community, and we undertook the following obligations toward you:
1. We shall not build, in our cities or in their neighbourhood, new
monasteries, churches, convents, or monks cells, nor shall we
repair, by day or by night, such of them as fall in ruins or are situated in the quarters of the Muslims.
2. We shall keep our gates open wide for passersby and travellers. We
shall give board and lodging to all Muslims who pass our way for
three days. We shall not give shelter in our churches or in our
dwellings to any spy, nor hide him from the Muslims.
171
172
Appendix I
3. We shall not teach the Qura4n to our children.

4. We shall not manifest our religion publicly nor convert anyone to
it. We shall not prevent any of our kin from entering Islam if they
wish it.
5. We shall show respect towards the Muslims [IM: and we shall
show them the way], and we shall rise from our seats when they
wish to sit.
6. We shall not seek to resemble the Muslims by imitating any of
their garments, the qalansuwa, the turban, footwear, or the parting
of the hair. We shall not speak as they do, nor shall we adopt their
kunyas.
7. We shall not mount on saddles, nor shall we gird swords nor bear
any kind of arms nor carry them on our persons.
8. We shall not engrave Arabic inscriptions on our seals.
9. We shall not sell fermented drinks [IM, IQ:nor shall we keep pigs
in their vicinity].
10. We shall clip the fronts of our heads.
11. We shall always dress in the same way wherever we may be, and
we shall bind the zunna4r round our waists.
12. (a) We shall not display our crosses or our books in the roads or
markets of the Muslims [IM, IQ:nor shall we conduct processions (lit. go out) on Palm Sunday and Easter].
(b) We shall only use clappers in our churches very softly [IM,
IQ:and we shall not display the cross on them].
(c) We shall not raise our voices in our church services or in the
presence of Muslims, nor shall we raise our voices when following our dead.
(d) We shall not show lights on any of the roads of the Muslims or
in their markets.
(e) We shall not bury our dead near the Muslims.
13. We shall not take slaves who have been allotted to the Muslims.
14. We shall not build houses overtopping the houses of the Muslims.
When I brought this letter to Umar, may God be pleased with him, he
added:We shall not strike any Muslim.
We accept these conditions for ourselves and for the people of our
community, and in return we receive safe-conduct. If in any way we
violate these undertakings for which we ourselves stand surety, we forfeit our covenant (dhimma), and we become liable to the penalties for
contumacy and sedition.
Appendix II
Al-Sha4fi3s Version of the Pact to Be Accorded
to Non-Muslim Subjects
The document adduced here is based on Bernard Lewiss translation of

al-Sha4fi3s version of the pact to be accorded to the Christians by the
ima4m given in Kita4b al-umm (see Lewis, Islam, pp. 21923).
al-sha4fi3s version of the pact to be accorded

to non-muslim subjects
If the Ima4m wishes to write a document for the poll-tax (jizya) of nonMuslims, he should write:
In the name of God, the Merciful and the Compassionate.
This is a document written by the servant of God so-and-so,
Commander of the Faithful, on the 2nd of the month of Rab3 I, in the
year such-and-such, to so-and-so son of so-and-so, the Christian, of the
descendants of such-and-such, of the people of the city of so-and-so.
I accord to you and to the Christians of the city so-and-so that which
is accorded to the dhimm s, in conformity with what you have given to
me and the conditions I have laid down concerning what is due to you
and to them, and I have agreed to your request and accorded to you and
to them, on behalf of myself and of all the Muslims, safe-conduct (ama4n),
for as long as you and they maintain all that we have required of you,
namely:
You will be subject to the authority of Islam and to no contrary authority. You will not refuse to carry out any obligation which we think fit to
impose upon you by virtue of this authority.
If any one of you speaks improperly of Muhammad, may God
bless and save him, the Book of God, or of His religion, he forfeits the
173
174
Appendix II
protection (dhimma) of God, of the Commander of the Faithful, and of

all the Muslims; he has contravened the conditions upon which he was
given his safe-conduct; his property and his life are at the disposal of the
Commander of the Faithful, like the property and lives of the people of
the house of war (da4r al-harb).
If one of them commits fornication with a Muslim woman or goes
through a form of marriage with her or robs a Muslim on the highway or
subverts a Muslim from his religion or gives aid to those who made war
against the Muslims by fighting with them or by showing them the weak
points of the Muslims, or by harbouring their spies, he has contravened
his pact (ahd), and his life and his property are at the disposal of the
Muslims.
If he commits some lesser offence against the property or the honour
of a Muslim or against an infidel under Muslim protection, with a pact
or safe-conduct, he shall be punished.
We shall supervise all your dealings with Muslims. If there is anything
in which you are engaged which is not lawful for a Muslim, we shall
reject it and punish you for it. If you sell a Muslim something we hold
forbidden, such as wine, pig, blood, or carrion, and the like, we shall
annul the sale, confiscate the price if it has been paid, and not return the
thing to you if it still exists, but pour it out if it is wine or blood and burn
it if it is carrion; if the purchaser has already consumed it, we shall not
oblige him to pay for it, but we shall punish you for it.
You shall not give a Muslim anything to eat or drink which is forbidden, nor marry him in the presence of witnesses chosen from among you,
nor by wedding rites we hold to be invalid.
We shall not supervise transactions between you and your coreligionists or other unbelievers nor inquire into them as long as you are
content. If the buyer or the seller among you desires the annulment of
a sale and comes to us to ask for this, we shall annul it or uphold it in
accordance with the provisions of our law. But if the payment has been
made and the purchase consumed, we shall not order restitution, for this
would count as a completed sale between polytheists.
If one of you or any other unbeliever applies to us for judgment, we
shall adjudicate according to the law of Islam. But if he does not come to
us, we shall not intervene among you.
If you commit manslaughter against a Muslim or a protected person
(mua4had), whether protected by you or by others, your clan is liable for
the blood price as with the Muslims. Your clan consists of your paternal
kinsmen. If the offender is one of you who has no kin, he himself is liable
Al-Sha4fi s Version of the Pact to Non-Muslim Subjects
175
for the blood price with his own property. If he kills with intent, he is subject to retaliation unless the heirs are content to receive the blood price,
in which case they must get it at once.
If any of you steals and the victim takes him before a judge, his hand
shall be cut off if his crime is punishable by this penalty, and he shall
make restitution.
If anyone commits slander and a legal penalty (hadd) is due, it shall be
inflicted on him; if there is no legal penalty, he shall be punished at discretion so that the laws of Islam may be applied among you in these matters,
both specified and unspecified.
You may not display crosses in Muslim cities, nor proclaim polytheism, nor build churches or meeting-places for your prayers, nor strike
clappers, nor proclaim your polytheistic beliefs on the subject of Jesus,
son of Miriam, or any other to a Muslim.
You shall wear the girdle (zunna4r) over all your garments, your cloaks
and the rest, so that the girdles are not hidden. You shall differentiate
yourselves by your saddles and your mounts, and you shall distinguish
your and their headgear (qalansuwa) by a mark which you shall place
on your headgear. You shall not occupy the middle of the road or the
seats in the market, obstructing Muslims.
Every free adult male of sound mind among you shall have to pay
a poll-tax (jizya) of one dna4r, in good coin, at the beginning of each
year. He shall not be able to leave his city until he pays his poll-tax or
appoints someone to pay it on his behalf, with no further liability until
the beginning of the year. The poor among you is liable for the poll-tax,
which should be paid for him. Poverty does not free you from any obligation, nor does it abrogate your pact (dhimma) You are subject to no
taxes on your money other than the poll-tax as long as you stay in your
country or travel around in the lands of the Muslims otherwise than as
a merchant. You may in no circumstances enter Mecca. If you travel for
trade, you shall pay the Muslims a tenth part of all your merchandise.
You may go wherever you wish in the lands of the Muslims, except
Mecca, and reside wherever you wish in the lands of the Muslims, except
the Hija4z, where you may only stay for three days in any city, after which
you must leave.
Whoever has hair under his garments, has attained puberty, or has
completed his fifteenth year before this, is subject to these conditions if he
accepts them. If he does not accept them, he has no covenant.
Your children under age, boys below puberty, persons of unsound
mind, and slaves are not liable for the poll-tax. But if the madman
176
Appendix II
recovers his reason, the child attains puberty, or the slave is emancipated
and follows your religion, they are all liable for the poll-tax.
These conditions are binding on you and on those who have accepted
them. Those who reject them we cast out.
We owe you protection for yourselves and for your property which it
is lawful for you to hold according to our laws, against anybody, Muslim
or other, who seeks to wrong you, as we would protect our own persons
and property, and we administer justice to you in matters under our own
jurisdiction as we would do with our own property. But no one among
you can ask us to protect any forbidden thing which you own, such as
blood, carrion, wine, or pigs, as we would protect lawful property. We
shall not prevent you from having them, but we shall not allow you to
display them in the cities of the Muslims. If a Muslim or any other buys
such merchandise we shall not compel him to pay the price, because these
are forbidden things and therefore have no price which could be legally
enforced. But we shall restrain him from troubling you in this, and if he
persists he shall be punished, though not by enforcing payment for what
he took from you.
You must observe all the conditions which we have imposed.
You may not deceive a Muslim nor give aid to their enemies by word
or deed.
This is the pact and covenant of God, and the greatest obligation to
respect this covenant which God has ever imposed on any of His creatures. You have the pact and covenant of God, the protection (dhimma)
of so-and-so, Commander of the Faithful, and the protection of the
Muslims to carry out their obligations toward you.
Those of your children who reach the age of puberty are in the same
position as you are, in regard to what we have given to you and in the
obligation to observe all the conditions which we have laid down for you.
If you change or modify anything, then the protection of God, of
so-and-so the Commander of the Faithful, and of the Muslims shall be
withdrawn from you. If anyone of those to whom we gave this was not
present when we wrote it, and hears of it and accepts it, the conditions
stated in it are binding on him and on us. If he does not accept it, we cast
him out.
Witnesses.
Notes
Introduction
1. See, for example, the Doctrina Iacobi Nuper Baptisati in Gilbert Dagron and
Vincent Droche, Juifs et chrtiens dans lOrient du VIIe sicle, Travaux et
Mmoires, 11 (1991), pp.47229, at p.209; for an English translation see
R. G. Hoyland, Seeing Islam as Others Saw it, Princeton 1997, pp.578; the
sermons and letters of Sophronius, Patriarch of Jerusalem during the conquest, in Hoyland, Seeing Islam, pp.6773, and many other sources cited and
surveyed by Hoyland.
2. M. Levy-Rubin, Changes in the Settlement Pattern of Palestine Following
the Arab Conquest, in K. Holum and H. Lapin, eds., Shaping the East,
forthcoming.
3. Throughout this book the term Shurut4 will refer only to the canonical text
defining the status of non-Muslims under Muslim rule. This text is cited among
many others by Abu4 Bakr al-Khallal,
4 Ahl al-milal wa-al-ridda wa-al-zanadika
4
al-sala
min kitab
wa-tar
q
t4 wa-al-farad
4 al-jami
4 lil-Khallal4 Ab Bakr Ahmad
al-Baghdad al-Hanbal
b. Ha ru
4 n
4 b. Yazd
, 2 vols., ed. Ibrah

4 m
3 b. Hamd
b.
Sulta n,
4 Riyadh 1996; Abu4 al-Maal
4 3 al-Musharraf b. al-Murajja4 b. Ibrah
4 m
3
al-Maqdis,3 Fada 4 il bayt al-maqdis wa-al-khall wa-fada 4 il al-sham
4 , ed. O.
Livne-Kafr,3 Shfaram 1995, pp.556; Abu4 al-Qasim
4
Al3 b. al-Hasan
b. Asakir,
4
Tarkh
madinat dimashq, Beirut 1995, vol. II, 120, 1749 (includes five different versions of the document); Al3 b. Ahmad
b. Hazm,
al-Muhalla
,4 ed. Ahmad
Muhammad Shakir, Beirut 1980, vol. VIII, pp.3467; Ibn Qayyim al-Jawziyya,
Ahka
m
4 ahl al-dhimma, Damascus 1961, vol. II, pp.65762 (two versions).
The name does not refer to various other documents which try to do the
same but have not been canonized such as, for example, the document adduced
by Abu4 Abdalla4h Muhammad b. Idr3s al-Sha4fi3s Kita4b al-umm, Cairo 1968,
vol. IV, pp.11819, trans. in B. Lewis, Islam from the Prophet Muhammad
to the Capture of Constantinople, New York 1974, vol. II, pp.21923; or the
one cited by Yaqu4b b. Ibra4h3m Abu4 Yu4suf in Kita4b al-khara4j, Cairo 1353 H,
pp.1389. This issue is discussed in Chapter 2.
177
178
Notes to pp. 25
4. The most important works on the subject are A. S. Tritton, The Caliphs and
their Non-Muslim Subjects, London 1930; H. Zaya4t, Sima4t al-nasa4ra4 wa-alyahu4d f al-isla4m:al-salb wa-al-zunna4r, wa-al-ama4ma wa-al-ghiya4r; shuru4t
al-umariyya, al-Machriq, 43/2 (1949), pp. 161252; E. Strauss (Ashtor),
The Social Isolation of Ahl al-Dhimma, in O. Komls, ed., tudes orientales
la mmoire de Paul Hirschler, Budapest 1950, pp. 7394; A. Fattal, Le
statut lgal des non-musulmans en pays dIslam, Beirut 1955, esp.pp.609;
A. Noth, Abgrenzungsprobleme zwischen Muslimen und nicht-Muslimen:
Die Bedingungen Umars (a-uru4t al-umariyya) unter einem anderen
Aspekt gelesen, Jerusalem Studies in Arabic and Islam, 9 (1987), pp.290
315, trans. M. Muelhaeusler in R. G. Hoyland, ed., Muslims and Others
in Early Islamic Society, Ashgate 2004, pp. 10324; M. R. Cohen, Under
Crescent and Cross:The Jews in the Middle Ages, Princeton 1994, pp.5474;
M. R. Cohen, What was the Pact of Umar? A Literary-Historical Study,
Jerusalem Studies in Arabic and Islam, 23 (1999), pp.10057; M.R. Cohen,
Islamic Attitudes and Policies, in The Cambridge History of Judaism:Jews
in the Medieval Islamic World, vol. V:Robert Chazan and Marina Rustow,
eds., The Medieval Period, Seventh through Fifteenth Centuries, forthcoming;
Y.Friedmann, Tolerance and Coercion in Islam, Cambridge 2003; M. LevyRubin, Shuru4t Umar and its Alternatives: The Legal Debate throughout
the Eighth and Ninth Centuries over the Status of the Dhimms, Jerusalem
Studies in Arabic and Islam, 30 (2005), pp. 170206; Daniel E. Miller, From
Catalogue to Codes to Canon: The Rise of the Petition to Umar among
Legal Traditions Governing non-Muslims in Medieval Islamicate Societies,
Ph.D. thesis, University of Missouri-Kansas City, 2000 (I owe thanks to Prof.
Mark Cohen for referring me to this work).
5. These will be discussed in detail in the book.
6. On Umars image in Muslim historiography and tradition see A. Hakim,
Umar b. al-Khatta4b:Some Aspects of his Image as an Ideal Leader in Early
Islamic Traditions, Ph.D. thesis, Tel-Aviv University, 2002 (Hebrew).
7. Tritton, The Caliphs, p.12; Fattal, Le statut, p.68.
8. See Noth, Abgrenzungsprobleme; see chap.2.
9. Cohen, Under Crescent and Cross, p.57. See his comprehensive discussion
of the petition form of the Shuru4t in Cohen, What was the Pact of Umar?.
10. This is the text called Juz fh shuru4t al-nasa4ra4, attributed in the text to
Qa4d3 Abu4 Muhammad Abd Alla4h b. Ahmad b. Zabr al-Qa4d3, 255329/870
940; Cohen, What was the Pact of Umar?, pp.11016, 13157.
11. See Noth, Abgrenzungsprobleme.
12. Miller, From Catalogue to Codes to Canon, pp.35.
13. For bibliography, see chap.1.
14. P. Crone and M. Cook, Hagarism: The Making of the Islamic World,
Cambridge 1977.
15. See Michael G. Morony, Iraq After the Muslim Conquest, Princeton 1984.
16. See e.g. L. I. Conrad, Theophanes and the Arabic Historical Tradition:Some
Indications of Intercultural Transmission, Byzantinische Forschungen, 15
(1990), pp.144; L. I. Conrad, The Conquest of Arwa4d:A Source-Critical
Study in the Historiography of the Early Medieval Near-East, in A. Cameron
Notes to pp. 59
179
and L. I. Conrad, eds., The Byzantine and Early Islamic Near-East, Studies in
Late Antiquity and Early Islam 1, Princeton 1992, vol. I, pp.317401.
17. Hoyland, Seeing Islam.
18. Chase F. Robinson, Empire and Elites after the Muslim Conquest: The
Transformation of Northern Mesopotamia, Cambridge 2000.
19. For examples of such uses of historical researches towards political ends see
M. R. Cohen, The Neo-Lachrymose Conception of Jewish Arab History,
Tikkun, 1/2 (MayJune 1991), pp.1116.
1. The Roots and Authenticity of the Surrender Agreements
in the Seventh Century
1. See A. Noth, Die literarische berlieferten Vertrge der Eroberungszeit als historische Quellen fr die Behandlung der unterworfenen Nicht-Muslims durch
ihre neuen muslimischen Oberherren, in T. Nagel et al., eds., Studien zum
Minderheitenproblem im Islam, vol. I, Bonn 1973, pp.282304, including a
comprehensive bibliography; see also D. Hill, The Termination of Hostilities
in the Early Arab Conquests, 634656 AD, London 1971.
2. Fattal, Le statut, p.58:Celle-ci nest pas un trait en bonne et due forme; elle
ne porte ni date ni limite de temps; elle nest pas toujours crit. Rares sont
ceux qui crivent larabe cette poque, et plus sont ceux qui le lisent parmi
les Byzantines et les Iraniens. Les termes de la convention sont gnralement brefs et vagues; les texts les plus prcis et les plus dtaills sont les plus
suspects.
3. On this treaty see also below; on the term baqt see pp. 345, 39.
4. See M. Hinds and H. Sakkout, A Letter from the Governor of Egypt to the King
of Nubia and Muqurra concerning EgyptianNubian Relations in 141/758,
in W. al-Qad
4 ,3 ed., Studia Arabica and Islamica:Festschrift for Ihsa
n
4 Abbas4 ,
Beirut 1981, pp.20929, esp.p.211, citing P. M. Holt, The Nilotic Sudan,
in The Cambridge History of Islam, vol. II:P. M. Holt, Bernard W. Lewis, and
B. Lewis, eds., The Further Islamic Lands, Islamic Society and Civilization,
Cambridge 1970, pp.32744, at p.328; M. Brett, The Arab Conquest and the
Rise of Islam in North Africa, in The Cambridge History of Africa, vol. II:J. D.
Fage, ed., c. 500 BCAD 1050, Cambridge 1978, pp.490555, at p.506.
5. The early Abbasid period witnessed the creation of the since authoritative
structure of Islamic law A by-product of the search for certitude concerning prophetic precedent was thus a magnificent efflorescence of historywriting, in which the careful scholarly reconstruction of the story of the
creation of the Nubian baqt formed one very small episode: J. Spaulding,
Medieval Christian Nubia and the Islamic World:A Reconsideration of the
Baqt Treaty, International Journal of African Historical Studies, 28 (1995),
pp.57794, at p.577.
6. Hinds and Sakkout, A Letter, p. 220, l. 19: al-baqt alladh sulihtum
alayhi (the baqt about which a peace agreement was made with you); and
l. 534:fa-innahu qad dhakara l inna alaykum baqt sann lam tuaddu4hu
wa-ma baathtum bihi min al-baqt (For he [Salm] mentioned to me that you
are liable to the baqt of outstanding years, which you have not made over).
180
Notes to pp. 911
7. Ibid., p.224, l. 63.

8. See Noth, Vertrge; Albrecht Noth, Quellenkritische Studien zu Themen,
Formen und Tendenzen frhislamischer Geschichtsberlieferung, part 1:
Themen und Formen, Bonner Orientalische Studien 25, Bonn 1973, pp.6071;
Wadad
4 al-Qad
4 ,3 Madkhal ila4 dirasat
4 uhud
4 al-sulh
al-islamiyya
4
zaman al-futuh
4 ,
in M. A. Bakhit and I. Abbas, eds., Proceedings of the Second Symposium (4th

conference) on the History of Bilad
4 al-Sham
4 during the Early Muslim Period
up to 40 AH/640 AD, vol. II (Arabic), Amman 1987, pp.193269.
9. Noth, Vertrge, pp. 3056. Noth also refers here to the agreement of
al-Sha4m cited by Abu4 Yu4suf in Kita4b al-khara4j which is probably a later
revision of an agreement: on this see Levy-Rubin, Shuru4t, pp. 1989. The
other agreement cited, that of al-Raqqa, as presented in Ahmad b. Yahya4
al-Bala4dhur3, Kita4b futu4h al-bulda4n, ed. M. J. de Goeje, Leiden 1866, p.173,
is an exception, and is certainly forged, as Noth claims, since it includes the
prohibition on building new prayer houses, exhibiting the na4qu4s and the
cross, and holding Easter processions.
10. Noth, Vertrge, p.286.
11. Qa4d3, Madkhal.
12. Regarding early legal documents in Islam see M. Lecker, A Pre-Islamic
Endowment Deed in Arabic Regarding al-Wahda in the Hija4z, paper presented at the Huitimes Rencontres Sabeenes, Aix-en-Provence, 2003;
repr. in M. Lecker, People, Tribes and Society in Arabia Around the Time of
Muhammad, Aldershot:Variorum 2005, IV; M. Lecker, The Constitution of
Medina, Princeton 2004.
13. For a short survey and bibliography of the history of international treaties in the ancient Near East as well as in Greece, the Hellenistic world,
and Rome see Peter Kehne et al., International Treaties, in H. Cancik,
H.Schneider, eds., Brills New Pauly, Antiquity volumes, Leiden 2010, and
bibliography.
14. See Peter Kehne, Treaties, Upholding of, in ibid. For a good and compact
survey and bibliography regarding the Ancient Near East see D. L. Magnetti,
The Function of the Oath in the Ancient Near Eastern International Treaty,
American Journal of International Law, 72 (1978), pp.81529; C.Phillipson,
The International Law and Custom of Ancient Greece and Rome, London
1911, vol. I, pp.375419.
15. See e.g. the treaty with Carthage, Polybius, Histories, 3.22; 3.24; 3.25,
trans. in A. C. Johnson, P. Robinson Coleman-Norton, and F. C. Bourne,
Ancient Roman Statutes, Austin 1961; Poly. 3.27, trans. in Johnson et al.,
Statutes, no. 13; with the Aetolians, Livy 38.11; Poly. 21.32, trans. Johnson
et al., Statutes, no. 24; or with Antiochus III, Livy 38.38; Poly. 21.42, trans.
Johnson et al., Statutes, no. 27.
16. See M. Weinfeld, Covenant Terminology in the Ancient Near East and its
Influence on the West, Journal of the American Oriental Society, 93 (1973),
pp.1909.
17. Ibn Hisha4m, al-Sra al-nabawiyya, vol. IV, Beirut 1411 H, p.143, cited later
by others.
Notes to pp. 1112
181
18. See S. Shaked, From Iran to Islam:Notes on Some Themes in Transmission,

Jerusalem Studies in Arabic and Islam, 4 (1984), pp.3167, at pp.345, n. 17.
19. Weinfeld, Covenant Terminology, n. 15.
20. On these see B. Dignas and E. Winter, Rome and Persia in Late Antiquity,
Cambridge 2007, pp. 11951; R. Blockley, East Roman Foreign Policy,
Leeds 1992, pp. 566. The existence of such treaties prior to the Muslim
conquest has already been noted by Chase Robinson. See Robinson, Empire
and Elites, pp.67.
21. See Blockley, Policy, pp. 1518; see D. A. Miller, Byzantine Treaties and
Treaty-Making:5001025 AD, Byzantino-Slavica, 32 (1971), pp.5676;
E. Chrysos, Byzantine Diplomacy, AD 300800: Means and Ends, in J.
Shepard and S. Franklin, eds., Byzantine Diplomacy, Aldershot 1992,
pp.2539, at p.30.
22. E.g. the exchange of letters between Sha4pu4r and Constantius in 3578
which ended in failure in Ammianus Marcellinus, Ammianus Marcellinus,
XVII.5.1XVII.5.15, Latin text with Eng. trans. by J. C. Rolfe, Loeb Classical
Library, London 19502, vol. I, pp. 3329; Chosroes exchange of letters
with Mauritius in 590 in Theophylactus, Theophylacti Simocattae Historiae,
ed. C. De Boor, re-ed. P. Wirth, Stuttgart 1972, IV.10.7IV.14.5, trans. M.
and M. Whitby as The History of Theophylact Simocatta, Oxford 1986,
pp.116123. For letters that ratify agreements, see below the examples of
the treaties of 363 and 562.
23. See P. Huyse, Die dreisprachige Inschrift a4buhrs I. an der Ka4ba-i-Zaradut,
Corpus Inscriptionum Iranicarum, part III: Pahlavi Inscriptions, vol. I,
London 1999; Dignas and Winter, Rome and Persia, pp.11922; see also Z.
Rubin, Nobility, Monarchy and Legitimation under the Later Sasanians, in
John Haldon and L. I. Conrad, eds., The Byzantine and Early Islamic NearEast VI:Elites Old and New in the Byzantine and Early Islamic Near-East,
Studies in Late Antiquity and Early Islam 1, Princeton 2004, pp.23573, at
pp.2434.
24. Rubin, Nobility, p.122, n. 14, citing L. Trmplemann, Triumph ber Julian
Apostata, Jahrbuch fr Numismatik und Geldgeschichte, 25 (1975), p.115.
25. See Blockley, Policy, pp.57, 1578; Dignas and Winter, Rome and Persia,
pp.12230.
26. See Blockley, Policy, p.136.
27. For an English translation of his description see Dignas and Winter, Rome
and Persia, pp.1224.
28. Ammianus, Ammianus Marcellinus, XXV.7.5XXV.7.13, trans. Rolfe,
vol.II, pp.5306.
29. Eutropius also calls this treaty pax ignobilis:Eutropi Breviarium ab urbe
condita, ed. F. Ruehl, Leipzig 1919, X, 17, trans. H. W. Bird as The Breviarium
ab urbe condita of Eutropius, the Right Honourable Secretary of State
for General Petitions: Dedicated to Lord Valens, Gothicus Maximus and
Perpetual Emperor, Liverpool 1993, p. 69, as does also Agathias, IV.25.7,
ed. and trans. R. Keydell, Agathiae myrinaei historiarum libri quinque, Berlin
1967, text vol. I, pp.1556, trans. vol. II, p.128.
182
Notes to pp. 1315
30. Ammianus, Ammianus Marcellinus, XXV.7.1314, trans. Rolfe, vol. II,

pp.5346.
31. Zosimus, 3.31, text B. Kindt, F. van Haeperen and CENTAL, eds., Thesaurus
Zosimi, Historia Nova, Louvain-la-Neuve 2008, trans. R. T. Ridley as
Zosimus:New History, Sydney 1982, p.66.
32. Ammianus, Ammianus Marcellinus, XXV.9.2, trans. Rolfe, vol. II, p.549.
This same claim is repeated by Eutropius, Eutropi Breviarium, X.17, p.77;
Eutropius, The Breviarium, p.69.
35. See also P. Heather, Foedera and Foederati in the Fourth Century, in W.
Pohl, ed., Kingdoms of the Empire:The Integration of Barbarians in Late
Antiquity, Leiden 1997, pp. 5774, at p. 63, who notes that Ammianus
usage of foedus and foederati are entirely in accord with ancient Roman
precedent.
36. On these see Blockley, Policy, pp.579, 61; Socrates, Historia Ecclesiastica,
IX.4, Socrate de Constantinople Histoire ecclsiastique, ed. G. C. Hansen,
trans. P Prichon and P. Maraval, introduction and notes P. Maraval, Paris
20047, talks about a hundred-year peace treaty. It is not clear whether this
refers to 408 or to 422 (Blockley, Policy, p.57, n. 37; John Malalas, Ioanni
Malalae Chronographia, ed. I. Thurn, Berlin 2000, XIV.23, talks about a
fifty-year treaty). See Dignas and Winter, Rome and Persia, pp.1378.
37. G. Greatrex, Rome and Persia at War:502532, Leeds 1998, pp.11418;
Dignas and Winter, Rome and Persia, pp.1004.
38. Marcellinus, The Chronicle of Marcellinus, text, trans., and comm. B. Croke,
Byzantina Australiensia 7 (Australian Association for Byzantine Studies),
Sydney 1995, p.34.
39. Procopius, History of the Wars, trans. H. B. Dewing, London 195471 vol. I,
p.69.
40. Pseudo-Joshua. The Chronicle of Pseudo-Joshua the Stylite, trans. and annot.
F. R. Trombley and J. W. Watt, Liverpool 2000, p.99.
41. These earlier roots of this legal tradition will be referred to on pp. 2930.
42. Zachariah of Mitylene, Historia Ecclesiastica Zachariae Rhetori Vulgo Adscripta, ed. E. W. Brooks, Corpus Scriptorum Christianorum Orientalium, Scriptores
Syri 39, Louvain 1921, pp.334; F. J. Hamilton and E. W. Brooks, The Syriac
Chronicle known as that of Zachariah of Mitylene, London 1899, p.163.
43. The significance of the formal aspects of such treaties is apparent also
in Marcellinus description of the treaty of 533: After the fierce, drawnout struggle conducted against the Persians with Roman toil, peace (pax)
was finally reached with the Persians through Rufinus the Patrician and
Hermogenes the Master of Offices, both ambassadors being sent by our
emperor. By a term of this agreed treaty pledges of good faith (munera concordiae) were sent afterwords by each emperor to the other. See Marcellinus,
Chronicle, p.44.
44. See R. C. Blockley, ed. and trans., The History of Menander the Guardian,
Liverpool 1985, pp.5571; the treaty pp.717.
45. Ibid., p.71.
Notes to pp. 1517
183
46. Ibid., p.77.

47. See Dignas and Winter, Rome and Persia, pp. 14851, citing Chronicon
Paschale 284628 AD, trans. L. M. Whitby and M. Whitby, Liverpool 1989,
p. 189 n. 491; Nikephoros, Nicephori Archiepiscopi Constantinopolitani
Opuscula historica, ed. C. de Boor, Leipzig 1880, Historia Syntomos, 22b
23b, pp.1920; C. Mango, Nikephoros, Patriarch of Constantinople:Short
History, Text, Translation and Commentary, Dumbarton Oaks Papers, 10
(1990), Washington DC, p.15; Theophanes, Theophanis Chronographia, ed.
C. de Boor, Leipzig 18835, AM 6118, vol. I, p.327.
48. Blockley, Policy, pp.1056, notes that their control of southern Mesopotamia
had led the Sasanians to adopt the organizational, administrative and legalistic skills prevalent there, including the diplomatic practices.
49. Heather, Foedera, pp.69, 71. For a recent comprehensive survey and analysis of the Barbarians role see P. Heather, Empires and Barbarians:The Fall
of Rome and the Birth of Europe, Oxford 2010.
50. Ammianus, Ammianus Marcellinus, XVII.10.3, trans. Rolfe, vol. I, pp.360
1, XVII.10.9, trans. Rolfe, vol. I, pp.3723, XVII.10.710, trans. Rolfe, vol.
I, pp.3735.
51. Ammianus, Ammianus Marcellinus, XVII.12.13, trans. Rolfe, vol. I,
pp.3767.
52. Ammianus, Ammianus Marcellinus, XVII.13.21, trans. Rolfe, vol. I,
pp.3801.
53. Ammianus, Ammianus Marcellinus, XVII.12.13, trans. Rolfe, vol. I, pp.
3767:numquam antea pignora foederis exhibere conpulsi.
54. Priscus, in R. C. Blockley, The Classicising Historians of the Later Roman
Empire, vol. II, Liverpool 1983, pp.2247.
55. This is in contrast to older views which claimed that foedus could not follow
deditio. The opposite, in fact, is true. Most submissions were conditional
rather than unconditional. See Heather, Foedera, pp.612.
56. Only several significant examples will be adduced here, many more cases are
mentioned in the sources. See Blockley, Policy, pp.596, 12964; see p.8
(regarding Constantines agreement with the Tervingi which lasted into the
fourth and fifth centuries); pp.605, 159 regarding the agreements with the
Huns in 434/5, 442, 447; see also H. Wolfram, History of the Goths, Berlin
and Los Angeles 1988, pp.6, 192, 403, n. 146.
57. Ammianus, Ammianus Marcellinus, XX.5.810, trans. Rolfe, vol. II,
pp.345.
58. Ammianus, Ammianus Marcellinus, XXX.3.56, trans. Rolfe, vol. III,
pp.31617.
59. Ammianus, Ammianus Marcellinus, XXXI.12.9, trans. Rolfe, vol. III,
pp.4689; see also P. Heather, The Goths, Oxford 1996, p.137.
60. The main source here is Themistius, Themistii Orationes quae Supersunt, ed.
H. Schenkl, Leipzig 1965974, XVI, 211, trans. in P. Heather and D. Moncur,
Politics, Philisophy, and Empire in the Fourth Century: Select Orations of
Themistius, Translated Texts for Historians 36, Liverpool 2001, pp.2801.
Themistius oration is highly apologetic, clearly in the face of criticism
regarding the autonomy that was granted to them. This criticism is clearly
184
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
Notes to pp. 1718

apparent in Synesius de Regno, 1415, Patrologia Graeca, vol. LXVI, cols.
108898, who clearly blames Theodosius for granting them the status of
symmachoi. see A. H. M. Jones, The Later Roman Empire 284602, Oxford
1964, vol. I, p.157, and vol. III, p.29, n. 46, where the sources regarding
this treaty are cited.
See Blockley, Policy, pp.401.
Blockley, Classicising Historians, p.226.
Priscus, cited in ibid., pp.2301.
E. Landau-Tasseron, Alliances in Islam, in M. Bernards and J. Nawas,
eds., Patronate and Patronage in Early and Classical Islam, Leiden 2005,
pp.149; E. Landau-Tasseron, Alliances among the Arabs, al-Qantara, 26
(2005), pp.14174; A. Marsham, Rituals of Islamic Monarchy:Accession
and Succession in the First Muslim Empire, Edinburgh 2009, pp.256.
See Marsham, Rituals of Islamic Monarchy, pp.256.
Ibid., pp. 278; see also Ahd al-Umma, cited and translated in Lecker,
Constitution.
See especially the treaty made c. 550 between Abd al-Muttalib b. Ha4shim,
grandfather of Muhammad, and leading figures in Mecca, cited in Marsham,
Rituals of Islamic Monarchy, pp. 278, and additional information there,
and in Landau-Tasseron, Alliances among the Arabs.
G. W. Bowersock, The GreekNabatean Bilingual Inscription at Ruwwa4fa,
Saudi-Arabia, in J. Bingen et al., eds., Le monde grec, pense, littrature,
documents:Hommages Claire Praux, Brussels 1975, pp.51417, cited in
Z. Rubin, Mavia, Queen of the Saracens, Cathedra, 47 (1988), pp.2549
(Hebrew), at pp.423.
Rubin, Mavia, pp.4546. Rubin notes, following F. Altheim and R. Stiehl,
Die Araber in der alten Welt, vol. II, Berlin 1965, pp.31920, that it is possible that Imru4 al-Qays had been an ally of the Persian Bahra4m III, who was
overthrown in 293 by his brother Narseh. Following this event he may have
chosen to desert, and fled to the Romans, became their ally, and converted to
Christianity, as claimed by Abu4 Jafar Muhammad b. Jar3r al-Tabar3, Tarkh

al-rusul wa-l-mulu4k, ed. M. J. de Goeje et al., Leiden 18791901, vol. I.2,
p.834; regarding this issue see also n. 133 in C. E. Bosworth, annot. and
trans., The History of al-Tabar, vol. V: The Sa4sa4nids, the Byzantines, the
Lakhmids, and Yemen, Albany 1999, p.44.
Rufinus, Historia Ecclesiastica, XI, 6; Theodoretus, Historia Ecclesiastica, IV,
23; Socrates, Historia Ecclesiastica, IV, 36; Sozomen, Historia Ecclesiastica,
VI.38. In addition to Rubin, Mavia, see also Altheim and Stiehl, Die
Araber, pp. 32832; J. S. Trimingham, Mawwya the first Christian Arab
Queeen, Theological Review of the Near-East School of Theology, 1
(1978), pp.310; G. W. Bowersock, Mavia, Queen of the Saracens, in W.
Eck et al., eds., Studien zur antiken Sozialgeschichte: Festschrift Friedrich
Vittinghoff, Cologne and Vienna 1980, pp. 47795; P. Mayerson, Mavia,
Queen of the Saracens, Israel Exploration Journal, 30 (1980), pp.12331; I.
Shahid, Byzantium and the Arabs in the Fourth Century, Washington 1984,
pp.13843.
Sozomen, Historia Ecclesiastica, VI.38.1.
Notes to pp. 1920
185
72. Blockley, East Roman Foreign Policy, pp.1612. Z. Rubin translates this
passage in Sozomen, Historia Ecclesiastica, VI.38.1 as broken, supposing
that the treaty was still in effect and broken actively by Mavia, but the
Romans are not presented as demanding that she uphold the treaty, and in
fact sent ambassadors to offer her a new treaty:When the war became difficult, it seemed necessary to send an embassy regarding peace (per eirenes)
to Mavia. It is said that she openly refused the envoys sent to her regarding
the treaty (spondas), if a certain Bishop by the name of Moses is not
ordained for her people. See the French translation in A. J. Festugire and
B.Grillet, Sozomne histoire ecclsiastique, VVI, Paris 2005, p.457:Vers
ce temps, le roi des Sarrasins tant mort, la trve avec les Romains fut
rompue.
73. This is also Shahids opinion:Fourth Century, p.143.
74. Ammianus, Ammianus Marcellinus, XXXI.16.46, trans. Rolfe, vol. III,
pp.5023, cited by Rubin, Mavia, p.41.
75. R. A. B. Mynors, Panegyrici Latini, Oxford 1964, II.22.3; Pacatus, Panegyric
to the Emperor Theodosius, trans. C. E. V. Nixon, Liverpool 1987, p. 34,
cited by Rubin, Mavia; see also Shahid, Fourth Century, pp.2034.
76. Z. Rubin, Mavia, p.41.
77. Sozomen, Historia Ecclesiastica, VI.38.1415.
78. Blockley, Classicising Historians, pp.4047. A similar example of a head of
a tribe who converted to Christianity and and received the title of phylarch
is Aspebetos; see D. J. Chitty, The Desert a City: An Introduction to the
Study of Egyptian and Palestinian Monasticism under the Christian Empire,
Oxford and New York 1966, pp.834, 88.
79. C. Mller, ed., Fragmenta Historicorum Graecorum, Paris 1928, vol. 4,
p.179; see I. Shahid, Byzantium and the Arabs in the Sixth Century, vol. I,
part 1, Washington 1995, pp.15360.
80. Shahid (p. 159) believes that this hegemonia, which basically should be
translated as praefecture, must have been only an honorary one, although
the text does not support this.
81. Compare for example Constantines treaty with the Goths in 332. See
Heather, Foedera, pp.667.
82. Procopius sayings in Wars, 3.11.34, 8.5.1314 regarding the changes in
the term foederati (see Heather, Foedera, pp.5860) seem to refer not to
the institution of the treaty or foedus, but rather to the imprecise use of the
term, which in his days was applied not only to groups with whom a foedus
was signed, but also often to others who served, it seems, as mercenaries. He
in fact seems to protest against this, saying:But at the present time there is
nothing to prevent anyone from assuming this name, since time will by no
means consent to keep names attached to the things to which they were formerly applied, but conditions are ever changing and men pay little heed
to the meaning which they originally attached to a name (Wars, 3.11.34).
In fact, in the preceding paragraph Procopius recounts how Justinian sends
the tyrant of Sardinia a letter with an envoy, promising him an alliance (symmachia) (Wars, 3.10.32). It should also be noted that this refers specifically
to the Latin term foederati, which Procopius, although writing in Greek,
186
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
Notes to pp. 2023

adduces in its Latin form (kai phoideratoi epiklethentes. houto gar autous
tote Latino4n ekalesan Romaioi:Wars, 3.8.13).
See Blockley, ed. and trans., Menander the Guardian:for the description of
the negotiations see pp.5571, the treaty, pp.717. Johnson et al., Statutes,
nos. 52, 110.
Priscus in Blockley, Classicising Historians, pp.2247.
See Malalas, Chronographia, p.270, trans. E. Jeffreys, M. Jeffreys, and R.
Scott as The Chronicle of John Malalas, Melbourne 1986, p.143.
Malalas, Chronographia, chap.11.3, p.205, ll. 359; trans. Jeffreys et al.,
pp.1434.
Pseudo-Dionysius, Pseudo-Dionysius of Tel-Mahre Chronicle, Part III, trans.
with notes and introduction by W. Witakowski, Liverpool 1996, p.64:At
the same time the Persians attacked Antioch and it was conquered, ravaged
and burnt they burned it with fire and destroyed it. They stripped it and
removed even the marble slabs, with which the walls were overlaid, and took
them away to their country, since they also were building in their country a
city like this one and named it Antioch too.
See n. 30.
See Zosimus, Thesaurus, book 3, paras. 334; trans. Ridley, pp. 678;
Malalas, Chronographia, pp. 3367; trans. Jeffreys et al., p. 183. See
C. Mango, Nisibis, in A. P. Kazhdan et al., eds., Oxford Dictionary of
Byzantium, New York 1991, vol.III, p.1488.
Such situations were not limited to the eastern part of the empire. A situation in which the cities of Thessaly and Macedonia were thrown back and
forth between the Barbarians and the empire and paid a dear price is described
in Zosimus, Thesaurus book 4, paras. 312, trans. Ridley, pp.845.
See Pseudo-Joshua, The Chronicle of Pseudo-Joshua the Stylite, introduction
by Trombley and Watt, pp.xxxviili, esp.lxvxlvii.
Pseudo-Joshua, Incerti auctoris Chronicon ps.-Dionysianum vulgo dictum
I, ed. and trans. J. B. Chabot, Paris 1927 (text), Louvain 1949 (Latin trans.)
vol. I, p.280, trans. Trombley and Watt, pp.613 and notes.
Thus for example when the Roman duces made ready to march against
the Persians in the vicinity of Tella they were hindered by the fact that
the Roman forces that were with them had dispersed to strip the dead.
Olympius, the dux, gave an order to light a beacon and sound the trumpets
in order to gather them, but this call aroused the Persians rather than the
dispersed Roman fighters. As a result the Roman army was badly defeated.
See Pseudo-Joshua, Chronicon, p.277, trans. Trombley and Watt, pp.567.
Pseudo-Joshua, Chronicon, trans. Trombley and Watt, introduction,
pp.xlvixlvii.
Pseudo-Joshua, Chronicon, pp. 27580; trans. Trombley and Watt,
pp.5163.
Pseudo-Joshua, Chronicon, p.300, trans. Trombley and Watt, pp.945.
Pseudo-Joshua, Chronicon, p.307, trans. Trombley and Watt, p.103.
For these attacks see Procopius, Wars, vol. I, passim; see also Malalas,
Chronographia.
Notes to pp. 2324
187
98. Theophanes, Chronographia, vol. II, p. 296, trans. Mango, p. 425;

Chronicon ad Annum Domini 1234 pertinens, ed. and trans. J. B. Chabot,
Louvain 191620, vol. I, pp.224 (text), 176 (trans.).
99. Regarding the date see B. Flusin, Saint Anastase le Perse et lhistoire de la
Palestine au dbut du viie sicle (text, French translation, and commentary), Paris 1992, vol. II, commentaire, pp.704.
100. Sebeos, The Armenian history attributed to Sebeos, trans. with notes R. W.
Thomson, historical commentary by James Howard-Johnston, assistance
from Tim Greenwood, Liverpool 1999, vol. I, p.58.
101. Chron. 1234, pp.224 (text), 176 (trans.).
102. Histoire Nestorienne (Chronique de Sert), ed. and trans. Monsignor A.
Scher, Patrologia Orientalis, 13/4, Turnhout 1973, p.526.
103. See e.g., n. 175:aliquot oppida et castella Persarum in deditionem accepit
vel vi expugnavit.
104. See e.g. Procopius, Wars, passim, where there are numerous examples.
105. Regarding this term see pp. 245.
106. See Phillipson, Law and Custom, vol. II, pp. 23941; see Onasander,
On the General, in Aeneas Tacticus, Asclepiodotus, Onasander, with an
English Translation by the Members of the Illinois Greek Club, Cambridge,
MA, and London 1977, chaps. 378.
107. Kehne, Treaties, Upholding of.
108. C. T. Lewis and C. Short, A Latin Dictionary, Oxford 1879, p.747, s.v. 1.
fides, IIB (esp.IIB2).
109. See Phillipson, Law and Custom, vol. I, p.391.
110. E. Badian, Deditio, Brills New Pauly. Antiquity volumes, ed. Hubert
Cancik and Helmuth Schneider, Leiden 2010 (Brill Online).
111. Some claim that the act of surrender meant actually giving yourself up
to Roman trust, an act which had no legal significance but was rather a
plea for the mercy of the Romans. See Y. Shahal, Deditio per Pactionem,
MA thesis, Hebrew University of Jerusalem, 2005 (Hebrew), citing E.
Badian, Hegemony and Independence, Prolegomena to a Study of the
Relations of Rome and the Hellenistic States in the Second Century BC,
in J. Harmatta, ed., Proceedings of the VII Congress of the International
Federation of the Societies of Classical Studies, Budapest 1984, pp.397
414; W. Dahlheim, Struktur und Entwicklung des rmischen Vlkerrechts
im dritten und zweiten jahrhundert v. Chr., Munich 1968, pp.4852; A.
M. Eckstein, Glabrio and the Aetolians:A Note on Deditio, Transactions
and Proceedings of the American Philological Association, 125 (1995),
pp.27189, esp.p.276. Others maintain that by accepting the deditio in
fidem the Romans accepted responsibility for the safety and well-being
of the vanquished. See E. S. Gruen, Greek Pistis and Roman Fides,
Athenaeum, 60 (1982), pp. 5068; A. Watson, International Law in
Archaic Rome, Baltimore and London 1993, pp.503; D. Nrr, Aspekte
des rmischen Vlkerrechts die Bronztafel Alcantara, Munich 1989,
p.136.
112. Shahal, Deditio per Pactionem. I am grateful to Prof. Hannah Cotton for
referring me to this work.
188
Notes to pp. 2429
113. This term does not appear in classical Greek, but was widely used in the
Byzantine period as a Greek adaptation of the Latin pactum. See for example John Malalas in the case of the agreement made by Antioch with the
Persians:kata suntaxin philiken kai pakta kata idian proairesin pakta
eirenes kai hupotages stesanto4n, Chronographia, chap. 3.11, p. 205, ll.
369; the term appears often in the Egyptian Papyri in Greek in reference
to contracts of marriage or business contracts. See the data-base of the
Papyri in CD ROM no. 7 of the Packard Humanities Institute, Papyri,
under pakton, passim.
114. Phillipson, Law and Custom, vol. I, p.380.
115. H. G. Liddell and R. Scott, GreekEnglish Lexicon, Oxford 1948, s.v.
homologia, no. 3.
116. Shahal, Deditio per Pactionem, p.48.
117. Ibid., p.34 citing Polybius, Histories 36.9.15.
118. Shahal, Deditio per Pactionem, p.34, citing Cicero, Pro.Balbo 15
119. Shahal, Deditio per Pactionem, p.40, citing Cicero, Pro.Balbo 29.
120. Heather, Foedera, p.616.
121. Ibid., p.656.
122. Ibid., p.71.
123. See n. 86.
124. Malalas, Chronographia, ch. 11.3, p.205, ll. 3539; trans. Jeffreys et al.,
pp.1434.
125. Eutropius, Eutropi Breviarium, book 10, chap.16, p.76; Eutropius, The
Breviarium, p.69.
126. Ammianus, Ammianus Marcellinus, XXIV. 2. 12; see also XXIV.1.69,
where the capture of Ana4ta4 is described. Although not said in so many
words, it is most probably a surrender with agreement: But on drawing
near the walls he [i.e. Julian] considered that a battle must be accompanied
by many dangers, and accordingly, partly in mild terms, partly in harsh and
threatening language, he urged the defenders to surrender. They asked for a
conference with Ormizda, and were induced by his promises and oaths to
expect much from the mercy of the Romans. Finally, driving before them
a garlanded ox, they came down in suppliant guise. At once the whole fortress was set on fire; Pusaeus, its commander, later a general in Egypt, was
given the rank of tribune. As for the rest, they were treated kindly, and with
their families and possessions were sent to Chalcis, a city of Syria.
127. Ibid., XXIV.2, 1821; trans. Rolfe, p.421.
128. Procopius, Wars, 2.27.4546; trans. Dewing, vol. I, p. 515, with my
changes.
129. See Liddell and Scott, Lexicon, s.v. homologia, 3b.
130. Procopius, Wars, 2.26.1222.
131. This represents in fact a more general Persian policy. Thus, for example,
when the Persians, who had conquered Amida, agreed to relinquish it to
the besieging Byzantines, a proposal was discussed among them to the
effect that the Persians will hand the city over to the Romans for the sum
of 1000 pounds of gold (see ibid., 1.9.4). The inhabitants of Hierapolis
(Membij) agree to pay 2,000 centenaria if Khusro would raise the siege (see
Notes to pp. 2934
132.
133.
134.
135.
136.
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
189
ibid., 2.6.2425), and Megas offered Khusro 10 centenaria to withdraw

from the territory of the Byzantine empire (ibid., 2.6.25).
Ibid., 2.7.3435, trans. Dewing, p.323.
Ibid., 2.5.13, trans. Dewing, pp. 297, 299.
Ibid., 2.5.1431, trans. Dewing, pp. 299305.
Ibid., 2.5.2833, trans. Dewing, pp. 303, 305.
Sebeos, Armenian History, vol. I, p.63.
Regarding the Syriac terminology, see also p. 34.
Chron. 1234, vol. I, pp.226 (text), 1778 (trans.).
Ibid., pp.235, l. 27 (text), 184 (trans.).
Pseudo-Joshua, Chronicon, pp.315 (text), 116 (trans.).
On this term see the discussion below.
Amended here from ma4n, belief or faith, to ama4n, as in the previous
sentence.
Eutychius (Sa3d b. Bitr3q), Kita4b al-tarkh, ed. L. Cheikho, Beirut 1905
part II, pp.56 (my translation).
See nn. 11718.
Shahal, Deditio per Pactionem, p.34, citing Polybius, Histories, 36.9.15.
E. W. Lane, ArabicEnglish Lexicon, Cambridge 1984, p.101B.
J. Schacht, Ama4n, EI2, vol. I, pp.42930.
See also E. Tyan, Institutions du droit public, Paris, 1954, vol. I, pp.607,
426.
See al-Bala4dhur3, Bulda4n, p.23; Yahya4 b. A"dam al-Qurash3, Kita4b al-khara4j,
ed. T. W. Juynboll, Leiden 1896, para. 89.
Abu4 Yu4suf, Kita4b al-khara4j, p. 51 (ala4 an yasu4nahum wa-yahquna
dima4ahum).
Al-Bala4dhur3, Bulda4n, p.59.
Ibid., p.65.
Abu4 Ubayd al-Qa4sim b. Salla4m, Kita4b al-amwa4l, Cairo 1353 H, p. 20,
para. 52 (=Abu4 Ubayd al-Qa4sim b. Salla4m, Kita4b al-amwa4l, ed. M. Ima4ra,
Beirut 1989, p.92, para. 52).
An exception is found in al-Tabar3, Tarkh, vol. II.3, p.1769, when during
a raid on Cyprus (125 H) a group of the inhabitants chose the protection of
the Muslims jiwa4r al-muslimn and were transferred to al-Sha4m, while
others chose Byzantium and moved there.
See al-Bala4dhur3, Bulda4n, pp.113, 115, 116, 127, 130, 133, etc.; see the
appendix to Qa4d3, Madkhal, including many versions of agreements,
where the term ama4n is prevalent.
In early north-west Semitic inscriptions the word mn means firmness, steadfastness, faithfulness, probably to indicate a sure covenant.
See J. Hoftijzer and K. Jongeling, Dictionary of the North-West Semitic
Inscriptions, vol. I, Leiden 1995, p.72). In a Semitic inscription on a statue
of Hadad found in Zinjirili the idiom ve-aman karat b is used, referring to
the making of a treaty, using the verb karat to cut, common later both in
Hebrew (karat brit) and in Greek (horkia temnein). The word amana (sic)
for covenant appears in the book of Nehemiah 10:1, 11:23, as well as in the
190
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
Notes to pp. 3435

Covenant of Damascus from Qumran (e.g. Covenant of Damascus 20, l. 12
in M. Broshi, ed., The Damascus Covenant Reconsidered, Jerusalem 1992,
p. 47: ve-maasu be-brit ve-amana asher qiymu be-eretz dameseq ve-hu
brit hadasha. See also 4Q455.2). In the Babylonian Talmud the word carries the meaning of faith, trust, just like the basic meaning of the Greek
pistis and the Latin fides. Thus, amana hayu4 devareinu4 means our words
were based on trust, shtar amana is a bill of indebtedness signed in trust,
shamar amana means to deal in good faith, and amanat ha-midot means
honesty in measures (See M. Jastrow, Dictionary of the Targumim, the
Talmud Babli and Yerushalmi, and the Midrashic Literature, vol. I, New
York 1950, pp.77b78a).
See C. Brockelmann, Lexicon Syriacum, Gttingen 1908, p. 653, s.v.
qeya4ma4, 8a/b.
F. Lkkegaard, Bakt, EI2, vol. I, p.966.
Hinds and Sakkout, A Letter, p.229.
Ibid., pp.220 (text), 227 (trans.), which expressly relates to what you owe
of the baqt about which a peace agreement was made with you; and line
534 (pp.222 (text), 228 (trans.)):For he [Salm] mentioned to me that
you are liable to the baqt of outstanding years, which you have not made
over.
Taq3 al-D3n Ahmad ibn Al3 al-Maqr3z3, Kita4b al-mawa4iz wa-l-itiba4r bidhikr al-khitat wa-l-a4tha4r, Cairo 1270 H, vol. I, p.199.
Bala4dhur3, Bulda4n, p.199.
Ibid.
Ibid., pp. 133, 200. See R. Dozy, Supplemnt aux dictionnaires arabes,
Leiden 1881 (repr. 1991) vol. II, p.377.
See Phillipson, Law and Custom, vol. I, p.391.
It should be noted that this chronicle relies on Dionysius of Tell-Mahre4, a
Jacobite patriarch (d. 845), who relied in his turn on earlier sources. On
this source see L. I. Conrad, Syriac Perspectives on Bila4d al-Sha4m during
the Abbasid Period, in M. A. Bakhit and R. Schick, eds., Bila4d al-Sha4m during the Abbasid Period, 132 AH/750 AD451 AH/1059 AD:Proceedings
of the Fifth International Conference, vol. I, Amman 1991, pp. 144, at
pp.1740; Hoyland, Seeing Islam, pp.41619; see also the introduction
to his forthcoming Theophilus of Edessas Chronicle (590750s) and the
Circulation of Knowledge in Late Antiquity and Early Islam, Translated
Texts for Historians, Liverpool forthcoming (2011).
For these expressions see Pseudo-Joshua, Chronicon, p.315 (text); trans.
Trombley and Watt, p.116; Chron. 1234, pp.240, ll. 1314, 248 l. 10249
l. 14, 254, ll. 2021, 255, ll. 89, 2567 (text). See also p. 31.
Regarding the term qya4ma4, see p. 54.
See J. Payne-Smith, A Compendious Syriac Dictionary, Oxford 1903,
p.230; for examples see Chron.1234, vol. I, pp.248, l. 18, 249, l. 1011,
256 (text); for early usage see Pseudo-Joshua, Chronicon, chap.7, para. 98.
The term riksu u ma4mtu (bond and oath) was common in Mesopotamia
and Anatolia from the middle of the second millennium BCE, while in
Greek the horkia pista is found in Homer, and the expression horkos kai
Notes to pp. 3539
172.
173.
174.
175.
176.
177.
178.
179.
180.
181.
182.
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
194.
195.
191
syntheke appears in Greek treaties. See Weinfeld, Covenant Terminology,

pp.1901.
Chron. 1234, vol. I, p. 255, ll. 89 (text). See also parallel usages on
pp.248, 249.
See e.g. the case when Pharos was conquered by the Romans; its people
were given their polis, their ancestral customs (tous patrious nomous),
and their land. See A. M. Eckstein, Pharos and the Question of Roman
Treaties of Alliance in the Greek East in the Third Century BCE, Classical
Philology, 94 (1999), pp.395418, at p.398. When Antiochus III took over
Jerusalem he swore, according to Josephus, to allow them to live according to their ancestral customs kata tous patrious nomous. See Josephus,
Antiquitates Iudaicae, 12.142.
See pp. 234.
A. Noth, Zum Verhltnis von kalifaler Zentralgewalt und Provinzen in
umayyadischer Zeit: Sulh-Anwa. Traditionen fr gypten und den
Iraq, Die Welt des Islams, 14 (1973), pp.15062.
Qa4d3, Madkhal, pp.25167.
Al-Tabar3, Tarkh, vol. I, pp. 2897900; for translation see R. S.
Humphreys, annot. and trans., The History of al-Tabar, vol. XV: The
Crisis of the Early Caliphate, Albany 1985, pp.1025.
On the usage of this term see M. Lecker, Appendix E:The Term Muwa4daa,
in Lecker, Constitution, pp.2045.
See al-Bala4dhur3, Bulda4n, p.201; trans. P. K. Hitti as The Origins of the
Islamic State, New York 1916, vol. I, pp.31516 (except last sentence).
Qura4n 2.279:fa-dhanu4 bi-harbin; see Lane, Lexicon, vol. I, p.42 A, s.v.
adhina.
See pp.334, nn. 1516.
See Abu4 Ubayd, Amwa4l, pp.201, para. 52 (=Amwa4l, ed. Ima4ra, p.92,
para. 52); al-Bala4dhur3, Bulda4n, p.79.
Al-Bala4dhur3, Bulda4n, p. 59; the agreement given by the Prophet to the
people of Ayla was called amana (sic!):

. See e.g. Ibn Asa4kir, Tarikh, vol. II, p. 41; Abu4 al-Fida4 Isma43l
b. Umar b. Kath3r, al-Sra al-nabawiyya, vol. IV, Beirut 1976, p. 30.
Al-Bala4dhur3, Bulda4n, pp.5960.
Ibid., pp.656.
Ibid., p.60.
Ibid., pp.645.
Ibn Abd al-Hakam, Futu4h Misr, p.85.
Ibid., p.89.
Qa4d3, Madkhal, pp.202, 213.
Ibid., p.21718.
See pp. 1819.
Qa4d3, Madkhal, p. 217. See Abu4 Ubayd, Amwa4l, p. 146, para. 400
(=Amwa4l, ed. Ima4ra, p.235, para. 401).
Qa4d3, Madkhal, p. 217; Tabar3, Tarkh, vol. I, pp. 2593, p. 217; Abu4
Ubayd, Amwa4l, ed. Ima4ra, pp. 2956, para. 522 (=Amwa4l, pp. 2078,
192
196.
197.
198.
199.
200.
201.
202.
203.
204.
205.
206.
207.
208.
209.
210.
211.
212.
213.
214.
215.
216.
217.
218.
219.
Notes to pp. 3943

no. 520). Regarding the Nubian baqt see pp. 89; regarding the term see
p. 35.
AlBala4dhur3, Bulda4n, pp.2012.
Ibid., p.218.
Ibn Asa4kir, Tarkh, vol. II, pp.3545.
Ibid., vol. VI, pp.1921.
Ibid., vol. LXIV, pp.12535.
See chap.2.
See Qa4d3, Madkhal, pp.2238.
Phillipson, Law and Custom, vol. I, pp.38590.
See all three versions of the Ruha4 agreement:Qa4d3, Madkhal, p.251; see
also Cohen, Pact of Umar, p.143.
Qa4d3, Madkhal, p. 256 (Tifl3s). See also p. 259 (Baalbakk): shahda
Alla4hu, wa-kafa4 bi-Alla4hi shahdan.
Probably referring to Babylon/Memphis (Manf). See A. D. Beihammer,
Quellenkritische Untersuchungen zu den gyptischen Kapitulationvertrgen
der Jahre 640646, Vienna 2000, p.64, n. 156.
Tabar3, Tarkh, vol. I, p.25889; H. A. Juynboll, annot. and trans., The
History of al-Tabar, vol. XIII:The Conquest of Iraq, Southwestern Persia,
and Egypt, Albany 1989, p.171.
Maqr3z3, Khitat, vol. I, p.200.
In the Constitution of Medina the treaty also ends with the statement wainna Alla4h ala4 atqa4 ma4 f ha4dhihi al-sahfa wa-abarrihi; see Ibn Hisha4m,
Kita4b srat rasu4l Alla4h, ed. F. Wstenfeld, Gttingen 1858, vol. I, part 1,
pp.343.
They were, however, placed in the temples of the gods. See Johnson et al.,
Statutes, nos. 53, 54.
For the treaties see Qa4d3, Madkhal, appendix.
There are three versions cited (see ibid., p.251). Two are typical surrender agreements, while al-Baladhur
4
s
3 version (al-Baladhur
4
,3 Buldan
4 , p.207)
includes also the demand to repair the bridges and guide those who are lost.
See n. 207.
For these two latter cases, see Qa4d3, Madkhal, pp.25960.
Ibid., p.258, Bihquba4dh.
Ibid.
See note 174; see Eckstein, Pharos, pp.3989.
On this see Donald W. Baronowski, The Status of the Greek Cities of Asia
Minor after 190 BC, Hermes, 119 (1991), pp.45063 esp.pp.4612, n.
13. On several other occasions the restitution of liberty, land and laws
follows defeat by Rome or deditio, actual or contemplated:Livy 25.23.4,
29.21.7, 31.31.7, 33.30.2 (Polyb. 18.44.2), 33.32.5 (Polyb. 18.46.5),
37.54.26, 38.44.4, 45.29.4; Polyb. 36.4.4. In each case, the result or the
intention of the Roman proclamation was to guarantee a communitys
independence, subject to some degree of Roman tutelage. On the restitution of freedom by Rome to defeated or surrendered peoples, see Dalheim,
Struktur, pp.6998.
Baronowski, Status, p.452, citing Polyb. 21.46.810.
Notes to pp. 4347
193
220. See W. Dittenberger, Sylloge Inscriptionum Graecorum (SIG), 3rd edn.,

vol.II, Leipzig 191517, no. 618; trans. in Johnson et al., Statutes, no. 23.
221. This is something like the autonomy these cities had under Seleucid rule.
See Baronowski, Status.
222. See Johnson et al., Statutes, e.g. nos. 19, 20, 25; many others appear in H.
Bengston, Staatsvertrge des Altertums, Kommission fr Alte Geschichte
und Epigraphik des Deutschen Archaeologischen Instituts, Munich 1960
9, as well as in various inscriptions revealed in the Greek world, to be
found in Dittenberger, SIG.
223. See above, Treaties with the Sasanians; Treaties with the Barbarians and the
Arabs.
224. See Polybius, Histories; Diodoros Siculus, Dio Cassius, passim.
225. See above and nn. 1734.
226. Blockley, ed. and trans., Menander the Guardian, pp.747.
227. This appears in different variations in the ama4n to Misr, as it appears in
Abu4 Ubayd, al-Tabar3, and Ya4qu4t: see Qa4d3, Madkhal, pp. 2567; or
Tifl3s in Abu4 Ubayd, al-Bala4dhur3, al-Tabar3, and Ya4qu4t:see ibid.
228. See e.g. in the case of Jurja4n, ibid., p.255; Ma4h D3na4r, ibid., p.260; Ma4h
Bahra4dha4n, ibid., p.261; Qu4mis, ibid., p.263; A"dharbayja4n, ibid., p.264;
Mu4qa4n, ibid., p.266.
229. Strength or power:see Lane, Lexicon, p.1799, col. 3.
230. Qa4d3, Madkhal, p.256, Tifl3s.
231. Ibid., Marwaru4dh.
232. Ibid., Tifl3s.
233. Ibid., p.263, Qu4mis.
234. Ibid., p.266, Mu4qa4n.
235. Ibid., pp.256, 257, 261, 266.
236. Ibid., p.256, Tifl3s.
237. Ibid., p.251 (al-Ruha4), p.260 (al-Ruha4, another version).
238. Ibid., p. 261 (Ma4h Bahra4dha4n); the Hara4t treaty has: wa-isla4h ma4 taht
yadayhi, ibid., p.265; see also Abu4 Yu4suf, Kita4b al-khara4j, p.138.
239. See James B. Pritchard, ed., Ancient Near Eastern Texts Relating to the Old
Testament, trans. and annot. W. F. Albright et al., Princeton 1969 (hereafter
ANET), pp.2035, the treaty between Mursilis the Hittite king and DupiTessub of Amurru; the treaty between Supiluliumas the Hittite and Aziras
of Amurru; Sefire III (c. 750 BCE), in J. A. Fitzmyer, The Aramaic Treaties of
Sefire, rev. edn, Rome 1995, pp.13741 (=ANET, p.6601); the vassal treaties of Esarhaddon, ANET, pp.53441. Despite the existing differences due,
no doubt, to a great extent to the long span of time the striking similarity
found between the Hittite, Assyrian, and Aramaic treaties has been noted and
discussed by various scholars. For a survey see Fitzmyer, Treaties, pp.1626.
240. See Bengston, Staatsvertrge, vol. III, the treaty between Rhodos and
Hierapytna, pp. 31417, where both poleis take upon themselves as
mutual obligation to come to each others aid with all their might in case
their polis, harbours, or military bases, their routes, or their laws and stable
democracy are threatened. They also commit to being well-wishers, friends,
and allies always. See ibid. for many additional examples.
194
Notes to p. 47
241. See for example the treaty between Rome and the Aetolians, Johnson et al.,
Statutes, no. 24; Rome and Antiochus III, ibid., no. 27; Rome and the Jews,
ibid., no. 35; treaty with Cnidos in 45 BCE, ibid., no. 110.
242. See e.g. the treaty of 382 with the Goths, p. 17, and the treaty of 434 with
the Huns, ibid.
243. See pp. 1820.
244. The use of terms such as vassal or feudal, borrowed from medieval
Europe, may of course raise some eyebrows. However, it has been shown
by Widengren as justified. On this issue see Rubin, Nobility, p. 246, n.
27, who cites and translates Widengrens definition in his Feudalismus im
alten Iran, Cologne and Oplanden 1967, p.12, n. 10, according to which
feudalism consists in the existence of a warrior class that committed itself
to military service; the obedience of this warrior class towards the lord as
symbolized by a ceremony; and enfeoffment of the individual warriors with
the possession of an estate in return for active military service. See, however, Richard Fryes reservations in his Feudalism in Sasanian and Early
Islamic Iran, Jerusalem Studies in Arabic and Islam, 9 (1987), pp.1318,
although he notes that adequate categories of explanation or new expressions are yet to be found to describe the society or political system of both,
p.18; M. Shaki, Class System, Encyclopedia Iranica, vol. III, pp.65458.
245. G. Widengren, Recherches sur le fodalisme iranien, Orientalia Suecana,
5 (1956), pp.79182, at pp.8995.
246. Abul-Qasim Firdaus3, Sha4h-na4ma, ed. Djala4l Kha4ligh3-Mutlagh, New
York and Costa Mesa, 19882008, vol. II, p. 62, ll 8478 (=Firdaws3,
Sha4hna4mah, French trans. J. Mohl, Tehran 19336, vol. I, p.448).
Line 847:kunu4n khilat-i sha4h ba4yad nukhust yak- ahd-u muhr- baru4-bar durust
ki ta4 zinda ba4shad bi-Ma4zandara4n parastish kunand-ash hama mihtara4n
One of the MS variants has manshu4r instead of muhr:see vol. II, p.62,
n. 15. (The Russian translation is to grant him a royal edict authenticated
by a seal; see Firdousi, Sha4h-na4ma, prepared by T. B. Banu, A. Lahuti, and
A. A. Starikov, Moscow 1957, vol. I, p.407).
Line 848 So that as long as he is alive all the nobles in Ma4zandara4n
would serve him.
An addition in one of the MSS (see vol. II, p.62, n. 17) might clarify the
meaning of granting the edict and the seal:
yak- cha4kir- nk ba4shad tu-ra4
firistad ham ba4zh-i andar-kvara4
He will be a good servant to you [i.e. to Kavus] He will keep sending
fitting tribute
247. Firdaus3, Sha4h-na4ma, vol. II, pp.634, vv. 86071 (=Firdaws3, Sha4hna4mah,
vol. I, p.449).
Among the description of the gifts granted to Rustam by Kavus:
p.63, line 859:nibishta yak- ahd-i u4
bi-mushk-u may-u u4d
bar harr
dast-i dabr
In the Arabic translation by al-Bunda4r3 (Abu4 al-Qa4sim Firdaus3, Sha4hna4ma, trans. al-Fath b. Al3 al-Bunda4r3, ed. Abd al-Waha4b Aza4m, Cairo
1350/1932): !
" #"$ %%
& , the term manshu4r refers to letters
patent.
Notes to pp. 4849
195
248. Firdaus3, Sha4h-na4ma vol. IV, pp. 3558, vv. 2870915 (=Firdaws3,
Sha4hna4mah, vol. IV, pp.2016). See also Ibn al-Balkh3, Farsnamah, ed. G.
Lestrange and R. A. Nicholson, London 1921, p.43, who describes how
Kay Ka4vu4s liberated Rustam and gave him a pact (ahd), awarding him
Sista4n and Za4bulista4n.
249. Moses Khorenatsi, History of the Armenians, book II, chap.84, trans.
R. W. Thomson, Harvard 1980, p.236:Trdat, delighted with the news,
wrote an edict that he [Mamgon] was to have authority over all [the
lands] that he promised and he made him a prince with the title
Mamgonean after his own name; Thomas Artsruni, History of the
House of Artsrunik ed. and trans. R. W. Thomson, Detroit 1985, book I,
chap.7, p.116.
250. On the dates of Sha4pu4rs reign see R. Frye, The Political History of Iran
under the Sasanians, in E. Yarshater, ed., The Cambridge History of Iran,
vol. III(1): The Seleucid, Parthian and Sasanian Periods, Cambridge 1983,
p.119.
251. On the status of the nobility under Sasanian rule see Rubin, Nobility,
p.243; for the inscriptions see M. Back, Die sassanidischen Staatsinschriften,
Tehran and Leiden 1978, pp.3728; Huyse, Ka4ba-i-Zardut.
252. H. Humbach and O. Skjaerv, The Sasanian Inscription of Paikuli, Wiesbaden
197883.
253. P. Pourshariati, Decline and Fall of the Sasanian Empire, London 2007.
254. Sebeos, Armenian History, vol. I, p.5.
255. This refers to Kava4dh I (r. 488531). According to Lazars history he was
appointed by Valarsh, who reigned before Kava4dh, 4848. See ibid., nn.
18,26.
256. The sparapet, or commander-in-chief, of the Mamikonean house. See ibid.,
vol. I, p.4, n. 18.
257. Ibid., pp. 67; see also Pourshariati, Decline, p. 126, citing Sebeos,
Armenian History, pp.173, and 133; citing ibid., pp.434, on this sensitive issue of the marzuba4nate of Armenia.
258. Pourshariati, Decline, p.275.
259. Al-Tabar3, Tarkh, vol. I, pp.26656; Qa4d3, Madkhal, p.265.
260. Al-Tabar3, Tarkh, vol. I, p.2900; Qa4d3, Madkhal, p.265.
261. See Rubin, Nobility, p. 246; V. G. Lukonin, Political, Social and
Administrative Institutions: Taxes and Trade, in E. Yarshater, ed., The
Cambridge History of Iran, vol. III (2):The Seleucid, Parthian and Sasanian
Periods, Cambridge 1983, pp.681746, at p.700; see also A. Perikhanian,
Iranian Society and Law, in Yarshater, ed., The Cambridge History of
Iran, vol. III(2), pp.62780, at p.632.
262. See Pourshariati, Decline, p.248; al-Tabar3, Tarkh, vol. I, p.2658, G. Rex
Smith, annot. and trans., The History of al-Tabar, vol. XIV:The Conquest
of Iran, Albany 1994, p.29.
263. See Pourshariati, Decline, pp. 2745; al-Tabar3, Tarkh, vol. I, p. 2664;
'($ $)+ ,- %/ 01 $ 2'3. (G. R. Smith, annot. and trans., The History
of al-Tabar, vol. XIV: The Conquest of Iran, Albany 1994, p. 35: Our
tribute to you will be the military assistance we render youbut do not
humiliate us with tribute.).
196
Notes to pp. 4952
264. See Pourshariati, Decline, p.251, citing al-Tabar3, Tarkh, vol. I, p.2659,
Smith, Conquest, p.30.
265. This is the only source to claim that there was one agreement for the
whole territory of al-Sha4m. Other sources such as al-Bala4dhur3, al-Tabar3,
and al-Yaqu4b3 all adduce traditions regarding specific agreements for the
various cities of al-Sha4m, including Jerusalem, Tiberias, Askalon, Hims,
Damascus, and others. See Hill, Termination, pp. 6575. It thus seems
quite conspicuous that this is a later invention intended to generalize the
local agreements made. On this version see the following chapter.
266. Abu4 Yu4suf, Kita4b al-khara4j, p.138.
267. See pp. 89.
268. Johnson et al., Statutes, no. 3, p.7.
269. Ibid., no. 27, pp.256.
270. See p. 17.
271. Peter the Patrician, fragments 1314, trans. in Dignas and Winter, Rome
and Persia, p.124.
272. Blockley, ed. and trans., Menander the Guardian, pp. 703; Dignas and
Winter, Rome and Persia, p.1413.
273. R. H. Charles, trans., The Chronicle of John (c. 690) Coptic Bishop of
Nikiu, London 1916, chap.120, paras. 1721.
274. See e.g. Johnson et al., Statutes, nos. 24, 27; regarding the 363 treaty see
Ammianus, Ammianus Marcellinus (nn. 2831): When this shameful
treaty (decretum) was concluded, lest anything contrary to the agreements
(pactum) should be done during the truce, distinguished men were given
by both sides as hostages:from our side Nemota, Victor, and Bellovaedius,
tribunes of famous corps, and from the opposite party Bineses, one of the
distinguished magnates, and three satraps besides of no obscure name; for
the 363 treaty see p. 13; regarding the Goths see p. 17.
275. Charles, trans., John of Nikiu, chap.117, paras. 13.
276. On the surrender of Jerusalem and the agreement signed between the
Muslims and the inhabitants of the city see al-Tabar3, Tarkh, vol. I,
pp. 24036. See also M. Gil, A History of Palestine 6341099, trans. E.
Broido, New York 1992, paras. 689.
277. Al-Tabar3, Tarkh, vol. I, p.2406. De Goeje and Y. Friedmann, who translated this text into English (Y. Friedmann, annot. and trans., The History of
al-Tabar, vol. XII:The Battle of al-Qa4disiyyah and the Conquest of Syria
and Palestine, Albany 1992, p. 192, n. 711), believe that the expression
their churches and crosses makes little sense in this context and considered it to be a mistake. This is also why it was not included in the translation. It is not, however, inconceivable that Jerusalems exiled inhabitants
retained the right to the possessions they had left behind upon leaving the
city, until such time as they reached a place of safety.
278. See Muhammad b. Abdalla4h al-Azd3, Tarkh futu4h al-sha4m, Cairo 1970,
p.140 (E. W. N. Lees, comp., The Fotooh al-Sham being an Account of the
Moslim Conquests in Syria, by Aboo Ismail Mohammad bin Abd Allah,
al-Azdi al-Bacri, Calcutta 1854, p.123). In the case of Tiberias, Shurahb3l
Notes to pp. 5256
279.
280.
281.
282.
283.
284.
285.
286.
287.
288.
289.
290.
291.
292.
293.
294.
295.
296.
297.
197
b. Hasana made the usual agreement and guaranteed the inhabitants the
safety of their lives, possessions, children, churches, and houses, with the
exception of what they abandoned and left behind them; see al-Bala4dhur3,
Bulda4n, p.115.
S. D. Goitein, Did Omar Prohibit the Stay of the Jews in Jerusalem?, repr.
in his Palestinian Jewry in Early Islamic and Crusader Times in the Light
of the Geniza Documents, ed. J. Hacker, Jerusalem 1980, pp.3642, questions the existence of such a treaty, chiefly because Sayf b. Umar is generally considered a highly unreliable source, but also because the supposed
treaty included an article prohibiting Jews from living in Jerusalem; this
despite the fact that Jews did, in fact, settle in the city shortly after the
Muslim conquest.
Gil, however, believes the treaty to be authentic, both because of its
phrasing and literary style and because it is quite likely that the treaty
contained an article banning the Jews from Jerusalem even if it wasnt
enforced later. On Sayf b. Umars reliability as a source, Landau-Tasseron
argues convincingly that Sayf is no more or no less reliable than other narrators:see E. Landau-Tasseron, Sayf ibn Umar in Medieval and Modern
Scholarship, Der Islam, 67 (1990), pp.126.
On the authenticity of the Jerusalem agreement see M. Levy-Rubin,
Were the Jews Prohibited from Settling in Jerusalem following the
Arab Conquest?:On the Authenticity of al-Tabar3s Jerusalem Surrender
Agreement, Jerusalem Studies in Arabic and Islam, 36 (2009), pp.6381.
Bala4dhur3, Bulda4n, p.200.
I discuss this issue in detail in Were the Jews Prohibited.
See pp. 367.
Al-Tabar3, Tarkh, vol. I, p.2898; Humphreys, Crisis, p.103.
See n. 152; n. 186.
See nn. 148, 156.
On and = ( and )see A. de Biberstein Kazimirski, Dictionnaire
ArabeFranaise, Paris 1860, vol. II, p.1592.
Damascus, Misr, Tifl3s, Jerusalem, Ludd, and Filast3n, see Qa4d3, Madkhal,
pp.252, 254, 256, 259, 260.
Ibid., p.256, Tifl3s.
I agree here with Fattal, who opposes Lammenss theory that this was a
parity treaty. See Fattal, Statut, pp.246.
See p. 9; regarding the term see pp. 345.
Ibn Abd al-Hakam, Futu4h misr, pp. 1889; al-Maqr3z3, Khitat, vol. III,
pp.2902; trans. in P. Forand, Early Muslim Relations with Nubia, Der
Islam, 48 (1972), pp.11221, at pp.11415. For a full list of the translations of this text see Hinds and Sakkout, A Letter, p.210, n. 5.
Forand, Early Muslim Relations with Nubia, pp.11415.
Ibid.
See ibid., ll. 535.
Ibid., p.115.
198
Notes to pp. 5661
298. See Hinds and Sakkout, A Letter, p.216.

299. Forand, Relations, pp. 11314; Qa4d3, Madkhal, pp. 20910; a similar
case of supplying slaves as part of the agreement appears in Ammianus,
Ammianus Marcellinus, 17.13.3, where the surrendering party offers to
pay a yearly tribute, as well as a levy of their able youth (dilectumque
validate iuventutis).
300. See Noth, Vertrge, pp.28690.
301. For examples see pp. 24, 30, 43.
302. Charles, trans., John of Nikiu, chap.120, para. 27.
303. See p. 37. Compare the gifts brought by the bishop of Sura to the Persians,
p. 30.
2. Shuru4t Umar and Its Alternatives
1. It should be emphasized, however, that these reports are exceptional and
that in most cases only taxes are stipulated. Such stipulations exist in the
traditions regarding Tiberias, where al-Bala4dhur3 reports that deserted
houses were to be given over to the Muslims (al-Bala4dhur3, Bulda4n,
p.116), while al-Yaqu4b3 and al-Tabar3 report that it was stipulated that
half the houses in the city were to be given over to the Muslims (al-Yaqu4b3,
Tarkh, Leiden 1882, vol. II, p.159; al-Tabar3, Tarkh, vol. I, p.2159).
Al-Bala4dhur3 adduces a tradition according to which Hims was divided
into lots and in each lot one Muslim house was to be built; in addition,
Muslims were to settle in all unoccupied houses (al-Bala4dhur3, Bulda4n,
p.131). On this question see Levy-Rubin, Changes.
2. See Miller, Catalogue to Codes, with which I became acquainted only in
2007, thanks to Prof. Mark Cohen.
3. See al-Turtu4sh3, Sira4j al-mulu4k, Cairo 1935, pp.2523; for additional editions and English translations see Cohen, Pact of Umar, n. 16. See also
Fattal, Le statut, pp.602 for French translation.
4. Ibn Asa4kir, Tarkh, vol. II, pp.120, 1749.
5. Ibn Qayyim al-Jawziyya, Ahka4m ahl al-dhimma, ed. S. al-Sa4lih, Damascus
1961, vol. II, pp.65762.
6. Ibid., p.664.
7. Cohen, Pact of Umar, pp.11011 and n. 32 there. For the various versions see pp.13740, 1412, 1456.
8. See Abu4 Bakr al-Khalla4l, Ahl al-milal, vol. II, pp.4314.
9. See Cohen, Pact of Umar, p.109 and n. 25. Cohen also cites additional
later versions. Miller adduces a number of versions in the appendix to
Catalogue to Codes.
10. Ahmad b. Husayn al-Bayhaq3, Sunan al-Bayhaq al-kubra4, Mecca 1994,
vol. IX, p.202.
11. Ibn al-Murajja4, Fada4il, pp.556.
12. See Tritton, The Caliphs, pp.1017; Fattal, Le statut, pp.669.
13. Tritton, The Caliphs, p.12; Fattal, Le statut, p.68.
14. Fattal, Le statut, p.68.
15. Ibn Asa4kir, Tarkh, vol. II, pp.175, 177, 178.
Notes to pp. 6164
199
16. Ibn Hajar al-Asqala4n3, Lisa4n al-miza4n, Beirut 1987, vol. III, p.18; Ibn Ab3
Ha4tim al-Ra4z3, Kita4b al-jarh wa-al-tadl, Hyderabad 1952, vol. II, part 1,
p.284.
17. H. P. Raddatz, Sufya4n al-Thawr3, EI2, vol. IX, pp. 7702: F. Sezgin,
Geschichte des arabischen Schrifttums, Leiden 196784, vol. I, pp.51819.
18. Abu4 Zakariya Yahya4 b. Sharaf b. Mar3, Tahdhb al-asma4, Beirut 1996,
vol.I, p.241; Abu4 al-Maha4sin Jama4l al-D3n Yu4suf b. Taghr3 Bird3, al-Nuju4m
al-za4hira, Cairo n.d., vol. I, p. 271; Muhammad b. Ahmad al-Dhahab3,
al-Ibar f khabar man ghabara, Kuwait 19606, vol. I, p.139.
19. Ibn al-Murajja4, Fada4il, pp. 556. Noth already noted the fact that
the isna4d goes back to the middle of the eighth century CE: see Noth,
Abgrenzungsprobleme, pp.291 (German), 104 (English), n. 4.
20. Noth, Abgrenzungsprobleme, pp.30410 (German), 11521 (English). On
the ghiya4r, see also chap.5, n. 204.
21. Noth, Abgrenzungsprobleme, pp.3023 (German), 11314 (English).
22. See chap.4. Information on the restrictions of Umar II is abundant. On the
dress regulations see Abu4 Yu4suf, Kita4b al-khara4j, pp.1278; Abu4 Bakr Abd
al-Razza4q b. Hamma4m al-Sana4n3, al-Musannaf, Beirut 19702, vol. VI, p.61.
On the prohibition on exhibiting the cross see Abu4 Yu4suf, Kita4b al-khara4j,
pp.1389; Abd al-Razza4q, Musannaf, vol. VI, p.61. For non-Muslim testimonies see Theophanes, Chronographia, vol. I, p. 399; Theophanes, The
Chronicle of Theophanes Confessor: Byzantine and Near-Eastern History
AD 284813, trans. C. Mango and R. Scott, Oxford 1997, p. 550; J. B.
Chabot, ed., Chronique de Michel le Syrien, Paris 18991910, vol. II, p.489.
24. See e.g. Abu4 Ubayd, Amwa4l, ed. Ima4ra, p. 179, para. 269 (=Amwa4l, p. 97,
para. 269); Abu4 Yu4suf, Kita4b al-khara4j, pp. 13844 passim; al-Sha4fi3, Kita4b
al-umm, vol. IV, p. 293; Abd al-Razza4q, Musannaf vol. VI, p. 61; Abdalla4h
Muhammad b. Ab3 Shayba, Musannaf Ibn Ab Shayba al-Ku4f, Riyadh 1409
H, vol. VI, p. 467. Especially interesting is a hadth adduced by Abu4 Ubayd
concerning the meeting between Umar and the people of Adhra4t, who
received him carrying swords. Umar was appalled, and ordered that they
be sent back, but Abu4 Ubayda remarked that it was sunnat al-ajam (the
custom of the non-Arabs) and such an action would be considered a violation of the treaty. Umar accepted this view, and received them. Abu4 Ubayd
goes on to explain that although Umar resented this custom he nevertheless
accepted it because their customs, prayer-house et alia were all part of the
treaty which could not be violated ( 45) 67 8 9 :/)<; : Abu4 Ubayd,
Amwa4l, ed. Ima4ra, p. 243, para. 426 (=Amwa4l, p. 152, para. 425).
25. The legal discussions over this term will be discussed below.
26. Abu4 Ubayd, Amwa4l, ed. Ima4ra, p. 179, para. 269 (=Amwa4l, p. 97, para.
269):= 25 7 %> - $ 7 A - 6C 8- %A +%/ %/ $
E 0F - = " 9 !
- E < $ $ .<$ - %4 7 %' > 27. Abd al-Razza4q, Musannaf, vol. X, p. 320 (cf. vol. VI, p. 60): "%/ $ $
7 <$ $F- %4 7 $F- :K 7 $ 7 LM 7 A 7 " - ;-%+ ,0F
/ 0F K = " A- $
M
M P)$ $ %' > 7 %> $F- >. See
200
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
Notes to pp. 6469

also other versions in Ibn Ab3 Shayba, Musannaf, vol. VI, p. 467; al-Bayhaq3,
Sunan, vol. IX, p. 2012.
Abd al-Razza4q, Musannaf, vol. X, p.320; cf. vol. VI, p.60.
See ibid., vol. VI, pp.5960, esp.p.60, no. 10001, vol. X, pp.31920. The
full term amsa4r al-muslimn appears in this discussion at the beginning of the
chapter The destruction of their churches and should they ring the na4qu4s?
in vol. X, p.319. Later he uses just amsa4r, but the meaning is clear.
Abd al-Razza4q, Musannaf, vol. VI, p.61, vol. X, p.320.
Abu4 Ubayd, Amwa4l, ed. Ima4ra, p. 179, para. 269 (=Amwa4l, p. 97, para.
269).
Abu4 Ubayd, Amwa4l, ed. Ima4ra, pp. 1813, paras. 2779 (=Amwa4l,
pp.99102, paras. 2779).
Abu4 Ubayd, Amwa4l, ed. Ima4ra, p.182, para. 277 (=Amwa4l, p.100, para.
277).
9
2- 6 " %/ $ , Sha4fi3, Kita4b al-umm, vol. IV, p. 293.
Ibid.: .. 0F 1 0 , $SAbu4 Ubayd explains that house of chastisement refers to churches, synagogues, and fire-temples, which are not to dwell together with mosques
in amsa4r al-muslimn: A 9 $< .9 P 9 1 6 P U 7
" $/ - 3$" ; 1+ U 7 5 .% ; 8W$ 1. Abu4 Ubayd,
Amwa4l, ed. Ima4ra, p. 177, para. 263 (=Amwa4l, p. 95, para. 263).
Muhammad b. al-Hassa4n al-Shayba4n3, al-Siyar, ed. M. Kha4ddu4r3, Beirut
1975, p. 265; Muhammad b. al-Hassa4n al-Shayba4n3, The Islamic Law of
Nations:Shayba4ns Siya4r, trans. M. Khadduri, Baltimore 1966, p.278.
Muhammad b. Jar3r al-Tabar3, Kita4b ikhtila4f al-fuqaha4, ed. J. Schacht,
Leiden 1933, p.236, cited in S. Ward, A Fragment from an Unknown Work
by al-Tabar3 on the Tradition Expel the Jews and the Christians from the
Arabian Peninsula (and the Lands of Islam), Bulletin of the School of
Oriental and African Studies, 53 (1990), pp.40720, at pp.41314.
Al-Shayba4n3, Siyar, p.265.
Muhammd b. Ahmad al-Sarakhs3, Sharh kita4b al-siyar al-kabr, ed. Abd
al-Az3z Ahmad, n.d., n.p., vol. IV, p. 1535 (no. 3016): %/ $M ;X

- $5+ (; - ;(+ " See also pp. 1531, 1533.
Ibid., pp.15334.
Abd al-Razza4q, Musannaf, vol. VI, pp. 5961, vol. X, pp. 31920; Abu4
Ubayd, Amwa4l, ed. Ima4ra, pp.1767, para. 262 (=Amwa4l, p.95, para. 262).
Al-Sarakhs3, Sharh, p.1529 (no. 3003).
The question of the residence of dhimms in Muslim amsa4r is discussed in
Friedmann, Tolerance, pp.923.
For a discussion of this see Ward, A Fragment, pp.40913.
See ibid., pp.41314.
Al-Sarakhs3, Sharh, pp.15367 (no. 3020).
Ibid., pp.15378 (no. 3023).
B. Jokisch, Islamic Imperial Law, Berlin and New York 2007, pp.34.
Ibid., pp.313, 1048.
See ibid., pp.34, 7796.
Ibid., pp.34.
Notes to pp. 6973

53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
201
See Cohen, Pact of Umar, p.128.

See below, and chap.4.
See pp. 634.
The dating of Kitab
4 al-kharaj4 and the writings of other early Muslim jurisprudents has been challenged by Norman Calder. In his book Studies in Early
Muslim Jurisprudence, Oxford 1993, Calder claimed that Abu4 Yusufs
4 Kitab
4
al-kharaj4 as well as al-Shafi
4 s
3 Kitab
4 al-umm and other texts are, in fact, a product of Muslim society in the third century. Calders theory has been rejected
by many scholars. See e.g. the opinions of Y. Dutton, Review of N. Calders
Studies in Early Muslim Jurisprudence, Journal of Islamic Studies, 5 (1994),
pp.1028; H. Motzki, The Prophet and the Cat:On Dating Maliks
4
Muwatt a
and Legal Traditions, Jerusalem Studies in Arabic and Islam, 22 (1998),
pp.1883 esp.n. 19; and M. Q. Zaman, Religion and Politics under the Early
Abbasids:The
4
Emergence of the Proto-Sunn Elite, Leiden 1997, pp.913.
Noth, Abgrenzungsprobleme, pp. 313 (German), 122 (English), n. 99,
para.2.
Ibid.
Abu4 Yu4suf, Kita4b al-khara4j, p. 138. The words ba and kansa are not
unequivocal, and could both be understood as both synagogue and church.
However, kansa is the most common word for church, and so it is most
likely that ba would stand here for another kind of prayer-house, most
probably a synagogue. This is also the translation in E. Fagnan, Livre de
limpt foncier, Paris 1921, p. 213. See e.g. Abu4 Ubayd, Amwa4l, ed. Ima4ra,
p. 177, para. 263 (=Amwa4l, p. 95, para. 263) where he ennumerates side
by side % ; 8W$ 1, most probably churches, synagogues, and
fire-temples.
Here the text is not clear; it may also be understood to mean that the Muslims
should fight an enemy who would attack the dhimms. But see Abu4 Yu4suf,
Kita4b al-khara4j, p.143, which will be cited below, where the text expressly
says that they would be allowed to keep their fortresses in order to defend
themselves.
This appears in some of the versions and omitted in others; see Abu4 Yu4suf,
Kita4b al-khara4j, p.138, n. 1.
On Makhu4l al-Sha4m3 see Husayn Atwa4n, al-Riwa4ya al-tarkhiyya f bila4d
al-sha4m f al-asr al-umaw, Beirut 1986, pp.1035, and see also bibliography there.
See p. 59. The display of these banners, which were most probably embroidered with icons or saints, may have been considered shirk. In addition, it
was also seen as an encroachment upon regal privileges, as banners and standards were customarily used by the caliphs army. See chap.5, nn. 31112.
Abu4 Yu4suf, Kita4b al-khara4j, p.138.
Ibn Asa4kir, Tarkh, vol. II, pp.1234.
Ibid., vol. VIII, pp.18793. See also Sezgin, Geschichte, vol. I, pp.2934.
Ibn Asa4kirs version, after the prohibition on introducing pork into Muslim
neighbourhoods, also includes the prohibition on introducing wine, which
(perhaps as a result of negligence) was omitted in Abu4 Yu4sufs version.
202
Notes to pp. 7377
69. Ibid.
70. " 9 =9 0 7: ibid., p. 144.
71. Ibid., p.146. Al-Nak3b is found between Tabu4k and Maa4n on the hajj route,
according to Mara4sid al-ittila4 ala4 asma4 al-amkina wal-biqa4, ed. T. G. J.
Juynboll, Leiden 1854, vol. III, p.228, while al-Kawa4thil is at the edge of
al-Sha4m:ibid., vol. II, p.517 (cited by Fagnan, Impt foncier, p.227).
73. Ibid.
74. Ibid.
75. Ibid., p.139.
76. For Dionysius see A. Palmer et al., The Seventh Century in the West-Syrian
Chronicles, Liverpool 1993, pp.1567; al-Bala4dhur3, Bulda4n, p.137.
78. Ibid., pp.13940.
79. Ibid., p.141.
80. Ibid.
81. See p.74, nn. 734.
82. 5 8W$ 1 F+
%9 ='' A 9 %9 L2 :$< ;-$) L 9 )%>
)$[ $FF-''A 9 %9 $2= 9 F&- $9 $3 < $- 9S- F +F].. See
Abd al-Razza4q, Musannaf, vol. X, p. 320; and cf. ibid., vol. VI, pp. 5960.
See also ibid., vol. X, p. 319, vol. VI. p. 60: $/C - 2 8W$ 1 F+ "
^
5
83. Ibid., vol. VI, p.61.
84. See Tritton, The Caliphs, p. 46, citing Severus b. al-Muqaffa: Y. Abd
al-Mas3h and O. H. E. Burmester, eds., History of the Patriarchs of the
Egyptian Church by Sawrus ibn al-Muqaffa4, Cairo 1943. pp.181, 185.
85. Al-Tabar3, Tarkh, vol. III, p.712.
86. Chron. 1234, vol. II, pp.16, 11 (text).
87. Ibid., pp.212 (text), 15 (trans.).
88. The conflict over the public display of the cross by the Christians has already
been discussed in S. H. Griffith, Images, Islam and Christian Icons, in P.
Canivet and J. P. Rey-Coquais, eds., La Syrie de Byzance lIslam VIIe
VIIIe sicles: Actes du Colloque international Lyon Maison de lOrient
Mditerranen Paris Institut du Monde Arabe 1115 Septembre 1990,
Damascus 1992, pp.12138.
89. Palmer et al., Chronicles, p.167; Theophanes, Chronographia, vol. I, p.342;
Chabot, ed., Michel le Syrien, vol. IV, pp.421 (text), 431 (trans.).
90. Ibid.
91. Presumably the reference here is to crosses drawn upon walls.
92. Palmer et al., Chronicles, pp.16970.
93. Ibid.
94. Theophanes, Chronographia, vol. I, p.342.
95. Ibn Asa4kir, Tarkh, vol. LXV, p.156.
96. See chap.3.
97. See n. 65.
98. See Noth, Abgrenzungsprobleme, pp. 122 (German), 313, n. 99, para. 2
(English), which says only that Abu4 Yu4suf preserved many traditions parallel
Notes to pp. 7780
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
203
to the Shuru4t, yet he obviously was not familiar with its text; he therefore
comes to the unnecessary conclusion that the Shuru4t did not yet exist. As
said above, Abu4 Yu4suf might well have chosen to ignore Shuru4t Umar.
Only in Ibn Asa4kirs fifth version (Tarkh, vol. II, p. 120) did I find a slightly
different version: " >`_ - $ $ $F $a] 7 $ "W$ %> $ )(7.
See ibid., p. 174: 0F)%Fb >`_ " $F $ $- $237 /5 7 ;pp. 176,
177, 178: $" >`_ $F $ $ ( ()/()) )(7. For parallel versions see
Cohen, Pact of Umar, pp. 139, 141.
See above, p.71.
The prohibition on displaying the cross on top of the churches appears
in most of the versions of the Shuru4t: see e.g. Ibn Qayyim al-Jawziyya,
Ahka4m, p. 659; Ibn Asa4kir, Tarkh, in all five versions (see n. 16); Abu4
al-Maa4l3 al-Musharraf b. al-Murajja43, Fada4il bayt al-maqdis, pp.567; Abu4
Muhammad Abd Alla4h b. Ahmad b. Zabr, Juz fh shuru4tal-nasa4ra4, in
Cohen, Pact of Umar, pp.13157, at pp.137, 139. It is, however, missing
in a few of the versions:see Ibn Qayyim al-Jawziyya, Ahka4m, pp.6612;
al-Turtu4sh3s version, Sira4j pp.2523; two other versions in Ibn Zabrs text,
in Cohen, Pact of Umar, pp.141, 146.
See e.g. Ibn Asa4kir, Tarkh, vol. II, p.120; Ibn Qayyim al-Jawziya, Ahka4m,
p.659.
Regarding this, see chap.5.
Ibn Asa4kir, Tarkh, vol. II, pp.121, 175, 176, 177.
See also al-Sha4fi3, Kita4b al-umm, vol. IV, p.292, discussed below.
Ibid., pp.2814. Regarding the dating see n. 53.
Fattal, Le statut, pp.7781; Tritton, The Caliphs, pp.1216.
See Cohen, Pact of Umar, pp.11920. Cohen assumes the existence of
the Shuru4t at the time of al-Sha4fi3, accepting Trittons theory, followed by
Fattal, that the Umar in question is Umar b. Abd al-Az3z, or alternatively
Noths proposition that the document originated in the days of Umar b.
al-Khatta4b:see pp.1013.
See Muhy3-l-D3n Abu4 Zakariyya4 Yahya4 b. Sharaf b. Mur3 al-Nawaw3, Sharh
al-Nawaw ala4 Sahh Muslim, Beirut 1392 H [1972], vol. XV, p.100; Abu4
al-Fida4 Isma43l b. Kath3r, al-Sra al-nabawiyya, vol. I, Beirut 1976, p.199;
see also English translation in The Life of the Prophet Muhammad, trans.
Trevor Le Gassick, Reading 1998, vol. I, p.142.
E.g. in the matter of the application of Muslim law to dhimms (al-Sha4fi3,
Kita4b al-umm, vol. IV, pp.2989); the laws concerning the sale of pigs or
wine (ibid., pp.299300); the prohibition on entering the Hija4z, with the
exception of a three-day stay for commercial purposes (ibid., pp.2503);
the jizya (ibid., pp.247, 2878).
See e.g. the question of mixed marriages discussed by Abd al-Razza4q,
Musannaf, vol. VI, p.78, Sha4fi3, Kita4b al-umm, vol. V, p.84; the question
of a dhimm giving evidence, Abd al-Razza4q, Musannaf, vol. VI, p.129;
judgment of dhimms, ibid., vol. X, pp.3213; payment of taxes and jizya,
ibid., vol. X, p.328, 333; the diya of the dhimm, Sha4fi3, Kita4b al-umm, vol.
VI, p.136; the question of uhu4d, ibid., pp.191ff.
204
Notes to pp. 8084
113. Al-Sha4fi3, Kita4b al-umm, vol. IV, p.280.

114. Ibid.
115. See e.g. Theophanes, who reports that the tongue of Petrus, Metropolitan
of Damascus, was cut out because he publicly reproved the impiety of
the Arabs; he goes on to tell about his namesake Petrus of Maioumas,
who called Muhammad a false prophet and the precursor of the
Antichrist: Theophanes, Chronographia, pp. 41617; Theophanes,
Chronicle, p. 577; for the story of George, a monk from Mar Saba,
and his friends who defied Islam in Spain and were martyred see M.
Levy-Rubin and B. Z. Kedar, A Spanish Source on Mid-Ninth Century
Mar-Saba, in J. Patrich, ed., The Sabaite Heritage in the Orthodox
Church from the Fifth Century to the Present: Monastic life, Liturgy,
Theology, Literature, Art, Archaeology, Orientalia Lovaniensia Analecta
98, Leuven 2001, pp.6372; see there additional information concerning this phenomenon.
117. See e.g. Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, pp.659, 661.
118. See chap.1, nn. 23065.
119. See also Ibn Asa4kirs parallel version, Tarkh, vol. II, pp. 1234 (see
n.66).
120. Sha4fi3, Kita4b al-umm, vol. IV, p.281.
121. " $&e 21 7 $$3 f 7, Ibn Qayyim al-Jawziyya, Ahka4m, vol. II,
p. 661, ll. 1112; " $&e 021) 7 $$3 $ $ - 7 $ "W$ - f) 7, Ibn
Asa4kir, Tarkh, vol. II, p. 176, see also similarly pp. 120, 174, 177, 179;
Ibn al-Murajja4, Fada4il, p. 56.
122. See chap.1 (nn. 233, 237, 258).
123. Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, pp.659, 651.
125. Ibid., p.281.
126. Ibid., p.283.
127. Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, p.659, ll. 1213; Cohen, Pact of
Umar, p.141, l. 13.
128. See e.g. Ibn Qayyim al-Jawziyyas second version, Ahka4m, vol. II, pp.6612.
129. Al-Sha4fi3, Kita4b al-umm, vol. IV, p.282; trans. in Lewis, Islam, p.221.
130. Al-Sha4fi3, Kita4b al-umm, vol. IV, p.282; trans. in Lewis, Islam, p.221.
131. Sha4fi3, Kita4b al-umm, vol. IV, pp.2923.
132. See p.66, nn. 35, 36.
133. Abu4 Yu4suf, Kita4b al-khara4j, p.138:see n. 65.
134. Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, p.659, ll. 34, p.661, l. 10; Ibn
Asa4kir, Tarkh, vol. II, pp.174, 176, 178. The other two versions prohibit
the repair of churches only in Muslim neighbourhoods:see pp.120, 177;
Ibn al-Murajja4, Fada4il, p.56.
135. See Abu4 Yu4suf, Kita4b al-khara4j, pp.1278.
136. Thus, the absolute prohibition on exhibiting the cross (=9 $M %Fi) 7
$ "W$ )is followed by a prohibition to parade it on the streets of the Muslims
and in their markets (0F< " %j ]- 7), and in the versions of
the Shuru4t that mention pigs (many of them prohibit only the sale of wine)
Notes to pp. 8489
137.
138.
139.
140.
141.
142.
143.
144.
145.
205
the text reads: We will not keep pigs in their neighbourhoods (0$() 7
%$ k$). See Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, p. 659; Ibn Asa4kir,
Tarkh, vol. II, p. 120.
Sha4fi3, Kita4b al-umm, vol. IV, p.283; trans. in Lewis, Islam, p.222.
See e.g. Ibn Asa4kir, Tarkh, vol. II, pp. 1756: $C 01 S $ 9 02< $ 01)
$ "K)= 9 01 $ j%] $ 2 $ $ $ "K)C: When you came against us
we asked you for ama4n for ourselves, our descendants, our property, and
the people of our community, and we undertook the following obligation
towards you.
Cohen, Pact of Umar, pp.10330.
Ibn Asa4kir, Tarkh, vol. LXVII, p.235.
R. Gottheil, An Answer to the Dhimmis, Journal of the American Oriental
Society, 41 (1921), pp.383457, at p.390.
Ibn al-Wa4sit3 (see n. 141): " 3$" 0F"W$ 5 %
= > 9 ;Ibn
Asa4kir, Tarkh, vol. LXVII, p. 235, ll. 12: (" " 1 5' k> S A3
%Fj mi) $F)7 " ;see also vol. II, p. 181.
See e.g. the discussion concerning the authenticity of this tradition in
regard to Damascus in al-Bala4dhur3, Bulda4n, p.123, where the occupation
of former Christian houses is explained by the fact that they were vacated
by their inhabitants who had fled to Byzantium; another case is that of
Tiberias, in ibid., p.116; al-Tabar3, Tarkh, vol. I, p.2159; in Hims the
city was divided into lots, in each of which a Muslim house was built; in
addition a quarter of the church of St John was confiscated, as well as the
houses evacuated by Christians who had fled. See al-Bala4dhur3, Bulda4n,
p.130; see also Levy-Rubin, Changes.
Ibn Asa4kir, Tarkh, vol. II, p.180, ll. 49, p.181, ll. 1216.
See chap.4.
3. The Date and the Ideology of the Ghiya4r Code
1. For the meaning and origin of the term see chap.5; see Miller, Catalogue
to Codes, pp.1657.
2. In a separate chapter, rather than as part of the general conquest agreement. See below.
3. On these, see chap.2.
4. See Abu4 Yu4suf, Kita4b al-khara4j, pp.1278.
5. See e.g. G. R. Hawting, The First Dynasty of Islam, London, and New
York 2000, pp.1518, 7681. For avid support of Umar II as a historical devout and pious Muslim, a vehement critic of the Umayyads, and an
adamant follower of the Ra4shidu4n see Hasan Q. Murad, Was Umar II A
True Umayyad?, Islamic Studies, 24 (1985), pp.32548.
6. See P. M. Cobb, Umar b. Abd al-Az3z, EI2, vol. X, pp.8212, at p.822A;
on the prevalence of this topos see A. Borrut, Entre tradition et histoire:gense et diffusion de limage de Umar II, Mlanges de luniversit
Saint-Joseph, 58 (2005), pp.32978, at pp.32936.
7. See E. Landau-Tasseron, The Cyclical Reform:A Study of the Mujaddid
Tradition, Studia Islamica, 70 (1989), pp.79113, at pp.84, 11213.
206
Notes to pp. 8991
8. Hawting, Dynasty, pp.18, 77:While there is no doubt that the acceptance

of Umar as a genuine caliph (khalfa), unlike the other Umayyads who count
only as kings (muluk), is based to some extent on historical facts and on this
caliphs personality and actions, it is also clear that much of the traditional
writing about him should be regarded as pious and moralistic story-telling in
keeping with the needs and outlook of tradition (p.77).
9. Fattal, Le statut, pp.979.
10. Abu4 al-Qa4sim Abd al-Rahma4n b. Abd al-Hakam, Sira4t Umar b. Abd
al-Azz, Cairo 1994, p.160; Abu4 Yu4suf, Kita4b al-khara4j, pp.1278; Abd
al-Razza4q, Musannaf, vol. VI, p. 61, no. 10004; Abu4 Ubayd, Amwa4l, ed.
Ima4ra, p.130, para. 137 (=Amwa4l, p.53, para. 137); Ibn Asa4kir, Tarkh,
vol. II, pp. 17980, 185; for a short reference see Chabot, ed., Michel le
Syrien, vol. II, pp.4889.
11. Cohen, Pact of Umar, pp.140, 147.
12. Y. Kalfon-Stillman, Arab Dress:A Short History from the Dawn of Islam to
Modern Times, ed. N. A. Stillman, Leiden 2000, p.44.
13. Ibn Asa4kir and Ibn Zabr, adducing the same hadth, have also qalansuwa
(Ibn Zabr, Juz, p.140; Ibn Asa4kir, Tarkh, vol. II, p.179), while Abu4 Yu4suf
also includes a prohibition on silk clothes (Kita4b al-khara4j, p.127).
14. Abu4 Yu4suf, Kita4b al-khara4j, p.127; the same version exactly is also cited by
Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, p.743.
15. Ibn Abd al-Hakam, Sira4t Umar, p.160.
16. Ibid.
17. Cohen, Pact of Umar, p.140.
18. Ibid., p.147.
19. Quoting exactly the same hadth and isna4d on pp. 17980 and another
isna4d, going back to al-Hakam b. Umar al-Ruayn3, on p.185, which does
not carry the same suffix mentioned here below.
20. Although the term mafru4q could also be interpreted as someone whose hair
is parted, this is not possible in this case, as it is well established that the
demand was to cut the forelocks and not to part them. This is said in so many
words in the second part of the edict cited by Abu4 Yu4suf, complaining that
the non-Muslims had abandoned the girdles, had started wearing their hair
thick and long, and had abandoned the trimming (of the forelocks): %+
n/52 %+ %- $( k+ 0Fj$ 9 !j$ ( Abu4 Yu4suf, Kita4b al-khara4j,
p. 128), and in Abd al-Razza4qs version: 0FM (' )0F< $K F
and they should be prohibited from parting their hair and should cut their
forelocks (Abd al-Razza4q, Musannaf, vol. VI, p. 61, no. 10004).
21. Contrary to Abu4 Yu4suf himself, who states at the beginning of the chapter
that the straps of their sandals must be doubled: Abu4 Yu4suf, Kita4b al-khara4j,
p. 127 ( ^ 0F$A) %] A( $).
22. The first version of Ibn Zabr and Ibn Asa4kir both omit the qaba4 and the
taylasa4n and add the qalansuwa.
23. Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, pp.7412.
24. See pp.1537 (nn. 283308).
25. See n. 64, 65; see also chap.5, nn. 22348.
Notes to pp. 9192
207
26. Abu4 Yu4suf, Kita4b al-khara4j, p.127; Abd al-Razza4q, Musannaf, vol. VI, p.61,
no. 10004.
27. Ibn Zabr, Juz, p. 146; Ibn Asa4kir, Tarkh, vol. II, p. 180: $- $' P $
''A 9 %9 9 % $F+3 $`k %9 F9 $ <; .
28. See Chabot, ed., Michel le Syrien, vol. IV, p. 456 (text), vol. II, pp. 4889
(trans.); Chron. 1234, vol. I, pp.3078 (text), vol. I, pp.23940 (trans.); Bar
Hebraeus, Chronicon Syriacum, ed. A. Fiat, Paris 1890, p.117.
29. See Borrut, Tradition et histoire, pp.34952:Hoyland, trans., Theophilus.
30. Chron. 1234, vol. I, p.307 (text).
31. Such as the imposition of the head tax, the prohibition on Christians serving
as witnesses, the regulation that the penalty for the murder of a Christian
should not be death but rather a compensation fee, and the annulment of the
tithes to churches and monasteries. See Chabot, ed., Michel le Syrien, vol.
IV, p.456 (text), vol. II, pp.4889 (trans.);, Chron. 1234, vol. I, pp.3078
(text), vol. I, pp.23940 (trans.).
32. Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, p. 742; Abu4 al-Abba4s Taq3
al-D3n Ahmad b. Abd al-Hal3m b. Taymiyya, Iqtida4 al-sira4t al-mustaqm
li-mukha4lafat asha4b al-jahm, Cairo 1369 H, vol. I, pp.1234. Both of the
isna4ds go back to Muhammad b. Qays, who was Umar b. Abd al-Az3zs
qa4ss (or qa4d):see Ibn Asa4kir, Tarkh, vol. LV, pp.10814; and to Sa3d b.
Abd al-Rahman b. Hasan (both seem to carry the same mistake: Haba4n
instead of Hasa4n) b. Tha4bit the poet, who served as a deputy of Yaz3d b.
Abd al-Malik and Hisha4m b. Abd al-Malik. See Ibn Asa4kir, Tarkh, vol.
XXI, pp.17783. A similar tradition is noted by Fattal, Le statut, p.98 and
n. 55, citing al-Turtu4sh3, Ibn Naqqa4sh, and al-Ibsh3h3. Fattal notes that a
famous scholar of the twelfth century, al-Ka4sa4n3, retelling the episode of the
Banu4 Taghlib, also claimed that Umar b. Abd al-Az3z was the initiator of the
ghiya4r:see Le statut, p.98, n. 56.
33. For the various versions of this tradition see Shahid, Fourth Century,
pp.42232. A similar story is attributed to al-Mahd3:see Chabot, ed., Michel
le Syrien, vol. IV, pp.4789 (text), vol. III, p.1 (trans.).
34. Hisha4m al-Kalb3, an early source (737819 or 821 CE), reports that the
Tanu4khids came out to meet al-Mahd3 on horseback, wearing ama4im
(turbans) on their heads. When he found out that they were Christian he
asked them to convert, but they refused. He then beheaded their chief, after
which they converted to Islam. Shahid, Fourth Century, p.431, and n. 61
citing al-Kalb3 from a manuscript of Ya4qu4ts al-Muqtada, who epitomizes
al-Kalb3s Jamhara.
35. See Hoyland, Seeing Islam, pp.3524, and see there, on p.352, n. 60, the
citation of Kha4lid from Azd3s Futu4h, saying:We do not allow Arabs to be in
a religion other than ours; see also ibid., pp.41516, the Ehnesh Inscription,
which mentions that the caliph al-Mahd3 ordered the Tanu4kh to become
Muslims. This is indeed sixty years later than Umar II, but may well signify
the deterioration in the relationship between Christian and Muslim Arabs.
36. On the issue of the dress code see chap.5.
37. Ibn Abd al-Hakam, Srat Umar.
208
Notes to pp. 9395
38. Ibid., pp.278.

39. Borrut, Tradition et histoire, pp.35964.
40. H. A. R. Gibb, The Fiscal Rescript of Umar II, Arabica, 2 (1955), pp.116,
at pp.12.
41. Ibn Abd al-Hakam, Srat Umar, p.87.
42. Ibid., p.89.
43. Ibid., p.90.
44. Ibid., pp.925.
45. Ibid., p.94.
46. Ibid., pp.1612.
47. E.g. his correspondence with Umar b. al-Wal3d (ibid., pp. 1503; Safwat,
Jamhara, vol. II, pp.36972 (nos. 41819); his letter to Sa4lim b. Abdalla4h
(Ibn Abd al-Hakam, Srat Umar, p.125; Safwat, Jamhara, vol. II, pp.3758);
his letter to Abu4 Bakr b. Muhammad b. Amru b. Hazm (Ibn Abd al-Hakam,
Srat Umar, p.72; Safwat, Jamhara, vol. II, pp.32839).
48. See M. Cook, Early Muslim Dogma, Cambridge 1981, pp.12436.
49. Ibid., p.7.
50. Ibid., p.165, chap.2, n.1, and pp.1012.
51. See ibid., p. 7; P. Crone and F. Zimmermann, The Epistle of Sa4lim b.
Dhawka4n, Oxford 2001, p.16 and p.148, n. 14.
52. Crone and Zimmermann, Epistle, p.15.
53. Cook, Dogma, pp.1620.
54. See Q 3:11012, 3:139, 2:61, 17:111, 42:45, 47:35. On this subject see H.
Ben-Shammai, The Idea of Election in Early Islam, in S. Almog and M.
Heyd, eds., Chosen People, Elect Nation and Universal Mission, Jerusalem
1991, pp. 14777 (Hebrew); Friedmann, Tolerance, pp. 349, esp. p. 35,
and n. 115, citing Muhammad b. Isma43l al-Bukha4r3, Kita4b al-ja4mial-sahh,
ed. M. L. Krehl and T. W. Juynboll, Leiden 18621908, vol. I, ed. L. Krehl,
Leiden 1864, pp.33940):Islam is exalted and nothing is exalted above it
(al-Isla4m yalu4 wa la4 yula4).
a 01 019 $ 019
55. Ibn Abd al-Hakam, Srat Umar, p. 79: $ <

%5
$ 0 9 %56< %- ] ] < p. 85: () $ A- %7 E
)( ' $)'9 F )( ' $ % AX ' .
56. Ibid., p.82.
57. Trans. in A. J. Arberry, The Qura4n Interpreted, London and New York
1955, vol. II, p.53.
58. Ibn Abd al-Hakam, Srat Umar, p.84.
59. $`& 3 0FA3 6 8() %& , ibid., p. 159.
60. On the use of mushriku4n for non-Muslims in general as a polemical term,
and reflecting, in fact, the original Qura4nic tone, see G. R. Hawting, Idolatry
and Idolators, Encyclopaedia of the Qura4n, Leiden and Boston 2002,
pp.47584, at pp.477, 479.
61. See a hadth expressing the same message and attributed to Umar in Shuru4t
al-nasa4ra from the tenth or eleventh century CE, in Cohen, Pact of Umar,
p. 148: 7 f9 $< $2 $ % $ $< E L21 $K 6 $+$ k+ ,-
0$/< 0$)+ 7 0F)$ 0F%1+ 7 0F)> 0F $+. This tradition is repeated
in later literature; see al-Qurtub3, Tafsr al-Qurtub, Cairo 1372 H, vol. IV,
Notes to pp. 95100
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
209
p. 179; Abdalla4h b. Ahmad b. Quda4ma al-Maqdis3, al-Mughn, Beirut 1405

H, vol. I, p. 270, vol. VI, p. 326.
Ibn Abd al-Hakam, Srat Umar, p.160.
See n. 20; M. Bravmann, The Ancient Arab Background of the Qura4nic
Concept al-Jizyatu an Yadin, Arabica, 13 (1966), pp.30714.
The qaba4 was a luxurious sleeved robe slit in front with buttons (muzarrar) made of fabric such as brocade (dba4j), and apparently of Persian provenance: see Kalfon-Stillman, Arab Dress, p. 12, and n. 17; see also Dozy,
Supplemnt, pp.35262.
Here literally a belt made of leathers, perhaps meaning that the belt could
be made of various kinds of leather:Ibn Asa4kir, Tarkh, vol. II, p.180 has
zunna4r min jild in the singular.
See chap.5 (nn. 16586, esp.n. 186).
See introduction, n. 13.
Regarding the epithet fa4ru4q, see G. Levi Della Vida, review of vol. V of L.
Caetanis Annali, in Rivista degli studi orientali, 4 (1911), pp.10746; G.
Levi Della Vida (M. Bonner), Umar (I) b. al-Khatta4b, EI2, vol. X, p.820;
P. Crone and M. Hinds, Gods Caliph, Cambridge 1986, pp.11314; Crone
and Cook, Hagarism, pp. 56; W. W. Barthold, Caliph Umar II and the
Conflicting Reports on his Personality, Islamic Quarterly, 15 (1971),
pp.6995, at pp.735.
See Tabar3, Tarkh, vol. III, pp. 13923; J. L. Kramer, annot and trans.,
The History of al-Tabar, vol. XXXIV: Incipient Decline, Albany 1989,
pp.923.
See Muhammad b. Jar3r al-Tabar3, Tafsr al-Qura4n, Beirut 1405H [1984],
vol. IV, p. 51; Ibn Kath3r, Tafsr Ibn Kathr, Beirut 1401H [1980], vol. II,
p. 350; al-Qurtub3, Tafsr, vol. I, p. 430; al-N3sa4bu4r3, al-Mustadaik ala
al-sahhayn, Beirut 1990, vol. IV, p.477; Ibn Ab3 Shayba, Musannaf, vol. IV,
p.216, no. 19437.
Fattal, Le statut, pp.989; Bar Hebraeus, Chronicon, p.117.
P. Crone, The Pay of Client Soldiers in the Umayyad Period, Der Islam, 80
(2003), pp.284300; I. Hasson, Les mawa4l dans larme musulmane sous
les premiers umayyads, Jerusalem Studies in Arabic and Islam, 14 (1991),
pp.176213.
See chap.5 (nn. 13649).
H. A. R. Gibb, ArabByzantine Relations under the Umayyad Caliphate,
Dumbarton Oaks Papers, 12 (1958), pp.21933, at p.233.
For the letter, see chap.5.
Al-Sha4fi3, Kita4b al-umm, vol. IV, p.282; trans. in Lewis, Islam, p.221.
4. The Enforcement of Shuru4t Umar

2. S. D. Goitein, A Mediterranean Society: The Jewish Communities of the
Arab World as Portrayed in the Documents of the Cairo Geniza, vol. II:The
Community, Berkeley, 1971, pp.2878.
3. Gil, History, p.159.
210
Notes to pp. 100102
4. This is also in concert with the opinion of Miller, who also noted the connection between the two, yet presumes that it was al-Mutawakkils decrees that
brought about the rise in popularity of the Shurut4 :see
Catalogue to Codes,
pp.2045. I would prefer to say that most probably al-Mutawakkils restrictions were based on an existing document, and that there is actually no reason
to doubt that this document was a version of Shurut4 Umar. Miller himself
admits that a broad array of isnad
4 s for the Standard/Cairo Petition in the
preserved texts diverge from Rab3 b. Thalab. These isnad
4 s demonstrate a
groundswell of support for the Petition to Umar among lesser known third
century had
th
transmitters from the Hija
z4 through Iraq
4 (ibid., p.204).
5. See Palmer et al., Chronicles, pp.16970. The conflict over the public display
of the cross by Christians is discussed in Griffith, Images.
6. See Dionysius of Tell-Mah re
4 in Palmer et al., Chronicles, text no. 12, AG
1015, p.78; see ibid., a reference to Chabot, ed., Michel le Syrien, vol. IV,
p.447, ll. 1720, who reports in the same context about an order to remove
all crosses.
7. See n. 25.
8. D. C. Dennett, Conversion and the Poll Tax in Early Islam, Cambridge, MA,
1950, pp.7884.
9. Abu4 Yu4suf, Kita4b al-khara4j, p. 127; see also Abd al-Razza4q, Musannaf,
vol.VI, p.61, no. 10004.
10. Chabot, ed., Michel le Syrien, vol. II, p. 489; Abd al-Razza4q, Musannaf,
vol.VI, p.61, no. 10004.
11. Chabot, ed., Michel le Syrien, vol. II, p.489.
12. See chap.2, nn. 308, and nn. 845; Fattal, Le statut, pp.1856; al-Turtu4sh3,
Sira4j, p.138; al-Tabar3, Tarkh, vol. II, p.1371.
13. See Abd al-Razza4q, Musannaf, vol. VI, pp.1920; and cf. vol. X, pp.31920.
14. Tritton, The Caliphs, pp.423.
15. See ibid., p.102 adducing evidence in favour, and Michael the Syrian (vol. II,
p.489), who claims that Umar forbade tithes and bequests to churches.
16. See Ibn Asa4kir, Tarkh, vol. XXXIV, pp.23659; Jama4l al-D3n Abu4-l-Haja4j
Yu4suf al-Mizz3, Tahdhb al-kama4l f asma4 al-rija4l, Beirut 1992, vol. XVII,
pp.1218.
17. Abu4 Yu4suf, Kita4b al-khara4j, pp.1278.
18. The usual term is zunna4r, the term used indeed by Abu4 Yu4suf himself in the
previous paragraph describing the prohibitions of his day. The use of the
term mintaqa, which was later kept for military and honorary belts, may
point to an early date for this tradition, before a clear distinction was made
between these two terms. See chap.5.
19. See ibid., pp.1278.
20. Theophanes, Chronographia, p.431; Theophanes, Chronicle, p.596.
22. Theophanes, Chronographia, p. 446; Theophanes, Chronicle, p. 616;
see also Chron. Zuqnin (Incerti Auctoris chronicon anonymum pseudoDionysianum vulgo dictum), vol. II, ed. J. B. Chabot, Louvain 1933, pp.104
105; trans. R. Hespel as Chron. Zuqnin (Incerti Auctoris chronicon anonymum pseudo-Dionysianum vulgo dictum), Louvain 1989, pp.1234.
Notes to pp. 102104
211

24. See n. 5.
25. See al-Bala4dhur3, Bulda4n, p. 193; al-Jahshiya4r3, Kita4b al-wuzara4 wa-alkutta4b, Cairo 1938, pp.3840; both these sources make a clear connection
between the ousting of prominent non-Muslims from the administration
and its Arabization, although of course practically this goal was much
harder to achieve; see A. A. Duri, Diwa4n, EI2, vol. II, p. 324; Hawting,
Dynasty, pp.635. There are, however, many traditions relating how Umar
b. al-Khatta4b adamantly prohibited the employment of non-Muslims in
government offices; even if this were indeed true, it was definitely unfeasible
at the time:for a survey on this issue see Fattal, Le statut, pp.24063.
26. See above.
27. Theophanes, Chronographia, p. 430; Theophanes, Chronicle, p. 594. The
name is miswritten in the text as Salim. See A. Grohmann (and H. Kennedy),
S a4lih b. Al3, EI2, vol. VIII, p.985.
28. Al-Tabar3, Tarkh, vol. III, p.712.
29. See ibid.; Abu4 Zakar3ya4 Yaz3d b. Muhammad al-Azd3, Tarkh al-Mawsil, ed.
A. Hab3ba, Cairo 1967 p.311; Tritton, The Caliphs, pp.11718; Fattal, Le
statut, pp.1001.
30. See M. Levy-Rubin, Continuatio of the Samaritan Chronicle of Abu4 l-Fath
al-Sa4mir al-Danaf, text and trans., Studies in Late Antiquity and Early
Islam, Princeton 2002, p.90.
31. For the full description and a citation of the edict itself see al-Tabar3, Tarkh,
vol. III, pp.138994; Kramer, Decline, pp.8994.
32. Abd al-Rahma4n b. Al3 b. Muhammad b. al-Jawz3, al-Muntazam f tarkh
al-mulu4k wa-al-umam, vol. XIII, Beirut and London 1992, p. 82; Abu4
al-Maha4sin Jama4l al-D3n Yu4suf b. Taghr3 Birdi, al-Nuju4m al-za4hira f mulu4k
misr wa-al-qa4hira, ed. Gautier H. A. Juynboll, Leiden 1861, pp.1745; Ibn
Naqqa4sh, Fetoua sur la condition des dhimms, ed. and trans. M. Belin,
Journal Asiatique, series 4, 18 (1851), pp.417516, 19 (1852), pp.97140.,
at 18 (1851), p.455 reports that al-Muqtadir forbade the employment of
dhimms in the government, gave an order to dismiss all dhimms already
employed, and went as far as executing one such dhimm employed by the
ha4jib Yu4nis; Fattal, Le statut, p. 103, reports that Muhammad al-Ikhsh3d
enforced a similar set of rules, but does not cite the actual source.
33. Ahmad b. Al3 al-Maqr3z3, Ittia4z al-hunafa4 bi-akhba4r al-umu4r al-fa4timiyn
al-khulafa4, Cairo 1967, vol. I, p. 132; Tritton, The Caliphs, pp. 54, 120,
130, and passim.
34. On the restrictions of al-Ha4kim see Y. Lev, Persecutions and Conversion
to Islam in Eleventh Century Egypt, Asian and African Studies, 22 (1988),
pp.7391, at pp.8083.
35. Al-Tabar3, Tarkh, vol. III, p.1389; Kramer, Decline, pp.8990.
36. Al-Tabar3, Tarkh, vol. III, pp.1392, Kramer, Decline, p.93.
37. Al-Tabar3, Tarkh, vol. III, p.1419; Kramer, Decline, p.128.
38. Ibn al-Jawz3, al-Muntazam f tarkh al-muluk wa-al-umam, vol. XI, Beirut
and London 1992, pp.2223. I would like to thank Prof. Y. Lev for kindly
sharing with me these references to Ibn al-Jawz3s Muntazam.
212
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
Notes to pp. 105108

Ibid., p.238.
Ibid., p.265.
Ibid., p.270.
Severus b. al-Muqaffa, History, vol. II, part 1, p.6.
Bar Hebraeus, The Chronography of Abu4l Faraj known as Bar-Hebraeus,
trans. E. A. W. Budge, Oxford 1932, vol. I, p.141.
Levy-Rubin, Continuatio.
Ibid., pp.913.
Al-Mutawakkils decrees are described in detail in al-Tabar3, Tarkh, vol.III,
pp.138994. See also translation, notes and bibliography in Kramer, Decline,
pp.8994.
See N. Rabbat, Tira4z, EI2, vol. X, pp.534538.
Undoubtedly during grand occasions and ceremonies, when these were
reserved for those of honourable rank. For this sense of martaba see J. Sadan,
Le mobilier au proche-orient medieval, Leiden 1976, pp.545.
See al-Tabar3, Tarkh, vol. III, p.1389.
Cf. ibid., p.1390:In addition, he ordered that their graves be made level
with the ground so as not to resemble the graves of the Muslims (Kramer,
Decline, p.91).
The change in font signifies words written in Samaritan script.
See al-Tabar3, Tarkh, vol. III, p.1390, who calls these idols suwar shaya4tn
(images of devils). The reason given for this is cited by Strauss (Ashtor),
Social Isolation, p.80:If the houses of the Dhimms will not bear distinctive signs the beggars approaching them will beg Alla4hs mercy upon
them and his forgiveness for their sins, and this say the theologians is
absolutely forbidden. This was surely a restriction that was meant to be
derogatory, since both shayta4n and wathan are definitely negative terms.
A Samaritan settlement near Ramla. It is miscopied here; later on, the correct name appears: see Levy-Rubin, Continuatio, Arabic text, p. 166 (MS
p.248); see also the Geographical Appendix.
This word, written here quite distinctly in Samaritan script, is transcribed in
Arabic as fanqa. I could not find any meaning for it in Arabic or in other
Semitic languages.
The word used here is goy in Hebrew literally a man belonging to a
different faith, a non-Jewish person.
Al-Tabar3, Tarkh, vol. III, p.1419.
See Ibn Asa4kir, Tarkh, vol. II, pp. 175, 176, 177, 179; Ibn Qayyim
al-Jawziyya, Ahka4m, vol. II, pp.659, 6612; Shuru4t al-nasa4ra4 in Cohen, Pact
of Umar, pp.1378, 139, 142, 146.
Al-Tabar3, Tarkh, vol. III, p.1393; Cohen, Pact of Umar, p.151.
Al-Tabar3, Tarkh, vol. III, p.1393; Kramer, Decline, pp.934.
See end of n. 32.
Ar3b b. Sad al-Qurtub3, Silat tarkh al-Tabar, ed. M. J. de Goeje, Leiden
1965, p.30.
Ibid.
Notes to pp. 108110
213
64. See Ibn Qayyim al-Jawziyya, Ahka4m, vol. I, pp.21242; for Ibn al-Naqqa4sh
see Fetoua, esp.18 (1851), pp.44055, 19 (1852), pp.97140.
65. See Levy-Rubin, Continuatio, pp.1034; Arabic text, p.169 (MS p.251).
66. The word used here is nabdh; this is an alcoholic drink usually made of
raisins or dates. It could, however, also mean wine expressed from grapes
(Lane, Lexicon, p.2757, s.v. nabdh).
67. Chabot, ed., Michel le Syrien, vol. II, p.489; Chron. 1234, vol. I, pp.3078
(text), 239 (trans.); Agapius b. Mahbu4b, Kita4b al-unwa4n, ed. and trans.
Alexandre Vasiliev, Patrologia Orientalis 8/3, Turnhout 1971, pp.5023.
69. Sha4fi3, Kita4b al-umm, vol. IV, pp.281, 283.
70. See e.g. Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, pp.659, 6612.
71. See M. Levy-Rubin, Ibn Tu4lu4n and his Attitude towards the Protected People
According to a Contemporary Samaritan Chronicle, in Y. Ben Aryeh et al.,
eds., Chapters in the History of Eretz Israel and its Settlement:Researches
Offered to Y. Ben Porat, Jerusalem 2003, pp. 34560 (Hebrew); E. K.
Corbet, The Life and Works of Ahmad ibn Tu4lu4n, Journal of the Royal
Asiatic Society (1891), pp. 527562; S. Lane-Poole, A History of Egypt
in the Middle Ages, London 1914, pp. 701; Z. M. Hassan, Ahmad ibn
Tu4lu4n, EI2, vol. I, pp.278279; H. Kennedy, The Prophet and the Age of the
Caliphates:The Islamic Near-East from the Sixth to the Eleventh Century,
London 1986, pp.31013; J. L. Bachrach, Palestine in the Policies of Tulunid
and Ikhshidid Governors of Egypt, AH 254358/868969, in A. Cohen and
G. Baer, eds., Egypt and Palestine:A Millennium of Association 8681948,
Jerusalem 1984, pp.5165; B. Lewis, Egypt and Syria, in The Cambridge
History of Islam, vol. I:P. M. Holt et al., eds., The Central Islamic Lands,
Cambridge 1970, pp.175230.
72. This is well exemplified in Goiteins statement that in Egypt too, we find
between 1130 and the end of the Fatimid rule in 1171 occasional attempts
to renew the restrictions about the clothing of the non-Muslims. These
attempts, however, recorded by the Arabic historians, were ephemeral,
andare not echoed in the Geniza papers:Goitein, Mediterranean Society,
vol. II, p.288.
73. Levy-Rubin, Continuatio, pp. 1026, Arabic text, pp. 16971 (MS
pp.2513).
74. See n. 33.
76. Regarding the enforcement of ghiya4r regulations under his rule, see A.
Scheiber, A New Fragment of the Life of Obadjah, the Norman Proselyte,
Kirjath Sepher, 30 (19545), pp. 938; Izz al-D3n Al3 b. Muhammad b.
al-Ath3r, al-Ka4mil f al-tarkh, Beirut 1965, p.186.
77. Goitein, Mediterranean Society, vol. II, p.287; Fattal, Le statut, p.104.
78. Tritton, The Caliphs, pp.11526; Fattal, Le statut, pp.96110.
79. See Elias of Nisibis, Eliae Metropolitae Nisibeni, ed. E. W. Brooks, Louvain
1954, vol. I, p.188 (text). Elias attributes this information to al-Tabar3, who
does indeed mention the riot and the looting and destruction that followed
214
80.
81.
82.
83.
84.
85.
86.
87.
Notes to pp. 110115

it, but does not mention the cause of it all, according to Elias i.e. riding on
horses. See al-Tabar3, Tarkh, vol. III, pp.21078. Al-Tabar3 does take the
trouble to mention that the government troops fought againt the mob and
that several days after the riot everything the mob had destroyed was rebuilt.
Shams al-D3n Abu4 Abdalla4h al-Muqaddas3, Ahsan al-taqa4sm f marifat
al-aqa4lm, ed. M. J. de Goeje, Leiden 1906 p. 429; trans. B. A. Collins as
The Best Divisions for Knowledge of the Regions:A Translation of Ahsan
al-taqasim fi marifat al-aqalim, Reading 1994, p.380.
E.g. the restrictions of 1121 in Goitein, Mediterranean Society, vol. II, p.287;
Scheiber, Obadjah, the Proselyte, p.98.
Fattal, Le statut, p.103.
Ibid.
See above and n. 72 there.
See Tritton, The Caliphs, and Fattal, Le statut, passim.
On al-Ha4kim see above, p. 104, n. 34 there; regarding Sala4h al-D3n see
Severus b.al-Muqaffa4, History, vol. III part 2, pp.978, trans. p.1646.
Regarding Spain see Fattal, Le statut, p.105, and further references there.
5. The Provenance of the Modes of Subordination of Non-Muslims
1. Regarding exceptional cases where additional stipulations were made see

chap.2.
2. See previous chapter.
3. Noth, Abgrenzungsprobleme, trans. Muelhaeusler, p.103 (English) where
he says:The latter group [i.e. the Orientalists:MLR] have interpreted the
Shuru4t as a testimony to increasing discrimination experienced by religious
minorities within the Islamic world an interpretation that appears to be
confirmed by Islamic interpretations and the history of the implementation
of these ordinances.
4. On this see below, The Concept of Ghiya4r.
5. Although not a minority, there is no doubt that non-Muslims in the early
Islamic empire did not belong to the ruling classes and may therefore be
considered closest in their social status to that of minorities.
6. R. Westbrook, ed., A History of Ancient Near Eastern Law, Leiden and
Boston 2003, pp.3742, 377.
7. Although manumitted slaves in Rome were awarded citizenship, they were
denied certain rights such as access to public office and to the legions. Their
voting rights were also limited. See J. Heinrichs, Freedmen (Rome), in
Cancik and Schneider, eds., Brills New Pauly.
8. W. Heffening, Das islamische Fremdenrecht bis zu den islamischefrnkischen Staatsvertrgen, Hanover 1925, repr. Osnabrck 1975, pp.8891;
Tyan, Institutions, p.426.
9. These last two terms are sometimes used interchangeably, although mawla4
in Islam is usually applied to non-Arab clients who attached themselves
to Arabs upon their conversion to Islam, and is often considered a lower
status than the half, who is an ally. Landau-Tasseron, Alliances in Islam,
pp.245.
Notes to pp. 115116
215
10. Westbrook, History, p.38; see there the case of Abraham, in Gen. 23, who
upon the purchase of the Cave of the Machpelah in order to bury Sarah,
was required to receive special permission from the local Hittite leaders;
anothercase is that of the king of Ugarit who gave a special grant of land to
an Egyptian.
11. Ibid.
12. Ibid., p.42.
13. See Cancik and Schneider, eds., Brills New Pauly, vol. VIII, Leiden and
Boston 2002, cols. 81014.
14. Phillipson, Law and Custom, vol. I, pp. 16576; Cancik and Schneider,
eds., Brills New Pauly, vol. VIII, cols. 81014; see M. Clerc, Les mtques
athniens:tude sur la condition lgale, la situation morale et le rle social
et conomique des trangers domicilis Athnes, Paris1893; repr. 1979;
D. Whitehead, The Ideology of the Athenian Metic, Cambridge 1977;
P.Gautier, Mtques, priques et paroiki:bilans et points dinterrogations,
in R. Lonis, ed., Ltranger dans le monde grec:Actes du colloque organis
par lInstitut dtudes Anciennes Nancy, Mai 1987, Nancy 1988, pp.2346;
M. I. Finley, Between Slavery and Freedom, Comparative Studies in Society
and History, 6 (1964), pp.23349.
15. See Phillipson, Law and Custom, pp.177, 21034; B. Nicholas, Ius Gentium,
p. 790 and T. C. Brennan, Praetor, p. 1240, in N. G. L. Hammond and
H.H. Scullard, eds., Oxford Classical Dictionary, 3rd edn, Oxford 1996.
16. Westbrook, History, p.38.
17. See Phillipson, Law and Custom, pp.193209; R. S. Bagnall and P. Derow,
Greek Historical Documents:The Hellenistic Period, Ann Arbor 1981, p.84.
18. See chap.1, e.g. the case when Pharos was conquered by the Romans:the
people of Pharos were given their polis, their ancestral customs (patrioi
nomoi), and their land. See Eckstein, Pharos, p.398.
19. Josephus mentions that the Jews in Egypt lived according to their own
ancestral customs in Egypt, and had their own Ethnarch, who had complete authority over the community:M. Stern, Greek and Latin Authors
on Jews and Judaism, Jerusalem 1974, vol. I:From Herodotus to Plutarch,
p.278, clauses 11617. When Antiochus III took over Jerusalem he undertook, according to Josephus, to allow the Jews to live according to their
ancestral customs kata tous patrious nomous. See Josephus, Antiq. Iud.,
12.142.
20. See A. Linder, The Jews in Roman Imperial Legislation, Detroit and Jerusalem
1987, pp.645.
21. Ibid., pp.679.
22. See ibid., p. 76, for example, who notes that the political capacity of the
Jews in the municipal sphere evolved in three stages:collective exemption in
the early fourth century; equalization of their status to that of the rest of the
population in the fourth and fifth centuries; and legal discrimination from
the beginning of the sixth century.
23. The existence of this formal status awarded to the Jews is debated among
scholars. See T. Rajak, Was There a Roman Charter for the Jews?, Journal
of Roman Studies, 74 (1984), pp. 10723; L. V. Rutgers, Roman Policy
216
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
Notes to pp. 116118

towards the Jews:Expulsions from the City of Rome during the First Century
CE, Classical Antiquity, 13 (1994), pp.5674.
Linder, Jews, p.178.
Ibid., p.86, no. 18, pp.17882.
The law as conveyed by Ulpian 5, 35 says:Conubium est uxoris iure ducendae facultas. Conubium habent cives Romani cum civibus Romanis: cum
Latinis autem et peregrinis ita, si concessum sit. Cum servis nullum est conubium (Conubium is the legal ability to marry a woman. Roman citizens
have the conubium to marry each other but, only by special dispensation,
to marry Latins and other foreigners. There is no conubium with slaves); see
G. Schiemann, Conubium, in Cancik and Schneider, eds., Brills New Pauly,
Antiquity vols.
Valentian I (r. 36475 CE) issued an edict prohibiting marriages between
inhabitants of the provinces and pagan barbarians. This is an application of
the classical conubium law rather than a religion-based law.
Linder, Jews, pp.269, 297, 4001.
Ibid., p.288.
Ibid., pp.327 (text), 330 (trans.).
Ibid., pp.845.
Ibid., p.77.
Ibid., pp.269, 325, 327, 330, 332, 333.
Ibid., pp.1247, 13842.
A law promulgated by Justinian in July 527 CE regarding heretics, Jews, and
Samaritans states the following:
We order, therefore, that none of the above-mentioned shall share in any
honour whatsoever, nor shall he put on an official belt (zone), neither civil
nor military, nor belong to any office Indeed we order that those who are
heretics, and above all pagans, Jews, Samaritans if they take part in any
of all those we have already recalled having obtained an honour, inscribed
in the advocates list, taken an office or put on an official belt, they shall
be thrown out on the spot from participating in these. For we want all the
above mentioned to be purged from association with such as these now and
for ever, not only in this glorious city, but in practically every province and
every place. There is absolutely nothing new in this, for at least the Divine
belt-documents given for most of the offices contain the specification that the
recipient should be Orthodox. Unless this shall be seen to be ours, which we
have restored and did not neglect as it was before, when it was unheeded by
some and barely survived save in documents alone.
See Linder, Jews, pp.756. Regarding the official belts, see pp. 1536.
Linder, Jews, pp.66, 78, 3715, 3938.
Ibid., pp.31319, 3689.
Fattal, Le statut, pp.1389; Justinian, Nov. 115:see Iustiniani Novellae, ed.
Z. A. Lingenthal, Leipzig 1881, part II, no. CXXXVI, chap.3, para. 14.
Linder, Jews, pp.237, 2845.
Ibid., pp.1967.
See ibid., pp. 3202; On the cessation of the Patriarchate see H. Shapira,
The Deposition of Rabban Gamaliel:Between History and Legend, Zion,
64 (1999), pp.538 (Hebrew), with comprehensive bibliography.
Notes to pp. 118120

42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
217
Linder, Jews, pp.18990, 1978, 264.

Ibid., p.264.
Ibid., pp.199200.
Ibid., pp.1945.
On the restrictions of civil rights of Jews and the deterioration of their status
as citizens of the empire see ibid., pp.656.
Ibid., pp.2845, no. 46.
See ibid., pp.3578 (text), 3601 (trans.).
On the Iranian class system see pp. 13741.
Perikhanian, Iranian Society, pp.6345.
For an English translation see A. Perikhanian, The Book of a Thousand
Judgments, New York 1980; M. Macuch, Das sasanidische Rechtsbuch
Ma4takda4n i haza4r Da4tista4n (Teil II), Abhandlungen fr die Kunde des
Morgenlandes XLV, 1 Wiesbaden 1981; M. Macuch, Ma4daya4n haza4r
da4desta4n (Book of a Thousand Judgments), Pahlavi Law-Book from the
Late Sasanian Period (First Half of the Seventh Century), in Encyclopedia
Iranica, online.
Middle Persian documents alternately use the terms a4za4t; hamna4f (lit.
agnate); a4dehk (lit. fellow countryman); and mart-i-shahr (citizen). See
Perikhanian, Iranian Society, pp.6345, and additional bibliography there.
Perikhanian, Iranian Society, pp.6345.
See Linder, Jews, nos. 11, 44, 59, and esp.61; for a summary of Roman legislation on this issue see pp.825.
See Perikhanian, Judgments, pp.289, clause 1, 1013.
Ibid., pp.289, clause 1, 1315, 1, 1617.
On the situation of Christians under Sasanian rule see J. R. Russel,
Christianity in Pre-Islamic Persia:Literary Sources, Encyclopedia Iranica,
vol. V, pp.5246; A. Christensen, LIran sous les Sasanids, Osnabrck 1971,
pp.26399.
See J. Neusner, A History of the Jews in Babylonia, vol. V: Later Sasanian
Times, Leiden 1970, pp. 1112, esp. 6972; G. Widengren, The Status of
the Jews in the Sasanian Empire, Iranica Antiqua, 1 (1961), pp. 11762,
at pp.1413; on the Jewish community in Babylonia see Y. Elman, Middle
Persian Culture and Babylonian Sages: Accomodation and Resistance in
the Shaping of Rabbinic Legal Tradition, in Charlotte E. Fonrobert and
Martin S. Jaffee, eds., Cambridge Companion to the Talmud and Rabbinic
Literature, Cambridge 2007, pp.16597; I. M. Gafni, The Jews of Babylonia
in the Talmudic Era:A Social and Cultural History, Jerusalem 1990; M. Beer,
The Babylonian Exilarchate in the Arsacid and Sasanian Periods, Jerusalem
1970; M. Beer, The Babylonian Amoraim:Aspects of Economic Life, Ramat
Gan 1982.
Widengren, Jews, p. 158; this was also attempted by Sha4pu4r II. See
Christensen, LIran, p.267.
Neusner, History, pp.134; Christensen, LIran, pp.26873.
Neusner, History, pp.78; Christensen, LIran, pp.2701.
See, chap.1.
Widengren, Jews, pp.1401; Christensen, LIran, p.110; Neusner, History,
pp.434; on the status of the resh galutha see now G. Herman, The Sasanian
218
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
Notes to pp. 120123

Exilarchate in Babylon, Ph.D. thesis, the Hebrew University, Jerusalem 2005
(Hebrew), chaps. 5, 6, pp.24166, forthcoming in English:A Prince without
a Kingdom, Texts and Studies in Ancient Judaism, Tbingen. On the status
of the Nestorian Catholicos see also chap.7. See also n. 308.
Widengren, Jews, p.155; Christensen, LIran, p.2701, 2849.
Widengren, Jews, p.156.
See Linder, Jews, pp.312.
Ibid., pp.467.
See A. Schminck, Basilika, in Kazhdan et al., eds., Oxford Dictionary of
Byzantium, vol. I, p. 265; A. Kazhdan, The Basilika as a Source, ibid.,
pp.2656.
See chap.2, nn. 535.
Linder, Jews, p.2301.
Ibid., pp.2989.
Although some jurisprudents continued the theoretical discussion, and interpreted Shuru4t Umar according to the developing needs and views of their
times. See Fattal, Le statut, pp.1748.
See p. 117, for the Byzantine law.
For the Byzantine law stating that we prohibit that any synagogue shall
be erected in a new building, granting leave to prop up the old ones which
threaten immediate ruin see Linder, Jews, pp. 327, 32930. The fact that
this was indeed the policy in the early days of Islam is corroborated well by
the story regarding the repair works that were made by Thomas, Patriarch
of Jerusalem in the beginning of the ninth century, when the dome of the
Church of the Holy Sepulchre had to be rebuilt after having collapsed. The
argument between Thomas and the Muslim authorities was about whether
he had rebuilt the dome according to its previous measurements, or had
taken this opportunity to enlarge it, so that it was now larger than the Dome
of the Rock. See Eutychius, Tarkh, pp.556.
Ibn Asa4kir, Tarkh, vol. II, p. 174 or p. 178; Ibn al-Murajja4, Fada4il, p. 56:
" >`_ - $F $ $ )(7.
Ibn Qayyim al-Jawziyya, Ahka4m, vol. I, p.659; al-Turtu4sh3, Sira4j, pp.3523;
see chap.2, nn. 838.
See al-Sha4fi3, Kita4b al-umm, vol. IV, pp. 132, 188; al-Shayba4n3, Kita4b
al-mabsu4t, ed. Abu4 al-Wafa4 al-Afgha4n3, Karachi, n.d., vol. V, p. 213
(al-Sarakhs3, Kita4b al-mabsu4t, Beirut 1994, vol. XIII, p. 130); Fattal, Le
statut, p. 149, who notes the affinity to the Byzantine law; Friedmann,
Tolerance, pp.1623.
Al-Turtu4sh3, Sira4j, p.401.
See R. Brunschvig, Abd, EI2, vol. I, pp.2440.
See Goitein, Mediterranean Society, vol. I:Economic Foundations, Berkeley
1967, pp.13047.
See Fattal, Le statut, pp.23640.
See n. 64.
See Ibn Zabr, Juz, pp.1512; Tabar
,3 Tarkh
, vol. III, p.1390; Ibn al-Jawz,3
al-Muntazam
, vol. XI, pp. 2223, 238; Ibn Naqqash,
4
Fetoua; Ahmad
b.
Al3 al-Qalqashand,3 Subh
al-asa4 f sina4 at al-insha,4 Cairo 1964, vol. XIII,
Notes to pp. 123126
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
219
pp 3668; see also the story about the Jewish tax-collector in Hasan b. Al3
al-Tu4s3 (= Niza4m al-Mulk), Siyar al-mulu4k, trans. H. Darke, London 1978,
pp.1703.
Q 3:28, 3:118, 4:115, 4:141, 4:144, 5:51, 5:57, 9:8, 9:10, 9:23, 60:1, 60:13.
Fattal notes that this doctrine was undoubtedly based on the Byzantine
legislation:Le statut, p.239.
Tabar3, Tarkh, vol. III, p.390; Kramer, Decline, p.90.
Friedmann, Tolerance, pp.12159; as early as the eighth century opinions
among Muslim jurists varied as to the latitude that should be given to an
apostate to repent before the execution of the capital punishment. See ibid.,
pp.12832.
See Fattal, Le statut, pp. 12934; Friedmann, Tolerance, pp. 16093,
esp.1612.
Fattal, Le statut, pp.1389.
Ibid., pp.13744; Friedmann, Tolerance, pp.35, 187.
Friedmann, Tolerance, p.35.
Linder, Jews, pp.2378, 285.
Fattal, Le statut, pp.1224; Friedmann, Tolerance, pp.14952.
Abu4 Ubayd, Amwa4l, ed. Ima4ra, p. 271, para. 484 (=Amwa4l, p. 179,
para.482).
See al-Sha4fi3, Kita4b al-umm, vol. IV, p.118; see chap.2, nn. 115, 116, citing Christian sources proudly relating defamation of Islam by Christians
during the eighth and the ninth centuries.
See below.
See P. Crone, Roman, Provincial and Islamic Law, Cambridge 1987,
pp.117, which provides a comprehensive survey of the previous views on
the question; see also P. C. Hennigen, The Birth of a Legal Institution:The
Formation of the Waqf in Third-Century AH Hanaf Legal Discourse,
Leiden 2004. Hennigen also supports the claim that Muslim law was based
on an array of laws from different Near Eastern societies:see pp.609.
Some, though not all, of these rules appear in Shuru4t Umar. These include
the laws regarding churches and synagogues (clause 1); conversion to
Islam (clause 4); and, partially, the law regarding the ownership of slaves
(clause13).
See Noth, Abgrenzungsprobleme.
The issue of rearing pigs does not appear in al-Turtu4sh3, Sira4j, but does
appear in Ibn al-Murajja4, Fada4il, p.56, and in Ibn Qayyim al-Jawziyya,
Ahka4m, vol. II, p.659.
See the versions in Ibn al-Murajja4, Fada4il, p.59; Ibn Qayyim al-Jawziyya,
Ahka4m, vol. II, p.659; Ibn Asa4kir, Tarkh, vol. II, p.179.
See n. 39; and nn. 923.
Ibn al-Murajja4, Fada4il, p. 57; Ibn Qayyim al-Jawziyya, Ahka4m, vol. II,
p.660; Ibn Zabr, Juz, p.137.
See an even more detailed version in Ibn Asa4kir, Tarkh, vol. II, pp. 179
80, regarding 7 > % 7 $9 7 < " - 0F$ - ]- 0F &2) 7
3 $)' 7 )& 7 9 A) (chap. 3, nn. 1921).
220
Notes to pp. 127128
107. See M. Kister, Do Not Assimilate Yourselves La4 tashabbahu4,

Jerusalem Studies in Arabic and Islam, 12 (1989), pp. 32153; on this
subject, see extensively below.
108. See Noth, Abgrenzungsprobleme, pp. 3057; trans. Muelhaeusler,
pp.11418.
109. See chap.3.
110. Kister, La4 tashabbahu4, pp.3235.
111. Ibid., p.349.
112. The intimate connection between the Sasanian and Muslim social ideologies has been demonstrated in L. Marlow, Hierarchy and Egalitarianism
in Islamic Thought, Cambridge 1997 pp.6677; see also D. Gutas, Greek
Thought, Arabic Culture: The Graeco-Arabic Translation Movement
in Baghdad and Early Abba4sid Society (2nd4th/8th10th Centuries),
London and New York 1998 p.29, which claims that theAbba4sid dynasty,
in addition to being the descendants of the prophet was at the same time
the successor of the ancient imperial dynasties in Ira4q and Ira4n, from the
Babylonians through the Sasanians, their immediate predecessors.
113. On the process of Islamization of Zoroastrian Iran see J. Choksy, Conflict
and Cooperation:Zoroastrian Subalterns and Muslim Elites in Medieval
Iranian Society, New York 1977.
114. P. Cartledge, Alexander the Great:The Hunt for a New Past, Woodstock
and New York 2004, p.190.
115. See T. Gergel, ed., Alexander the Great:Selected Texts from Arrian, Curtius
and Plutarch, New York 2004, p. 82 (Plutarch), pp. 1034 (Curtius);
M. Broisus, Alexander and the Persians, in J. Roisman, ed., Brills
Companion to Alexander the Great, Leiden 2003, pp. 16993. Broisus
claims that despite Alexanders adaptation of many features of Persian
kingship and imperial rule, theres no evidence to suggest that these adaptations were any more than superficial statements which lacked political
knowledge and understanding of the organization of the Persian empire
and its administration (p. 173); E. Fredricksmeyer, Alexander and the
Kingship of Asia, in A. B. Bosworth and E. J. Baynham, eds., Alexander the
Great in Fact and Fiction, Oxford 2000, pp.13666; Cartledge, Alexander,
pp.2013.
116. See A. B. Bosworth, Conquest and Empire: The Reign of Alexander the
Great, Cambridge 1988, pp.2847; Fredricksmeyer, Alexander.
117. T. A. Sinclair, A History of Greek Political Thought, London 1951, p.287.
118. The most prominent example of this being the depiction of the emperor
Justinian and the empress Theodora in the impressive mosaic in the
Church of San Vitale in Ravenna (dedicated in 547), in which the
emperor appears crowned in imperial purple, with his courtiers, guards,
and clerics, while the crowned Theodora is depicted with her brilliantly
gowned handmaidens. The edges of Theodoras robe are outlined by a
embroidered gold hem, a prestigious status symbol both in Byzantium
and in Persia, which was adopted in the caliphate under the name of tira4z.
On the Ravenna mosaics see I. Andreescu-Treadgold and W. Treadgold,
Procopius and the Imperial Panels of San Vitale, Art Bulletin, 99 (1997),
pp.70823.
Notes to pp. 128131
221
119. See B. Croke, Justinians Constantinople, in M. Maas, ed., The Cambridge

Companion to the Age of Justinian, Cambridge 2005, pp.6086.
120. See S. Gordon, A World of Investiture, in S. Gordon, ed., Robes and
Honor:The Medieval World of Investiture, New York 2001, pp.122, at
pp.1012.
121. M. M. Mango, Status and its Symbols, in C. A. Mango, ed., The Oxford
History of Byzantium, Oxford 2002, pp.603, esp.p.60.
122. J. Rose, Sasanian Splendor:The Appurtenances of Royalty, in S. Gordon,
ed., Robes and Honor:The Medieval World of Investiture, New York 2001,
pp. 3558, at pp. 489, citing D. Talbot-Rice, Art of the Byzantine Era,
London 1963, pp.4452. C. A. Bromberg, Sasanian Royal Emblems in the
Northern Caucasus, in G. Gnoli and A. Panaino, eds., Proceedings of the
first European Conference of Iranian Studies, part 1, Rome 1990, pp.118
emphasizes the wide influence of royal Iranian emblems and symbols outside Sasanian society, saying: In the Byzantine world as in the Orient,
Sasanian concepts of majesty retained their value long after the Empire
had vanished before the onslaught of Islam, and the special emblems of the
Sasanian ruler spoke of authority and legitimacy to Byzantine, Armenian
or Muslim courtiers, or to Barbarian tribesmen (p.13).
123. See e.g. J. Haldon, Economy and Administration: How did the Empire
Work?, in Maas, ed., Age of Justinian, pp.2859, at p.36. He adds that
such opportunities were by definition very limited, however, so that the
vast majority of the population undoubtedly remained in the social group
into which they were born; see also R. MacMullen, Social Mobility and
the Theodosian Code, Journal of Roman Studies, 54 (1964), pp.4953.
124. On this issue see nn. 1945.
125. See Levy-Rubin, Changes.
126. This idea is expressed clearly by Marlow, Hierarchy, p.65:In the case of
Iran, the Arabs rapidly conquered the entire region of the former Sasanian
empire, and Iranians converted to Islam early and in large numbers. As a
result, in Iran (and Iraq) the Arabs came into contact with a living tradition
which had large numbers of spokesmen to ensure that it was perpetuated;
see also Morony, Iraq, pp.199208.
127. Silk was the most prestigious fabric in both the Byzantine and Sasanian
empires. The Byzantines, who were dependent on the Persians for its supply, went to great lengths to obtain it and to try to circumvent the need
for Persian mediation. In the sixth century the emperor Justinian managed to establish a silkworm industry in Constantinople. See G. Greatrex,
Byzantium and the East in the Sixth Century, in Maas, ed., Age of
Justinian, pp.477509, at pp.5012.
128. Different variations of this story appear in Muslim and Christian sources.
See e.g. Ibn al-Murajja4, Fada4il, p.49 (in the name of Muhammad b. A"idh
al-Qurash3, d. 233/4, see p.44, n. 2); Dionysius of Tell-Mahre4 in Palmer
et al., Chronicles, para. 75, p.162; Theophanes, Chronographia, p.339;
Agapius b. Mahbu4b, Unwa4n, pp.21315.
129. Dionysius of Tell-Mahre4 in Palmer et al., Chronicles, para. 75, p. 162.
On Umar Is entry to Jerusalem see Della Vida, Umar (I) b. al-Khatta4b,
p.820; and Crone and Hinds, Gods Caliph, pp.11314.
222
Notes to pp. 131135
130. Ahmad b. Yahya al-Bala4dhur3, Ansa4b al-ashra4f, ed. M. Schloessinger,

vol.IVa, Jerusalem 1971, p.125.
131. Al-Tabar3, Tarkh, vol. I, p.2402; Friedmann, Battle, pp.1889. Yalmaq
(pl. yala4miq) is an Arabicized form of the Persian word taken in its turn
from the Turkish, meaning a thick quilted coat. For this, see G. Clauson,
An Etymological Dictionary of Pre-Thirteenth-Century Turkish, Oxford
1972, p.92.
132. See al-Qalqashand3, Subh, vol. III, p. 266; Ibn Khaldu4n, who dedicates a full
chapter to The transformation of the caliphate into royal authority, notes
apologetically that Umar meant by royal Sasanian manners (%" )the
attitude of the Sasanian rulers, which consisted in doing worthless things,
constantly practicing oppression and neglecting God. Mua4wiyah replied
that he was not interested in royal Sasanian manners as such, or in the
worthlessness connected with them, but his intention was to serve God.
Therefore, Umar was silent (Abd al-Rahma4n b. Khaldu4n, al-Muqaddima
li-l-alla4ma Ibn Khaldu4n, Beirut n.d., p. 160; trans. F. Rosenthal as The
Muqaddimah, London 1967, vol. I, p. 417).
133. Ibn Khaldu4n, Muqaddimah, trans. Rosenthal, vol. I, p.417, vol. II, p.50.
134. Kalfon-Stillman, Arab Dress, pp.16, 2223; Muhammad b. Isa4 al-Tirmidh3,
Sunan, Beirut 1983, Abwa4b al-liba4s, vol. III, nos. 177489, 17929,
18078.
135. See G. E. von Grnebaum, Islam:Essays in the Nature and Growth of a
Cultural Tradition, London 1961, p.36.
136. See E. Bosworth, The Heritage of Rulership in Early Islamic Iran and the
Search for Dynastic Connections with the Past, Iran, 11 (1973), pp.5162.
137. Kalfon-Stillman, Arab Dress, pp.2961.
138. See Morony, Iraq, p.263.
139. Ibid., pp.3242.
140. See B. Spler, Iran in Frh-Islamischer Zeit, Wiesbaden 1952, pp.51516.
Al-Bala4dhur3 describes how, following the conquest of Kunduhar, the conqueror, Abba4d b. Ziya4d, saw there the tall qalansuwas of its people and
had some made in that fashion, and they were called (after him) al-Abba4d
iyya:al-Bala4dhur3, Bulda4n, p.434.
141. Kalfon- Stillman, Arab Dress, p.35.
142. See R. Talgam, The Stylistic Origins of Umayyad Sculpture as Shown in
Khirbat al-Mafjar and Mshatta, Ph.D. thesis, Hebrew University, Jerusalem
1996 (Hebrew with English abstract), pp.1656.
143. Ibid.
144. Ibid., pp.356.
145. See ibid., pp. 13651; R. Ettinghausen and O. Grabar, The Art and
Architecture of Islam 6501250, Harmondsworth 1987, pp.5670.
146. Kalfon-Stillman, Arab Dress, pp.367.
147. Ibid., p.42.
148. Ibid., p.47.
149. P. Sanders, Mara4sim, EI2, vol. VI, pp. 51820; P. Gignoux, Courts
and Courtiers iii, Encyclopedia Iranica, vol. VI, pp. 35664; Bosworth,
Heritage; D. Sourdel, Questions de cremonial abba4sside, Rvue des
tudes Islamiques, 28 (1960), pp.12148.
Notes to pp. 135136
223
150. See e.g. Pseudo-al-Ja4hiz, in the mid-ninth century, who says: 03$97
9% $ $A M$k P+%+ 1 E <) $)> 0F 9 E - 7 0) $
$F23 9 $/2<7 $F`> 5j ( ' Pseudo-al-Ja4hiz, Kita4b al-ta4j, ed. A. Z.
Pacha, Cairo 1914, p. 23); French trans. C. Pellat as Le livre de la couronne attribut Gahiz, Paris 1954, p. 51; Abu4 al-Hasan Al3 b. Husayn
al-Masu4d3, Muru4j al-dhahab wa-maa4din al-jawhar, ed. and French
trans. Barbier de Meynard and Pavet de Courteille as Les prairies dor,
Paris 186677, vol. II, pp. 1569; rev. edn C. Pellat, Beirut 1966, vol. I,
pp. 28687; rev. trans. C. Pellat as Masu4d, Les prairies dor, Paris 1962,
pp. 21819,.
151. See Ibn Khaldu4n, Muqaddimah, trans. Rosenthal, vol. II, p.48ff.72; see
S.Shaked, From Iran to Islam:On Symbols of Royalty, Jerusalem Studies
in Arabic and Islam, 7 (1986), pp.7591.
152. See Kalfon-Stillman, Arab Dress, p. 48, citing Abu4 al-Husayn Hila4l b.
al-Muhassin al-Sa4b3, Rusu4m da4r al-khila4fa, ed. M. Awwa4d, Baghdad 1964,
p. 912, trans. E. Salem as The Rules and Regulations of the Abba4sid
Court, Beirut 1977, p. 74. Regarding red as the color of royal clothing
in the Sasanian period, see J. Rose, Investiture (in The Sasanian Period),
Encyclopedia Iranica, vol. XIII, p.186, citing C. Manson Bier, Textiles, in
Prudence Oliver Harper, The Royal Hunter:Art of the Sasanian Empire,
New York 1978, p.123 (unfortunately, I was not able to consult this book).
153. See C. E. Bosworth, Laqab, EI2, vol. V, p.619:Within a formalized hierarchical society such as the Islamic one became in the Abba4sid period,
modes of address, insignia or rank and office, dress, etc., all contributed to
the fixing of a mans status in society and the state, and as such were prized
as the visible and audible symbols of success in the temporal world.
154. See Pseudo-al-Ja4hiz, Ta4j, p.23.
155. Abu4 Umar Ahmad b. Muhammad b. Abd Rabbih, al-Iqd al-fard,
Cairo 1321.
156. See Hila4l al-Sa4b3, Rusu4m; Hila4l al-Sa4b3, Rules. See also Tu4s3, Siyar.
157. See Rose, Investiture; Rose, Sasanian Splendor, pp.478.
158. On both of these see A. Tafazzoli, Sasanian Society, New York 2000,
pp. 1859; on the dihqa4ns see M. Zakeri, Sa4sa4nid Soldiers in Early
Muslim Society:The Origins of Ayya4ra4n and Futuwwa, Wiesbaden 1995,
pp.4356, 5569, 10112.
159. See chap.5, n. 49.
160. On this, see below.
161. Tafazzoli, Sasanian Society, pp.1837.
162. Ibid., p.41; Zakeri, Soldiers.
163. On the status and role of the dihqa4ns following the conquest see Morony,
Iraq, pp.2006. On conversions of dihqa4ns to Islam see esp.p.205; Zakeri,
Soldiers, pp.10112, regarding the issue of their conversion see pp.1025.
164. See e.g. Theophanes, Chronographia, p. 367; Theophanes, Chronicle,
pp.5245; Palmer et al., Chronicles, pp.2024; R. Frye, The Golden Age of
Persia:The Arabs in the East, London 1975, pp.1819; Tafazzoli, Sasanian
Society, p.43. This phenomenon was not sweeping, and there were dihqa4ns
who refused to cooperate and chose to flee or resist:see Tafazzoli, Sasanian
Society, pp.445.
224
Notes to pp. 137138
165. See Tafazzoli, Sasanian Society, pp.4359; Frye, Golden Age, pp.1626.
166. On Sasanian Iran see V. S. Curtis and S. Stewart, The Sasanian Era:The
Idea of Iran, vol. III, London 2008; Pourshariati, Decline; J. Wiesehfer,
Ancient Persia from 550 BC to 650 AD, London and New York 1996;
Christensen, LIran.
On the class system specifically see M. Shaki, Class System: iii,
Encyclopedia Iranica, vol. V, pp. 6548; Marlow, Hierarchy, pp. 6677;
Rubin, Nobility; Z. Rubin, Eastern Neighbours:Persia and the Sasanian
Monarchy (224651), in Jonathan Shepard, ed., Cambridge History of the
Byzantine Empire, vol. I, Cambridge 2008, pp.13955; Tafazzoli, Sasanian
Society, pp.2, 20; Perikhanian, Iranian Society, pp.6326; Frye, Golden
Age, p.18.
167. On this see P. O. Skjrv, Class System: i: in the Avesta, Encyclopedia
Iranica, vol. V, pp.6501, where it is noted that there may also have been
a tradition of four estates in the Avesta.
168. For a thorough review of this issue and a comprehensive bibliography see
Marlow, Hierarchy, pp.6872.
169. Perikhanian, Iranian Society, pp.6323; E. Benveniste, Les classes sociales dans la tradition avestique, Journal Asiatique, 221 (1932), pp.11734.
170. Tafazzoli, Sasanian Society, p. 27, citing Abul-Qasim Firdaus3, Sha4hna4ma, ed. E. Bertel et al., Moscow 196671, vol. VIII, pp.2979:Firdaus3,
Sha4hna4mah, vol. VI, pp.41317.
171. See Abu4 Mansu4r Abd al-Malik b. Muhammad b. Isma43l al-Thaa4lib3,
Ghurar akhba4r mulu4k al-Furs:Histoire des rois des Perses, ed. and trans.
H. Zotenberg, Paris 1900, p.608.
172. See chap.4, nn. 39 and nn. 624.
173. See Benveniste, Classes sociales, pp. 11920; 1304. On pp. 1324 he
adduces Firdaus3s version with some minor but illuminating corrections;
see also al-Tabar3, Tarkh, vol. I, p. 180: 5j +$5 5j :$5j ; $ m M
A ' $5` E+ 5j % $> 0F 5j k+[ %> $9$ M $$2 5j $F5$ ; ' al-Thaa4lib3, Ghurar, pp. 1112.
174. See Tafazzoli, Sasanian Society, p.41 for different versions by al-Tabar3,
Tarkh, vol. I, p.434; al-Thaa4lib3, Ghurar, pp.67, 6768 and more; see
also Zakeri, Soldiers, pp. 456. citing al-B3ru4n3, al-A"tha4r al-ba4qiya an
al-quru4n al-kha4liya, ed. E. C. Sachau, Leipzig 1879, pp.220221; trans. E.
C. Sachau as The Chronology of Ancient Nations, Leipzig 1897, p.206. On
Manu4shihr, see E. Yarshater, Iranian National History, in Yarshater, ed.,
Cambridge History of Iran, vol. III (1), pp.4334; A. Tafazzol3, Dehqa4n,
Encyclopedia Iranica, vol. VII, pp.2235.
175. See M. Grignaschi, Quelques spcimens de la littrature sassanide conservs dans les bibliothques dIstanbul, Journal Asiatique, 254 (1966),
pp.1135.
176. Ibid., p. 54 (French trans. p. 74); I. Abbas, Ahd Ardashr, Beirut 1967,
pp.624.
177. See Grignaschi, Spcimens, p.54 (French trans. p.74); and Abbas, Ahd
Ardashr, pp. 624; a similar text, which contains an admonition to
the king regarding the confusion of the social order, and describing the
Notes to pp. 138141
178.
179.
180.
181.
182.
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
225
four-class structure, is found in Pseudo-al-Ja4hiz, Ta4j, pp. 245 (see also

Pellat trans., Couronne).
Grignaschi, Spcimens, pp.45.
Kita4b al-ta4j f srat Anu4shirwa4n li-Ibn al-Muqaffa:Grignaschi, Spcimens,
intro. pp.45, 1018 (text), 12935 (French trans.).
Ibid., p. 103: 0F ' - K2k 0F+% =9 $ 8(
Christensen, LIran, pp.636, argues that it was written in the sixth century; Boyce in the introduction to M. Boyce, trans., The Letter of Tansar,
Rome 1968, pp.522, supports an early date with later additions; Shaked,
Transmission, p.55, n. 22, does not negate the possibility of its having
been written in the third century.
See Yarshater, ed., Cambridge History of Iran, vol. III (2), pp. 731, 933,
1184, 1272.
Nameh Tansar, ed. M. Minovi, Tehran 1932 p.19; Boyce, trans., Letter of
Tansar, p.44.
Nameh Tansar, pp.234; Boyce, trans., Letter of Tansar, pp.489.
idha4 kuntumu4 li-nna4si ahla siya4satin
fa-su4su4 kira4ma nna4si bi-rrifqi wal-badhli
wa su4su4 lia4m nna4si bi-dhdhulli yaslahu4
ala4 dhdhulli inna4 dhdhulla aslahu li-l-nadhli:
Nameh Tansar, p. 37 (=Letter of Tansar in Ibn Isfandiya4r, Tarkh
Tabarista4n, ed. Abba4s Iqba4l, Tehran 1942, p.35); Boyce, trans., Letter of
Tansar, p.60.
See Grignaschi, Spcimens, pp. 23, 91102 (text), 11128 (French
trans.).
Ibid., p.94.
See Christensen, LIran, pp. 31620; for the opposing views see
esp.Pourshariati, Decline; and Rubin, Nobility.
See Boyce, trans., Letter of Tansar, pp.434:
As for another passage in which you wrote The King of Kings has
had established new customs and new ways , know that the decay of
family and rank is twofold in nature. In the one case men pull down the
family and allow rank to be unjustly lowered; in the other it is time itself
without anothers endeavour which deprives them of honour and worth
and the splendour of position. Degenerate heirs appear who adopt boorish ways and forsake noble manners and lose their dignity in the sight
of the people. They busy themselves like tradesmen with the earning of
money, and neglect to garner fair fame. They marry among the vulgar and
those who are not their peers, and from that birth and begetting men of
lower character appear; and this is what is meant by decadence of rank.
The Kings of Kings has issued a decree to exalt and ennoble their rank,
whose like we have not heard from any man.
See Rubin, Nobility, pp.24659; Pourshariati, Decline, pp.2156.
Rubin, Nobility, p.243.
See R. Gyselen, Le gographie administrative de lempire sassanide: les
tmoignages sigillographiques, Paris 1989, pp.152, 156.
Pourshariati, Decline, p.92.
226
Notes to pp. 141144
194. See Marlow, Hierarchy, pp. 6690; see also the survey on social stratification in Iranian and Muslim society in M. Shaki, Class System iii: In
the Parthian and Sasanian Periods and A. Ashraf and A. Banuazizi, Class
System iv:Classes in Medieval Islamic Persia, Encyclopedia Iranica, vol.V,
pp.65467; on this particular issue see pp.6626, with references to A. K.
S. Lambton, Theory and Practice in Medieval Persian Government, London
1980, chap. 7; A. K. S. Lambton, Continuity and Change in Medieval
Persia, Albany 1988, pp.221346, esp.pp.2214.
195. See Marlow, Hierarchy; for a good survey of this subject see Shaki, Class
System iii; Ashraf and Banuazizi, Class System iv.
196. See Gutas, Greek Thought, pp.2960.
197. See I. Goldziher, Muhammedanische Studien, Halle 188990, vol. I, chap.3,
pp.10146, ed. and trans. C. R. Barber and S. M. Stern as Muslim Studies,
New York 1977, pp.98136; P. Crone, Mawla4, EI2, vol. VI, pp.87482;
R. Levy, The Social Structure of Islam, Cambridge 1957, pp.5367.
198. Crone, Mawla4,p.876, citing al-Bala4dhur3, Ansa4b, vol. IVb, p.10; see also
chap.3, n. 71.
199. See Crone, Mawla4, regarding marriage, diya, and positions, and references there.
200. See Goldziher, Muhammedanische Studien, vol. I, p. 267; Goldziher,
Muslim Studies, p.242.
201. See Levy, Social Structure, p.59; Muhammad b. Yaz3d al-Mubarrad, Kita4b
al-ka4mil, ed. W. Wright, Leipzig 1864, pp.71112.
202. Sha4fi3, Kita4b al-umm, vol. IV, p. 118. See also Levy-Rubin, Continuatio,
pp. 912: he should not sit in front on a velvet-like sofa ($> +%).
203. Al-Tabar3, Tarkh, vol. III, pp.139091,
204. Ibid., vol. I, p.962.
205. See nn. 10610.
206. See Nameh Tansar, p.19.
207. Ibid., p. 23. This is based on a slight emendation of the text, as well as
Boyces translation of it: %5+ $ /< instead of %5+ $ /<. I would
like to thank Dr Vera Moreen and Dr Julia Rubanovich for their aid in
reading the Persian text.
208. The Letter of Tansar was translated from Pahlavi into Arabic by Ibn
al-Muqaffa (d. c. 760 CE), a translation which was lost, and then again
into New Persian in the thirteenth century. Only the New Persian version
has survived (on this see Boyces introduction to Letter of Tansar). Still,
even in Ibn al-Muqaffas time, ghiya4r would already have been a taken
word, kept for non-Muslims, and therefore other analogous terms in
Arabic were most probably chosen by him. The words used in the New
Persian translation for difference, or distinction, i.e. tafa4wut, tamyz, and
taqdr, all originate in Arabic rather than in Pahlavi, and may have been
used by Ibn al-Muqaffa in the Arabic translation.
209. See Lukonin, Institutions, pp.68398.
210. Ibid., p.699700.
211. See ibid., pp.70913.
Notes to pp. 145146
227
212. Unlike the Sasanian Book of Ranks, which has been lost, but which,
according to al-Masu4d3, included at its apogee 600 ranks. See M. L.
Chaumont, Lorder des prsances la cours des Arsacides dArmenie,
Journal Asiatique, 254 (1966), pp.47197, at p.491, n. 3.
213. See Gignoux, Courts and Courtiers, citing Gnalogie de la famille de
saint Grgoire et vie de saint Nerses, Venice 1853, pp.329 (in Armenian);
Chaumont, Prsances; Shaked, Symbols of Royalty, pp. 7779;
N.Adontz, Armenia in the Period of Justinian, trans. with partial rev. and
bibliog. N. G. Garsoian, Lisbon 1970, pp.1868.
214. Gnalogie; Chaumont, Prsances, pp.4813.
215. Xenophon, Cyropaedia, viii.4.35:So when invited guests came to dinner,
he did not assign them to their seats at random, but he seated on Cyruss
left the one for whom he had the highest regard and the one who was
second in esteem he seated on his right for he thought it a good plan to
show publicly how much regard he had for each one He did not however assign the appointed places permanently, but he made it a rule that by
noble deeds any one might advance to a more honoured seat, and that if
anyone should conduct himself ill he should go back to one less honoured
(trans. W. Miller, Loeb Classical Library, Cambridge, MA and London
1979, pp. 3757; D. Khaleghi-Motlagh, Ba4r (audience), Encyclopedia
Iranica, vol. III, pp.7304.
216. Masu4d3, Muru4j, ed. de Meynard and de Courteille, vol. II, pp.1569; ed.
Pellat, vol. I, pp.2867; trans. Pellat, pp.21819; see n. 177, referring to
Pseudo-al-Ja4hiz, Ta4j, pp.245, trans. Pellat, pp.513.
217. See Sourdel, Questions, pp. 13940, citing Tabar3, Tarkh, vol. III,
p.1503.
218. See Hila4l al-Sa4b3, Rules, p.78.
219. Levy-Rubin, Continuatio, pp. 912; the text originally has h-m-a-l,
emended to khaml, which may be a velvet cushion, or cushion of feathers.
For martaba in this sense see Sadan, Mobilier, pp.546.
220. Sha4fi3, Kita4b al-umm, vol. IV, p.119.
221. See Ibn Battu4ta, Rihlat Ibn Battu4ta4, Cairo 1957 pp. 3536 describing the
court of the sultan Muhammd b. Tughluq: $$ ]%K `/ =9 3
+% $F<-; and again pp. 389, trans. as The travels of Ibn Battuta, AD
13251354, trans. with revisions and notes from the Arabic text edited by
C. Defremey and B. R. Sanguinetti, by H. A. R. Gibb, Cambridge 1958
2000, pp. 660, 666.
222. See al-Tabar3, Tarkh, vol. III, pp.138990. An additional related requirement is the prohibition on visiting Muslims in their homes. See Ibn
al-Murajja4, Fada4il, p.57; Ibn Qayyim al-Jawziyya, Ahka4m, vol. II, p.660.
223. See Kister, La4 tashabbahu4, p.327, n. 23, citing a series of sources.
224. Xenophon, Cyropaedia, viii.3.9:see also Gignoux, Courts and Courtiers,
p.358 (bottom). On the policemen and the court security police Gignoux
cites M. W. Stopler, Enterpreneurs and Empire:The Mura Archive, the
Mura Firm and the Persian Rule in Babylonia, Leiden 1985, p.63 n. 51,
with further references.
228
Notes to pp. 146148
225. See a more detailed version in Ibn Asa4kir, Tarkh, vol. II, pp.17980, and
see another variation on this on p.185; Ibn Zabr, Juz, pp.139, 140; see
also chap.3, nn. 1921.
226. See E. H. Peck, Clothing vii: From the Arab Conquest to the Mongol
Invasion, Encyclopedia Iranica, vol. V, pp.76070.
227. Al-Tabar3, Tarkh, vol. I, pp.20256, 2038, and see Zakeri, Soldiers, p.75
for additional references.
228. See Zakeri, Soldiers, p.75 and references there.
229. Al-Jahshiya4r3, Wuzara4, p.3; Tafazzoli, Sasanian Society, p.27.
230. Al-Tabar3, Tarkh, vol. I, p. 434; W. M. Brinner, annot. and trans., The
History of al-Tabar, vol. III:The Children of Israel, Albany 1985, p.23.
231. Al-Masu4d3, Muru4j al-dhahab, vol. I, pp. 2478, section 662. Tafazzoli,
Sasanian Society, p.43.
232. Al-Ja4hiz, Kita4b al-baya4n wal-tabyn, Cairo 1956, pp. 11819; KalfonStillman, Arab Dress, pp.427; see also Sourdel, Questions, pp.1334.
233. See A. S. Shahbazi, Clothing ii:In the Median and Achaemenid Periods,
Encyclopedia Iranica, vol. V, p. 729; E. H. Peck, Clothing iv: In the
Sasanian Period, Encyclopedia Iranica, vol. V, p.745; regarding the continuity in fashion see ibid., section viii; G. Widengren, Some Remarks on
Riding Costume and Articles of Dress among Iranian Peoples in Antiquity,
Arctica: Studia Ethnographica Upsaliensia, 11 (1956), pp. 22876, at
pp.2756.
234. Abu4 Yu4suf, Kita4b al-khara4j, pp.1278; see chap.3.
235. Widengren, Riding Costume, p.259; Kalfon-Stillman, Arab Dress, p.12.
236. On the prestige of silk, see p. 128.
237. For this hairstyle see also al-Tirmidh3, Sunan, vol. III, Abwa4b al-liba4s, no.
1778, and below, clause 10.
238. Levy-Rubin, Continuatio, pp.241 (text), 91 (trans.).
239. N. Rabbat, Tira4z, EI2, vol. X, pp. 5348; Kalfon-Stillman, Arab Dress,
pp.1223.
240. See M. Shaki, Class System iii: In the Parthian and Sasanian Periods,
Encyclopedia Iranica, vol. V, p.654.
241. See A. S. Shahbazi, Clothing, Encyclopedia Iranica, vol. V, pp. 7256;
W. Hinz, Darius und die Perser, Baden-Baden 19769, p.61; Widengren,
Riding Costume, p. 240; M. Boyce, History of Zoroastrianism, Leiden
1982, p.21.
242. See Kalfon-Stillman, Arab Dress, p. 122; for a general survey see e.g.
M.Reinhold, On Status Symbols, Classical Journal, 64/7 (1969), pp.3004.
243. See n. 152.
244. Hila4l al-Sa4b3, Rules, p.61.
245. It may be that the Iranian social colour code was at the root of the concept
that became central in Islamic culture according to which colour serves as
a mark of group identity (not necessarily a social group) or of religio-political allegiance. This does not mean that the same specific colours continued
to be the mark of certain groups, but rather that a certain colour chosen
could be declared an exclusive social or political indicator.
Notes to pp. 148152
246.
247.
248.
249.
250.
251.
252.
253.
254.
255.
256.
257.
258.
259.
260.
261.
262.
263.
264.
265.
266.
267.
229
Thus the Abba4sids insisted that anyone who entered their court should
be dressed in black. This is stressed in a story told by Hila4l al-Sa4b3. When
Muhammad b. Umayr (d. 999 CE) entered the court dressed in white,
which he claimed was his familys ancestral colour, he was immediately
thrown out. On another occasion, when al-Mamu4n named Al3 al-Rida4
as his heir, he exchanged the traditional Abba4sid black for a formal green
attire, as well as flags and banners:see Gignoux, Courts and Courtiers,
p. 361; Ibn Khaldu4n, Muqaddimah, trans. Rosenthal, vol. II p. 51. The
Fa4timids would later choose white as their dynastic colour.
Al-Tabar3, Tarkh, vol. III, pp.138994; Kramer, Decline, pp.8994; on
these restrictions see chap.4.
Levy-Rubin, Continuatio, pp.241 (text), 91 (trans.); see chap.4.
See Kalfon-Stillman, Arab Dress, p. 44, citing Abu4 al-Tayyib Muhammad
al-Washsha,4 Kitab
4 al-muwashsha4 aw al-zarf wa-al-zurafa4 , Cairo 1953,
p.161; it should be noted, however, that there were some circumstances in
which yellow had other connotations and functions. Thus, Hilal4 al-Sa b
4 3 says
that descendants of the ansa r4 should wear yellow garments and turbans when
appearing before the caliph. He notes, however, that this is theoretical since
there are no great personages left from among them. See Kalfon-Stillman,
Arab Dress, p.48, citing Hilal4 al-Sa b
4 3 , Rusum
4 , pp.912 (Hilal4 al-Sa b
4 3 , Rules,
p.74). Regarding Wald
3 IIs unusual habit of wearing yellow. which may have
been due to Indian influences, see Kalfon-Stillman, Arab Dress, p.34.
Tirmidh3, Sunan, vol. III, Abwa4b al-liba4s, no. 1779; see also KalfonStillman, Arab Dress, p.24, citing al-Bukha4r3, Sahh, kita4b al-liba4s, ba4bs
334, vol. IV, ed. Krehl, p.87.
Al-Tabar3, Tarkh, vol. III, pp.138990.
1$ %kK 7 7, see Abd al-Razza4q, Musannaf, vol. VI, p. 122.
See Ibn al-Murajja4, Fada4il, p.228, n. 3 and references there.
Ibn Asa4kir, Tarkh, vol. II, p.183.
See Goldziher, Muhammedanische Studien, vol. I, p. 267; Goldziher,
Muslim Studies, p.242.
Ibid.
Hila4l al-Sa4b3, Rules, p.49.
Al-Tu4s3, Siyar, pp.14950.
For the following survey see mainly A. S. Shahbazi, Asb, Encyclopedia
Iranica, vol. II, pp.72430.
See W. Sundermann, Arte4ta4r, Encyclopedia Iranica, vol. II, p.661.
On the asba4ra4n see Zakeri, Soldiers, pp.5768.
Xenophon, Cyropaedia, viii, 2.7, trans. Miller, p.3345.
Jahshiya4r3, Wuzara4, p.7.
Tafazzoli, Sasanian Society, p.50, citing al-Bala4dhur3, Bulda4n, p.271.
See F. Vir, Faras, EI2, vol. II, pp.7847.
Hila4l al-Sa4b3, Rules, pp.1617.
See e.g. ibid., pp.75, 77, 80.
See al-Maqr3z3, Khitat, vol. I, p.387; Hila4l al-Sa4b3, Rules, p.65; see also
Sourdel, Questions, esp.p.123, n. 10.
230
Notes to pp. 152153
268. See Shahbazi, Asb, p. 734, citing Abu4l Fazl Bayhaq3, Ta4rikh Masu4d,
ed.Al3 Akbar Fayya4z, Mashhad 1350/1971, pp.158, 159, 355, 373, 526.
269. See Abu4 Yu4suf, Kita4b al-khara4j, p.117.
270. See al-Tabar3, Ta4rikh, vol. III, pp. 138990; Levy-Rubin, Continuatio,
pp.241, 91 (trans.).
271. Levy-Rubin, Continuatio, pp.241, 91 (trans.); Abu4 Yu4suf, Kita4b al-khara4j,
p.127; Al3 b. Muhammad al-Ma4ward3, Ahka4m al-sulta4nya wa-al-wila4ya4t
al-dnya, Cairo 1994, p.185.
272. Xenophon, Cyropaedia, viii.1.8.
273. P. Calmeyer, E. H. Peck, and W. Sundermann, Belt, Encyclopedia Iranica,
vol. IV, pp.1307; see Widengren, Riding Costume, passim, esp.the illustrations; see Zakeri, Soldiers, pp.4950 on the military equipment of the
asa4wira4, which included, inter alia, a lance, a sword, and a spear.
274. See e.g. Hila4l al-Sa4b3, Rules, p.63:On procession days the chief chamberlain comes fully attired in black robe and black turban, wearing sword
andbelt.
275. See Hila4l al-Sa4b3, Rules, p.73, describing the caliph in audience:He wears
a black rusa4fiyyah on his head, and adorns himself with the sword of the
prophet may Allah bless him. He also keeps another sword on his left
between the two cushions of the throne. Al-Tu4s3, Siyar, p.94, talks about
twenty special sets of arms studded with gold, jewels and other ornaments, which were kept ready so that twenty pages finely attired can take
these weapons and stand round the throne when delegations arrive, since
today, there is no king on earth greater than The Master of the World So
it is fitting that wherever other kings possess one of a thing, our sovereign
should have ten for he has at his command all spiritual and material
resources, coupled with sound judgment. In fact, he lacks nothing of majesty and dominion. See also al-Qalqashand3, Subh, vol. III, pp.46871; M.
Canard, L crmonial ftimite et le ceremonial Byzantine, Byzantion, 21
(1951), pp.355420, at pp.38893; see below, clause 11.
276. See Ibn Khaldu4n, Muqaddimah, trans. Rosenthal, vol. II, pp.605; J. Allan,
Khatam, EI2, vol. IV, pp.11025; Sourdel, Questions, p.135.
277. See Sourdel, Questions, p.135.
278. Bukha4r3, Sahh, vol. VII, liba4s, p. 48; Ibn Khaldu4n, Muqaddima, vol. II,
p.53; Ibn Khaldu4n, Muqaddimah, trans. Rosenthal, vol. II, p.61. On traditions regarding the Prophets seal see al-Tirmidh3, Sunan, vol. III, Abwa4b
al-liba4s, nos. 17929. See also chap.1, pp. 1415, the signing and sealing
of the treaty between the Byzantines and the Persians.
279. Ibn Khaldu4n, Muqaddima, vol. II, p.57; Ibn Khaldu4n, Muqaddimah. trans.
Rosenthal, vol. II, p.65; see also Hila4l al-Sa4b3, Rules, pp.1034.
280. See note 235.
281. See e.g. al-Tirmidh3, Sunan, vol. III, Abwa4b al-liba4s, nos. 180810.
282. See the various representations in Glass, Gilding, and Grand Design:Art
of Sasanian Iran (224642), exhibition catalogue, Asia Society, New
York 2007; B. Goldman, The Later Pre-Islamic Riding Costume, Iranica
Antiqua, 28 (1993), pp. 20139, see illustrations passim; Widengren,
Riding Costume, illustrations passim.
Notes to pp. 153156
231
283. See Bravmann, An Yadin. Bravmann therefore interprets jizya as the

reward given by the captive to his generous captor; see also C. F. Robinson,
Neck Sealing in Early Islam, Journal of the Economic and Social History
of the Orient, 48 (2005), pp.40141, at pp.40910, 41314, and n. 59.
284. Bravmann, An Yadin, p.311.
286. Tabar3, Tarkh, vol. III, p. 1393: E+ !j$ 8 0$ 09 >$+
!j$ $1 r+$"1 %)$)' & 5`
288. This version is repeated by Ibn Qayyim al-Jawziyya, who adheres to the
term mintaqa when citing the edict. See above, chap.3.
289. The usual form in Greek is zo4ne, while zo4narion appears much more rarely.
See zo4ne and zo4narion, in E. A. Sophocles, Greek Lexicon of the Roman
and Byzantine Periods (from BC 146 to AD 1100), Cambridge 1914,
p.558.
290. A. S. Tritton, Zunna4r, EI2, vol. XI, pp.5712; Tritton, The Caliphs, p.117;
on the zunna4r see also Noth, Abgrenzungsprobleme, pp. 3045, trans.
Muelhaeusler, pp.1315; Cohen, Under Crescent and Cross, pp.623.
291. S. Shaked, No Talking during a Meal: Zoroastrian Themes in the
Babylonian Talmud, in C. Bakhos and R. Shayegan, eds., Talmud in its
Iranian Context: Proceedings of the Conference Talmud in its Iranian
Context, Los Angeles, 67 May 2007, Tbingen 2010, pp.121. See p.5,
n. 19, and additional references there.
292. Ibid., p.6.
293. Ibid., p.7.
294. Ibid., pp.810, n. 31, and additional references there.
295. Babylonian Talmud, Shabbat 9b: . : ?
- ,- . . : .
296. On the influence of the ritual use of the kustg on the Jews in Mesopotamia
see Elman, Middle Persian Culture, pp.1812.
297. See W. Eilers, Banda, Encyclopedia Iranica, vol. III, pp. 6823; see also
E. H. Peck, Belt ii: In the Parthian and Sasanian Period, Encyclopedia
Iranica, vol. IV, pp.1326, who supports the distinction between the two
types of belts through archaeological evidence.
298. See Zakeri, Soldiers, pp. 704; see G. Widengren, Le symbolisme de la
ceinture, Iranica Antiqua, 3 (1968), pp.13355, at p.141.
299. Widengren, Riding costume, pp.254, 260; G. Widengren, Quelques rapports entre juifs et iraniens lpoque des parthes, Vetus Testamentum,
Supplements, 4, Leiden 1957, pp.197241, at pp.2278.
300. Huyse, Ka4ba-i Zardut; see also chap.1, n. 23.
301. Nathan ben Yehiel, Aruch Completum, ed. A. Kohut, Vienna 1930,
vol.VII, p.127 (Hebrew), cited by Herman, Exilarchate, p.147, n. 651.
302. On another occasion it is reported that Yazdgird fixed Rabbi Huna bar
Nathans belt (hamya4na4), a gesture which is interpreted to mean that he
was highly regarded by the king: see Babylonian Talmud, Zevahim 19a.
However, the word kamar is not used there. It may well be that the term
hamya4na4 functioned both as a general term for belt and as a specific one
232
303.
304.
305.
306.
307.
308.
309.
310.
311.
312.
313.
314.
315.
316.
317.
318.
319.
320.
321.
322.
323.
324.
325.
Notes to pp. 156159

denoting the religious function, as is indicated in Babylonian Talmud,
Shabbat 49b mentioned above, where a kamar and a hamya4na4 are both
called hamya4ney; see also Herman, Exilarchate, pp. 11213, nn. 508
509 and additional references regarding kamar there (Hebrew); Shaked,
NoTalking, pp.89.
Al-Bala4dhur3, Bulda4n, p.502; al-Tabar3, Tarkh, vol. II, pp.150910.
A. Khazdan, Belt, in Khazdan et al., eds., The Oxford Dictionary of
Byzantium, vol. I, p.280; J. Braun, Die liturgische Gewandung im Occident
und Orient, Darmstadt 1964 pp.11517.
Linder, Jews, pp.3578 (text), 3601 (trans.).
Kalfon-Stillman, Arab Dress, pp.502. Hila4l al-Sa4b3, Rules, p.65:He was
surrounded by a hundred of his private servants in good attire, with colored garments, belts, and swords with trappings studded with jewelry.
Ibid., p.65. See also above, clause 7.
See Sourdel, Questions, p. 143: al-Tabar3, Tarkh, vol. III, p. 1233;
C. E. Bosworth, annot. and trans., The History of al-Tabar, vol. V:Storm
and Stress along the Northern Frontiers of the Abba4sid Caliphate, Albany
1991, p.92; see also Masu4d3, Muru4j al-dhahab, vol. IV, pp.3567.
Al-Tabar3, Tarkh, vol. III, p.1302; Bosworth, Storm and Stress, p.178.
Xenophon, Cyropaedia, viii, 3.9.
See P. Sanders and A. K. S. Lambton, Mawa4kib, EI2, vol. VI, pp.84958;
see nn. 14851.
On this see chap.2, nn. 645.
Ra4ya is an army flag or banner:see Ra4ya, in Lane, Lexicon.
See A. S. Shahbazi, Deraf, Encyclopedia Iranica, vol. VII, pp. 31215;
al-Thaa4lib3, Ghurar, pp.389, describing how the Persian standard, which
was conceived as a good omen during battle, was captured and cut into
pieces by Umar.
Ibn Khaldu4n, Muqaddima, vol. II, p.42; Ibn Khaldu4n, Muqaddimah, trans.
Rosenthal, vol. II, p.48:One of the emblems of the royal authority is the
outfit (a4lah), that is the display of banners and flags and the beating of
drums and the blowing of trumpets and horns.
Sourdel, Questions, p.144.
See P. Chalmeta and A. K. S. Lambton, Mara4sim, EI2,vol. V, pp.51829;
Christensen, LIran, pp.21012; Spler, Iran, p.348.
Ibn Khaldu4n notes that the caliphs granted their permission to certain officials who were accompanied by a procession carrying flags and
other a4la:see Muqaddima, vol. II, p.44; Muqaddimah, trans. Rosenthal,
vol. II, p.50.
Ibid., pp.1516.
Ibid., p.16.
See Christensen, LIran, p. 131, and n. 8, citing M. K. Patkanian, Essai
dune histoire de la dynastie sassanides, daprs les renseignements fournis
Notes to pp. 159163
326.
327.
328.
329.
330.
331.
332.
333.
334.
335.
336.
337.
338.
339.
340.
341.
342.
343.
344.
345.
346.
347.
233
par les historiens armniens Journal Asiatique, 1 (1866), pp.101238, see

p.112.
See Sourdel, Questions, p.144; A. Metz, The Renaissance of Islam, trans.
Salahuddin Khuda Bakhsh and D. S. Margoliouth, Patna 1937, pp.1334.
Ibid.
See Chalmeta and Lambton, Mara4sim.
Ibn Khaldu4n, Muqaddima, vol. II, p.44; Ibn Khaldu4n, Muqaddimah, trans.
Rosenthal, vol. II, p.50.
Levy-Rubin, Continuatio, pp.251 (text), 104 (trans.).
Ibn al-Murajja4, Fada4il, p. 56 carries an exceptional version: $ +M; -%) 7
" 46 - we shall not raise our voices in the presence of Muslims,
with no reference to church services or readings.
Xenophon, Cyropaedia, viii.1.2934, trans. Miller, p.321.
al-Masu4d3, Muru4j al-dhahab, vol. I, p.219.
Pseudo-al-Ja4hiz, Ta4j, pp.11718.
Ibid., p.64.
Al-Ghaza4l3, al-Adab f al-dn, Cairo 1976 p.42.
The word that appears here in Samaritan script is clearly fanka; I have not
managed to discover its meaning.
Levy-Rubin, Continuatio, pp.242 (text), 91 (trans.).
Ibid., p.251 (text), p.104 (trans.).
Boyce, trans., Letter of Tansar, pp.489.
See Metz, Renaissance, pp.3912; A. S. Tritton, Muslim Funeral Customs,
Bulletin of the School of Oriental and African Studies, 9 (1938), pp.653
61; H. Algar, Burial, Encyclopedia Iranica, vol. IV, pp.5635, see p.564.
Kister, La4 tashabbahu4, pp.32153.
Pseudo-al-Ja4hiz, Ta4j, pp.1323.
Kalfon-Stillman, Arab Dress, p. 126; Al3 b. Yu4suf al-Qift3, Tarkh

al-Hukama4, ed. J. Lippert, Leipzig 1903, p.104.
See Kalfon-Stillman, Arab Dress, p.107. See there also the reference to Ibn
Abdu4n, who says to the muhtasibu4n that no individual among the tax collectors, constables, Jews and Christians should be allowed to dress in the
clothing of the nobility, nor that of a jurist, nor that of a person of good
standing; rather, they should be abhorrent and shunned:ibid., pp.1078.
Glossary
adha4n the call to prayer

ahl al-dhimma protected minorities under Muslim rule
ahl al-kita4b (people of the book) fellow monotheists
ahd (pl. uhu4d) covenant
ama4n safety, protection, safe-conduct, an assurance of safety
and security
amicitia (Lat.) friendship
am r (pl. umara4) ruler; military commander
Am r al-Mumin n (Commander of the Faithful) the caliph
amsa4r al-muslim n cities founded by the Arabs after conquest
anwatan conquest by force without a formal surrender agreement
aqd covenant
asa4wira4 (Pers./Ar.) cavalry
baqt agreement of safety
b a synagogue/church
deditio (Lat.) conquest by surrender
dhimma assurance of protection
dhimm member of a protected minority under Muslim rule
dihqa4n (Pers.) local notable in Sasanian Persia
fides (Lat.) honesty, confidence, good faith
foederati (Lat.) allies
foedus (Lat.; pl. foedera) treaty
ghiya4r restrictions placed on non-Muslims to differentiate them from
Muslims
hadd (pl. hudu4d) Quranic punishment for certain crimes
had th tradition; saying of the Prophet and/or his Companions
235
236
Glossary
hajj pilgrimage to Mecca

hamya4na4 belt worn by Jews in the Sasanian empire
homologia (Gk.) conquest by surrender; pact, agreement
ima4m prayer leader
imam ruler
ima4ma (pl. ama4im) turban
isna4d chain of transmission
jiwa4r agreement of protection
jizya tax payable by non-Muslims
jund (pl. ajna4d) administrative sub-district
ka4fir infidel, non-Muslim
kan sa synagogue/church
khila4f public differentiating signs between Muslims and non-Muslims
khutba Friday sermon
kita4b written compact
kita4b ama4n document listing the conditions that were agreed upon
in a sulh
mahd righteous leader who will rule before the end of the world
marzuba4n (also marzba4n/marzpan; Pers./Ar.) senior officer in the
Sasanian army; local Iranian official under Muslim rule
mawla4 (pl. mawa4l ) client; non-Arab Muslim
melta4 (Syr., word) honesty, confidence, good faith; agreement
melta4 d-qya4ma4 (Syr., word of covenant) surrender agreement
metoikos (Gr.; pl. metoikoi) resident alien
mintaqa (pl. mana4tiq) girdle
misr (pl. amsa4r) fortified city
mushriku4n polytheists; non-Muslims
na4qu4s (pl. nawa4q s) the clapper or gong used in the East for the
purpose of calling to prayer
pactio/pactum (Lat.; pakton (pl. pakta), Gr.) pact, agreement
pistis (Gr.; pl. ta pista) honesty, confidence, good faith; guarantee
qaba4 (Pers. kapa4h) luxurious robe
qa4d judge
Ra4shidu4n the four Rightly Guided Caliphs
shar a Islamic law
shirk polytheism
s ra biography
spa4badh general in the Sasanian army
sulh surrender agreement
sulhan conquest through surrender, conquered by means of a sulh
Glossary
sunna customs based on the practice of the Prophet
tabaqa (stratum) social class; non-Muslims
tafs r Quranic exegesis
tira4z an embroidered border on a garment
wal (pl. awliya4) government official; holy man, mystic
waz r vizier
zunna4r (pl. zana4n r) girdle, belt
237
Bibliography
Primary Sources
Abbas, I., Ahd Ardash3r, Beirut 1967.
Abd al-Razza4q b. Hamma4m al-Sana4n3, Abu4 Bakr, al-Musannaf, Beirut 19702.
Abu4 Bakr al-Khalla4l, Ahl al-milal wa-al-ridda wa-al-zana4dika wa-tarq al-sala4t
wa-al-farad min kita4b al-ja4mi lil-Khalla4l Ab Bakr Ahmad b. Ha4ru4n b.
Yazd al-Baghdad al-Hanbal, ed. Ibra4h3m b. Hamd b. Sulta4n, Riyadh 1996.
Abu4 al-Fida4 Isma43l b. Kath3r, al-Sra al-nabawiyya, vol. I, Beirut 1976; trans.
Trevor Le Gassick as The Life of the Prophet Muhammad, Reading 1998.
Abu4 Ubayd al-Qa4sim b. Salla4m, Kita4b al-amwa4l, Cairo 1353 H.
Kita4b al-amwa4l, ed. M. Ima4ra, Beirut 1989.
Abu4 Yu4suf Yaqu4b b. Ibra4h3m, Kita4b al-khara4j, Cairo 1352 H.
Abu4 Zakariya4 Yahya4 b. Sharaf b. Mar3, Tahdhb al-asma4, Beirut 1996.
Agapius b. Mahbu4b, Kita4b al-unwa4n, ed. and trans. Alexandre Vasiliev, Patrologia
Orientalis, 8/3, Turnhout 1971.
Ammianus Marcellinus, Ammianus Marcellinus, Latin text with Eng. trans.
J. C. Rolfe, Loeb Classical Library, London 19502.
Ar3b b. Sad al-Qurtub3, Silat tarkh al-Tabar, ed. M. J. de Goeje, Leiden 1965.
Artsruni, Thomas, History of the House of Artsrunik, ed. and trans. Robert W.
Thomson, Detroit 1985.
al-Asqala4n3, Ibn Hajar, Lisa4n al-miza4n, Beirut 1987.
al-Azd3, Abu4 Zakar3ya4 Yaz3d b. Muhammad, Tarkh al-mawsil, ed. A. Hab3ba,
Cairo 1967.
al-Azd3, Muhammad b. Abdalla4h, Tarkh futu4h al-sha4m, Cairo 1970.
Back, M., Die sassanidischen Staatsinschriften, Tehran and Leiden 1978.
al-Bala4dhur3, Ahmad b. Yahya4, Ansa4b al-ashra4f, ed. M. Schloessinger, Jerusalem
1971.
Kita4b Futu4h al-Bulda4n, ed. M. J. de Goeje, Leiden 1866; trans. P. K. Hitti as The
Origins of the Islamic State, New York 1916.
Bar Hebraeus, Chronicon Syriacum, ed. A. Fiat, Paris 1890; trans. E. A. W. Budge
as The Chronography of Abul
4 Faraj known as Bar-Hebraeus, Oxford 1932.
239
240
Bibliography
Bayhaq3, Abu4l Fazl, T a4rikh Masu4d, ed. Al3 Akbar Fayya4z, Mashhad
1350/1971.
al-Bayhaq3, Ahmad b. Husayn, Sunan al-Bayhaq al-kubra4, Mecca 1994.
Bengston, H., Staatsvertrge des Altertums, Kommission fr Alte Geschichte und
Epigraphik des Deutschen Archologisch Instituts, Munich 19609.
al-Bru
3 n
4 ,3 al-Atha
" r4 al-baqiya
4
an al-qurun
4 al-khaliya
4 , ed. E. C. Sachau, Leipzig 1879;
trans. E. C. Sachau as The Chronology of Ancient Nations, Leipzig 1897.
Blockley, R. C., The Classicising Historians of the Later Roman Empire, vol. II,
Liverpool 1983.
Blockley, R. C., ed. and trans., The History of Menander the Guardian, Liverpool
1985.
Bosworth, C. E., annot. and trans., The History of al-Tabar, vol. V:The Sa4sa4nids,
the Byzantines, the Lakhmids, and Yemen, Albany 1999.
The History of al-Tabar, vol. XXXIII: Storm and Stress along the Northern
Frontiers of the Abba4sid Caliphate, Albany 1999.
Boyce, M., trans., The Letter of Tansar, Rome 1968.
Brinner, W. M., annot. and trans., The History of al-Tabar, vol. III:The Children
of Israel, Albany 1985.
al-Bukha4r3, Muhammad b. Isma43l, Kita4b al-ja4mi al-sahh, ed. M. L. Krehl and
T.W. Juynboll, Leiden 18621908.
Chabot, J. B., ed., Chronique de Michel le Syrien, Paris 18991910 (repr. Brussels
1963).
Charles, R. H., trans., The Chronicle of John (c. 690) Coptic Bishop of Nikiu,
London 1916.
Chronicon ad Annum Domini 1234 pertinens, ed. and trans. J. B. Chabot, Louvain
191620; vol. I trans. J. B. Chabot, 1937; vol. II trans. A. Abouna, 1974.
Chron. Zuqnin (Incerti Auctoris chronicon anonymum pseudo-Dionysianum
vulgo dictum), vol. II, ed. J. B. Chabot, Louvain 1933; French trans. R.
Hespel as Chron. Zuqnin (Incerti Auctoris chronicon anonymum pseudoDionysianum vulgo dictum), Louvain 1989.
Chronicon Paschale 284628 AD, trans. L. M. Whitby and M. Whitby, Liverpool
1989.
al-Dhahab3, Muhammad b. Ahmad, al-Ibar f khabar man ghabara, Kuwait
19606.
Dittenberger, W., Sylloge Inscriptionum Graecorum, 3rd edn, Leipzig 191517.
Doctrina Iacobi Nuper Baptisati in Gilbert Dagron and Vincent Droche, Juifs
et chrtiens dans lOrient du VIIe sicle, Travaux et Mmoires, 11 (1991),
pp. 47229.
Elias of Nisibis, Eliae Metropolitae Nisibeni, ed. E. W. Brooks, vol. I (text),
Louvain 1954.
Eutropius, Eutropi Breviarium ab urbe condita, ed. F. Ruehl, Leipzig 1919;
trans., intro., and commentary H. W. Bird as The Breviarium ab urbe condita of Eutropius, the Right Honourable Secretary of State for General
Petitions: Dedicated to Lord Valens, Gothicus Maximus & Perpetual
Emperor, Liverpool 1993.
Eutychius (Sa3d b. Bitr3q), Kita4b al-tarkh, ed. L. Cheikho, Beirut 1905.
Fagnan, E., Livre de limpt foncier, Paris 1921.
Bibliography
241
Festugire, A. J. and B. Grillet, eds., Sozomne histoire ecclsiastique, VVI, Paris

2005.
Firdaus3, Abu4 al-Qa4sim, Sha4h-na4ma, trans. al-Fath b. Al3 al-Bunda4r3, ed. Abd
al-Waha4b Aza4m, Cairo 1350/1932.
Sha4h-na4ma, ed. E. Bertel et al., Moscow 196671.
Sha4h-na4ma, ed. Djala4l Kha4ligh3-Mutlagh, New York and Costa Mesa
19882008.
Firdaws3, Sha4hna4mah, ed. and trans. J. Mohl, Tehran 19336.
Flusin, B., Saint Anastase le Perse et lhistoire de la Palestine au dbut du viie sicle
(text, French trans., and commentary), Paris 1992.
Fragmenta Historicorum Graecorum, ed. C. Mller, Paris 1928.
Friedmann, Y., annot. and trans., The History of al-Tabar, vol. XII:The Battle of
al-Qa4disiyyah and the Conquest of Syria and Palestine, Albany 1992.
Gnalogie de la famille de saint Grgoire et vie de saint Nerses, Venice 1853
(Armenian).
al-Ghaza4l3, al-Adab f al-dn, Cairo 1976.
Hamilton, F. J. and Brooks, E. W., The Syriac Chronicle known as that of
Zachariah of Mitylene, London 1899.
Hila4l b. al-Muhassin al-Sa4b3, Abu4 al-Husayn, Rusu4m da4r al-khila4fa, ed. M.
Awwa4d, Baghdad 1964; trans. E. Salem as The Rules and Regulations of the
Abba4sid Court, Beirut 1977.
Histoire Nestorienne (Chronique de Sert), ed. and trans. Monsignor A. Scher,
Patrologia Orientalis, 13/4, Turnhout 1973.
Hoyland, R. G., Theophilus of Edessas Chronicle (590750s) and the Circulation
of Knowledge in Late Antiquity and Early Islam, Translated Texts for
Historians, forthcoming (2011).
Humbach, H. and Skjaerv, O., The Sasanian Inscription of Paikuli, Wiesbaden
197883.
Humphreys, R. S., annot. and trans., The History of al-Tabar, vol. XV: The
Crisis of the Early Caliphate, Albany 1985.
Huyse, P., Die dreisprachige Inschrift a4buhrs I. an der Ka4ba-i-Zaradut, Corpus
Inscriptionum Iranicarum III:Pahlavi Inscriptions, vol. I, London 1999.
Ibn Abd al-Hakam, Abu4 al-Qa4sim Abd al-Rahma4n, Futu4h misr, ed. C. C. Torrey,
New Haven 1922.
Sira4t Umar b. Abd al-Azz, Cairo 1994.
Ibn Abd Rabbih, Abu4 Umar Ahmad b. Muhammad, al-Iqd al-fard, Cairo 1321.
Ibn Asa4kir, Abu4 al-Qa4sim Al3 b. al-Hasan, Tarkh madnat dimashq, Beirut
19952000.
Ibn Ab3 Shayba, Abdalla4h Muhammad, Musannaf Ibn Ab Shayba al-Ku4f,
Riyadh 1409 H.
Ibn al-Ath3r, Izz al-D3n Al3 b. Muhammad, al-Ka4mil f al-tarkh, Beirut 1965.
Ibn al-Balkh3, Farsnamah, ed. G. Lestrange and R. A. Nicholson, London 1921.
Ibn Battu4ta, Rihlat Ibn Battu4ta, Cairo 1957; trans. H. A. R. Gibb as The Travels
of Ibn Battuta, AD 13251354, with revisions and notes from the Arabic
text, ed. C. Defremey and B. R. Sanguinetti, Cambridge 19582000.
Ibn al-Jawz3, Abd al-Rahma4n b. Al3 b. Muhammad, al-Muntazam f tarkh
al-mulu4k wa-al-umam, Beirut and London 1992.
242
Bibliography
Ibn al-Murajja4, Abu4 al-Maa4l3 al-Musharraf b. al-Murajja4 b. Ibra4h3m al-Maqdis3,

Fada4il bayt al-maqdis wa-al-khall wa-fada4il al-sha4m, ed. O. Livne-Kafr3,
Shfaram 1995.
Ibn Hazm, Al3 b. Ahmad, al-Muhalla4, ed. Ahmad Muhammad Shakir, Beirut
1980.
Ibn Hisha4m, Kita4b srat rasu4l Alla4h, ed. F. Wstenfeld, Gttingen 1858.
al-Sra al-nabawiyya, vol. IV, Beirut 1411 H [1990].
Ibn Kath3r, Tafsr Ibn Kathr, Beirut 1401 H.
Ibn Khaldu4n, Abd al-Rahma4n, al-Muqaddima lil-alla4ma Ibn Khaldu4n, Beirut
n.d.; trans. F. Rosenthal as The Muqaddimah, London 1967.
Ibn Naqqa4sh, Fetoua sur la condition des dhimm3s, ed. and trans. M. Belin,
Journal Asiatique, series 4, 18 (1851), pp. 417516, 19 (1852), pp. 97140.
Ibn Qayyim al-Jawziyya, Ahka4m ahl al-dhimma, ed. S. al-Sa4lih, Damascus
1961.
Ibn Taghr3 Bird3, Abu4 al-Maha4sin Jama4l al-D3n Yu4suf, al-Nuju4m al-za4hira f
mulu4k misr wa-al-qa4hira, ed. Gautier H. A. Juynboll, Leiden 1861.
al-Nuju4m al-za4hira f mulu4k misr wa-al-qa4hira, Cairo 196371.
Ibn Taymiyya, Abu4 al-Abba4s Taq3 al-D3n Ahmad b. Abd al-Hal3m, Iqtida4 al-sira4t
al-mustaqm li-mukha4lafat asha4b al-jahm, Cairo 1369 H.
Ibn Zabr, Abu4 Muhammad Abd Allah b. Ahmad, Juz fh shuru4t al-nasa4ra4, in
Cohen, Pact of Umar, pp. 13157.
al-Ja4hiz, Kita4b al-baya4n wa-al-tabyn, Cairo 1956.
al-Jahshiya4r3, Kita4b al-wuzara4 wa-al-kutta4b, Cairo 1938.
Johnson, A. C., P. Robinson Coleman-Norton, and F. C. Bourne, Ancient Roman
Statutes, Austin 1961.
Justinian, Iustiniani Novellae, ed. Z. A. Lingenthal, Leipzig 1881.
Juynboll, H. A., annot. and trans., The History of al-Tabar, vol. XIII: The
Conquest of Iraq, Southwestern Persia, and Egypt, Albany 1989.
Kramer, J. L., annot and trans., The History of al-Tabar, vol. XXXIV:Incipient
Decline, Albany 1989.
Lees, E. W. N., comp., The Fotooh al-Sham being an Account of the Moslim
Conquests in Syria, by Aboo Ismaail Mohammad bin Abd Allah, al-Azdi
al-Bacri, Calcutta 1854.
Levy-Rubin, M., Continuatio of the Samaritan Chronicle of Abu4 l-Fath al-Sa4mir
al-Danaf, text and trans., Studies in Late Antiquity and Early Islam,
Princeton 2002.
Lewis, C. T., and C. Short, A Latin Dictionary, Oxford 1879.
Linder, A., The Jews in Roman Imperial Legislation, Detroit and Jerusalem
1987.
Malalas, John, Ioanni Malalae Chronographia, ed. I. Thurn, Berlin 2000; trans.
E. Jeffreys, M. Jeffreys, and R. Scott as The Chronicle of John Malalas,
Melbourne 1986.
Mango, C., Nikephoros, Patriarch of Constantinople: Short History, Text,
Translation and Commentary, Dumbarton Oaks Papers, 10 (1990),
Washington DC.
al-Maqdis3, Abdallah b. Ahmad b. Quda4ma, al-Mughn, Beirut 1405 H.
Bibliography
243
al-Maqr3z3, Taq3 al-D3n Ahmad b. Al3, Ittia4z al-hunafa4 bi-akhba4r al-umu4r

al-fa4timiyn al-khulafa4, Cairo 1967.
Kita4b al-mawai z wa-al-itiba4r bi-dhikr al-khitat wa-al-a4tha4r, Cairo 1270 H.
Marcellinus, The Chronicle of Marcellinus, text, trans. and comm. B. Croke,
Byzantina Australiensia 7 (Australian Association for Byzantine Studies),
Sydney 1995.
al-Masu4d3, Abu4 al-Hasan Al3 b. Husayn, Muru4j al-dhahab wa-maa4din aljawhar, ed. and French trans. Barbier de Meynard and Pavet de Courteille as
Les prairies dor, Paris 186677; ed. and French trans. C. Pellat as Masu4d,
Les prairies dor, Paris 196297.
al-Maward3, Al3 b. Muhammad, Ahka4m al-sulta4nya wa-al-wila4ya4t al-dnya,
Cairo 1994.
al-Mizz3, Jama4l al-D3n Abu4 al-Haja4j Yu4suf, Tahdhb al-kama4l f asma4 al-rija4l,
Beirut 1992.
Moses Khorenatsi, History of the Armenians, trans. R. W. Thomson, Harvard
1980.
al-Muqaddas3, Shams al-D3n Abu4 Abdalla4h, Ahsan al-taqa4sm f marifat
al-aqa4lm, ed. M. J. de Goeje, Leiden 1906; trans. B. A. Collins as The Best
Divisions for Knowledge of the Regions:A Translation of Ahsan al-taqasim
fi marifat al-aqalim, Reading 1994.
Mynors, R. A. B., ed., Panegyrici Latini, Oxford 1964.
Nameh Tansar, ed. M. Minovi, Tehran 1932.
Nathan ben Yehiel, Aruch Completum, ed. A. Kohut, Vienna 1930.
al-Nawaw3, Muhy3 al-D3n Abu4 Zakariyya4 Yahya4 b. Sharaf b. Mur3, Sharh
al-Nawaw ala4 Sahh Muslim, Beirut 1392 H.
Nikephoros, Nicephori Archiepiscopi Constantinopolitani Opuscula historica,
ed. C. de Boor, Leipzig 1880.
al-N3sa4bu4r3, al-Mustadaik ala al-sahhayn, Beirut 1990.
Onasander, On the General, in Aeneas Tacticus, Asclepiodotus, Onasander,
with an English Translation by the Members of the Illinois Greek Club,
Cambridge, MA, and London 1977.
Pacatus, Panegyric to the Emperor Theodosius, trans. C. E. V. Nixon, Liverpool
1987.
Palmer, A., et al., The Seventh Century in the West-Syrian Chronicles, Liverpool
1993.
Pritchard, James B., ed., Ancient Near Eastern Texts Relating to the Old Testament,
trans. and annot. W. F. Albright et al., Princeton 1969.
Procopius, History of the Wars, trans. H. B. Dewing, London 195471.
Pseudo-Dionysius, Pseudo-Dionysius of Tel-Mahre Chronicle, Part III, trans. with
notes and introd. W. Witakowski, Translated Texts for Historians, Liverpool
1996.
Pseudo-al-Ja4hiz, Kita4b al-ta4j, ed. A. Z. Pacha, Cairo 1914; French trans. C. Pellat
as Le livre de la couronne attribut Gahiz, Paris 1954.
Pseudo-Joshua, Incerti auctoris Chronicon ps.-Dionysianum vulgo dictum I, ed.
and trans. J. B. Chabot, Paris 1927 (text), Louvain 1949 (Latin trans.); trans.
and annot. F. R. Trombley and J. W. Watt as The Chronicle of Pseudo-Joshua
the Stylite, Liverpool 2000.
244
Bibliography
al-Qalqashand3, Ahmad b. Al3, Subh al-asha4 f sina4at al-insha4, Cairo 1964.

al-Qurtub3, Tafsr al-Qurtub, Cairo 1372 H.
al-Ra4z3, Ibn Ab3 Ha4tim, Kita4b al-jarh wa-al-tadl, Hyderabad 1952.
Safwat, Ahmad Za4k3, Jamharat rasa4il al-arab, Cairo 1937.
al-Sarakhs3, Muhammad b. Ahmad, Sharh kita4b al-siyar al-kabr, ed. Abd al-Az3z
Ahmad, n.d., n.p.
Scheiber, A., A New Fragment of the Life of Obadjah, the Norman Proselyte,
Kirjath Sepher, 30 (19545), pp. 938.
Sebeos, The Armenian History Attributed to Sebeos, trans. with notes, R. W.
Thomson, historical commentary James Howard-Johnston with Tim
Greenwood, Liverpool 1999.
Severus b. al-Muqaffa, History of the Patriarchs of the Egyptian Church by
Sawrus ibn al-Muqaffa4, ed. Y. Abd al-Mas3h and O. H. E. Burmester, Cairo
1943.
al-Sha4fi3, Abu4 Abdalla4h Muhammad b. Idr3s, Kita4b al-umm, Beirut 1993.
al-Shayba4n3, Muhammad b. al-Hasan, The Islamic Law of Nations:Shayba4ns
Siya4r, trans. M. Khadduri, Baltimore 1966.
Kita4b al-mabsu4t, ed. Abu4 al-Wafa4 al-Afgha4n3, Karachi, n.d.
al-Siya4r, ed. M. Kha4ddu4r3, Beirut 1975.
Smith, G. R., annot. and trans., The History of al-Tabar, vol. XIV:The Conquest
of Iran, Albany 1994.
Socrates of Constantinople, Historia Ecclesiastica: Socrate de Constantinople
Histoire ecclsiastique, ed. G. C. Hansen, trans. Pierre Prichon and Pierre
Maraval, introd. and notes Pierre Maraval, Paris 20047.
Stern, M., Greek and Latin Authors on Jews and Judaism, Jerusalem 1974.
al-Tabar3, Muhammad b. Jar3r, Kita4b ikhtila4f al-fuqaha4, ed. J. Schacht, Leiden
1933.
Tafsr al-Qura4n, Beirut 1405 H.
Tarkh al-rusul wa-al-mulu4k, ed. M. J. de Goeje et al., Leiden 18791901.
al-Thaa4lib3, Abu4 Mansu4r Abd al-Malik b. Muhammad b. Isma43l, Ghurar akhba4r
mulu4k al-Furs:Histoire des rois des Perses, ed. and trans. H. Zotenberg, Paris
1900.
Theophanes, Theophanis Chronographia, ed. C. De Boor, Leipzig 18835; trans.
C. Mango and R. Scott as The Chronicle of Theophanes Confessor:Byzantine
and Near-Eastern History AD 284813, Oxford 1997.
Theophylactus, Theophylacti Simocattae Historiae, ed. C. De Boor, re-ed.
P.Wirth, Stuttgart 1972; trans. M. Whitby and M. Whitby as The History of
Theophylact Simocatta, Oxford 1986.
al-Tirmidh3, Muhammad b. Isa4, Sunan, Beirut 1983.
al-Turtu4sh3, Sira4j al-mulu4k, Cairo 1935.
al-Tu4s3, Hasan b. Al3 (= Niza4m al-Mulk), Siyar al-mulu4k, trans. H. Darke, London
1978.
al-Washsha4, Abu4 al-Tayyib Muhammad, Kita4b al-muwashsha4 aw al-zarf
wa-al-zurafa4, Cairo 1953.
Xenophon, Cyropaedia, trans. W. Miller, Loeb Classical Library, Cambridge, MA,
and London 1979.
al-Yaqu4b3, Tarkh, Leiden 1882.
Yahya4 b. A"dam al-Qurash3, Kita4b al-khara4j, ed. T. W. Juynboll, Leiden 1896.
Bibliography
245
Zachariah of Mitylene, Historia Ecclesiastica Zachariae Rhetori Vulgo Adscripta,

ed. E. W. Brooks, Corpus Scriptorum Christianorum Orientalium, Scriptores
Syri 39, Louvain 1921 (repr. 1953).
Zosimus, Thesaurus Zosimi, Historia Nova, ed. Bastien Kindt, Franoise
van Haeperen and CENTAL, Corpus Christianorum, Thesaurus Patrum
Graecorum, Louvain 2008; trans. R. T. Ridley as Zosimus: New History,
Sydney 1982.
Secondary Sources
Adontz, N., Armenia in the Period of Justinian, trans. with partial revisions and
bibliography N. G. Garsoian, Lisbon 1970.
Altheim, F. and R. Stiehl, Die Araber in der alten Welt, Berlin 1965.
Andreescu-Treadgold, I., and W. Treadgold, Procopius and the Imperial Panels of
San Vitale, Art Bulletin, 99 (1997), pp. 70823.
Bachrach, J. L., Palestine in the Policies of Tulunid and Ikhshidid Governors of
Egypt, AH 254358/868969, in A. Cohen and G. Baer, eds., Egypt and
Palestine: A Millennium of Association 8681948, Jerusalem 1984, pp.
5165.
Badian, E., Deditio, in Brills New Pauly, Antiquity volumes, ed. Cancik and
Schneider.
Hegemony and Independence, Prolegomena to a Study of the Relations of
Rome and the Hellenistic States in the Second Century BC, in J. Harmatta,
ed., Proceedings of the VII Congress of the International Federation of the
Societies of Classical Studies, Budapest 1984, pp. 397414.
Bagnall, R. S., and P. Derow, Greek Historical Documents:The Hellenistic Period,
Ann Arbor 1981.
Baronowski, Donald W., The Status of the Greek Cities of Asia Minor after 190
BC, Hermes, 119 (1991), pp. 45063.
Barthold, W. W., Caliph Umar II and the Conflicting Reports on his Personality,
Islamic Quarterly, 15 (1971), pp. 6995.
Beer, M., The Babylonian Amoraim: Aspects of Economic Life, Ramat Gan
1982.
The Babylonian Exilarchate in the Arsacid and Sasanian Periods, Jerusalem
1970.
Beihammer, A. D., Quellenkritische Untersuchungen zu den gyptischen
Kapitulationvertrgen der Jahre 640646, Vienna 2000.
Ben-Shammai, H., The Idea of Election in Early Islam, in S. Almog and M. Heyd,
eds., Chosen People, Elect Nation and Universal Mission, Jerusalem 1991,
pp. 14777 (Hebrew).
Benveniste, E., Les classes sociales dans la tradition avestique, Journal Asiatique,
221 (1932), pp. 11734.
de Biberstein Kazimirski, A., Dictionnaire ArabeFranaise, Paris 1860.
Blockley, R., East Roman Foreign Policy, Leeds 1992.
Borrut, A., Entre tradition et histoire:gense et diffusion de limage de Umar II,
Mlanges de luniversit Saint-Joseph, 58 (2005), pp. 32978.
Bosworth, A. B., Conquest and Empire: The Reign of Alexander the Great,
Cambridge 1988.
246
Bibliography
Bosworth, C. E., The Concept of Dhimma in Early Islam, in B. Braude and B.

Lewis, eds., Christians and Jews in the Ottoman Empire, New York 1982,
pp. 3751.
Bosworth, E., The Heritage of Rulership in Early Islamic Iran and the Search for
Dynastic Connections with the Past, Iran, 11 (1973), pp. 5162.
Bowersock, G. W., The GreekNabatean Bilingual Inscription at Ruwwa4fa,
Saudi-Arabia, in J. Bingen et al., eds., Le monde grec, pense, littrature,
documents:hommages Claire Praux, Brussels 1975, pp. 51417.
Mavia, Queen of the Saracens, in W. Eck et al., eds., Studien zur antiken
Sozialgeschichte: Festschrift Friedrich Vittinghoff, Cologne and Vienna
1980, pp. 47795.
Boyce, M., History of Zoroastrianism, Leiden 1982.
Braun, J., Die liturgische Gewandung im Occident und Orient, Darmstadt 1964.
Bravmann, M., The Ancient Arab Background of the Qura4nic Concept al-Jizyatu
an Yadin, Arabica, 13 (1966), pp. 30714.
Brills New Pauly:Encyclopedia of the Ancient World, Antiquity vols., ed. Hubert
Cancik and Helmuth Schneider, English edn Christine F. Salazar; Classical
Tradition vols. ed. Manfred Landfester, English edn Francis G. Gentry,
Leiden and Boston 2007 (and online 2010).
Brockelmann, C., Lexicon Syriacum, Gttingen 1928.
Broisus, M., Alexander and the Persians, in J. Roisman, ed., Brills Companion
to Alexander the Great, Leiden 2003, pp. 16993.
Bromberg, C. A., Sasanian Royal Emblems in the Northern Caucasus, in G.
Gnoli and A. Panaino, eds., Proceedings of the first European Conference of
Iranian Studies, part 1, Rome 1990, pp. 118.
Calder, N., Studies in Early Muslim Jurisprudence, Oxford 1993.
Canard, M., Le crmonial ftimite et le crmonial Byzantine, Byzantion, 21
(1951), pp. 355420.
Cartledge, P., Alexander the Great: The Hunt for a New Past, Woodstock and
New York 2004.
The Cambridge History of Iran,:see Yarshater.
Chaumont, M. L., Lordre des prsances la cours des Arsacides dArmenie,
Journal Asiatique, 254 (1966), pp. 47197.
Chitty, D. J., The Desert a City:An Introduction to the Study of Egyptian and
Palestinian Monasticism under the Christian Empire, Oxford and New York
1966.
Choksy, J., Conflict and Cooperation:Zoroastrian Subalterns and Muslim Elites
in Medieval Iranian Society, New York 1977.
Christensen, A., LIran sous les Sasanids, Osnabrck 1971.
Chrysos, E., Byzantine Diplomacy, AD 300800:Means and Ends, in J. Shepard
and S. Franklin, eds., Byzantine Diplomacy, Variorum, Aldershot 1992,
pp.2539.
Clauson, G., An Etymological Dictionary of Pre-Thirteenth-Century Turkish,
Oxford 1972.
Clerc, M., Les mtques athniens; tude sur la condition lgale, la situation
morale et le rle social et conomique des trangers domicilis Athnes,
Paris 1893 (repr. 1979).
Bibliography
247
Cobb, P. M., Umar b. Abd al-Az3z, EI2, vol. X, pp. 8212.

Cohen, M. R., Islamic Attitudes and Policies, in The Cambridge History of
Judaism: Jews in the Medieval Islamic World, vol. V: R. Chazan and M.
Rustow, eds., The Medieval Period, Seventh through Fifteenth Centuries,
forthcoming.
The Neo-Lachrymose Conception of Jewish Arab History, Tikkun, 1/2 (May
June 1991), pp. 1116, and responses in Tikkun, 1/3 (JulyAugust 1991),
pp. 967.
Under Crescent and Cross:The Jews in the Middle Ages, Princeton 1994.
What was the Pact of Umar? A Literary-Historical Study, Jerusalem Studies in
Arabic and Islam, 23 (1999), pp. 10057.
Conrad, L. I., The Conquest of Arwa4d: A Source-Critical Study in the
Historiography of the Early Medieval Near-East, in A. Cameron and
L. I. Conrad, eds., The Byzantine and Early Islamic Near East, vol. I, Studies
in Late Antiquity and Early Islam 1, Princeton 1992, pp. 317401.
Syriac Perspectives on Bila4d al-Sha4m during the Abbasid Period, in M. A.
Bakhit and R. Schick, eds., Bila4d al-Sha4m during the Abbasid Period, 132
AH/750 AD451 AH/1059 AD: Proceedings of the Fifth International
Conference, vol. I, Amman 1991, pp. 144.
Theophanes and the Arabic Historical Tradition: Some Indications of Intercultural Transmission, Byzantinische Forschungen, 15 (1990), pp. 144.
Cook, M., Early Muslim Dogma, Cambridge 1981.
Corbet, E. K., The Life and Works of Ahmad ibn Tu4lu4n, Journal of the Royal
Asiatic Society (1891), pp. 52762.
Croke, B., Justinians Constantinople, in M. Maas, ed., The Cambridge
Companion to the Age of Justinian, Cambridge 2005, pp. 6086.
Crone, P., Mawla4, EI2, vol. VI, pp. 87482.
The Pay of Client Soldiers in the Umayyad Period, Der Islam, 80 (2003),
pp. 284300.
Roman, Provincial and Islamic Law, Cambridge 1987.
Crone, P. and M. Cook, Hagarism:The Making of the Islamic World, Cambridge
1977.
Crone, P. and M. Hinds, Gods Caliph, Cambridge 1986.
Crone, P. and F. Zimmermann, The Epistle of Sa4lim b. Dhawka4n, Oxford
2001.
Curtis, V. S., and S. Stewart, The Sasanian Era:The Idea of Iran, London 2008.
Dahlheim, W., Struktur und Entwicklung des rmischen Vlkerrechts im dritten
und zweiten jahrhundert v. Chr., Munich 1968.
Daryaee, T., Sasanian Iran (224651 CE): Portrait of a Late Antique Empire,
Costa Mesa 2008.
Dennett, D. C., Conversion and the Poll Tax in Early Islam, Cambridge, MA,
1950.
Dignas, B., and E. Winter, Rome and Persia in Late Antiquity, Cambridge
2007.
Dozy, R. P. A., Dictionnaire dtaill des noms des vtements chez les Arabes,
Amsterdam 1845.
Supplemnt aux dictionnaires arabes, Leiden 1881 (repr. 1991).
248
Bibliography
Dutton, Y., Review of N. Calders Studies in Early Muslim Jurisprudence, Journal

of Islamic Studies, 5 (1994), pp. 1028.
Eckstein, A. M., Glabrio and the Aetolians: A Note on Deditio, Transactions
and Proceedings of the American Philological Association, 125 (1995),
pp. 27189.
Pharos and the Question of Roman Treaties of Alliance in the Greek East in
the Third Century BCE, Classical Philology, 94 (1999), pp. 395418.
Elman, Y., Middle Persian Culture and Babylonian Sages: Accomodation and
Resistance in the Shaping of Rabbinic Legal Tradition, in C. E. Fonrobert
and M. S. Jaffee, eds., Cambridge Companion to the Talmud and Rabbinic
Literature, Cambridge 2007, pp. 165197.
Ettinghausen, R., and O. Grabar, The Art and Architecture of Islam 6501250,
Harmondsworth 1987.
Fattal, A., Le statut lgal des non-musulmans en pays dIslam, Beirut 1955.
Finley, M. I., Between Slavery and Freedom, Comparative Studies in Society and
History, 6 (1964), pp. 23349.
Forand, P., Early Muslim Relations with Nubia, Der Islam, 48 (1972),
pp.11221.
Fredricksmeyer, E., Alexander and the Kingship of Asia, in A. B. Bosworth and
E. J. Baynham, eds., Alexander the Great in Fact and Fiction, Oxford 2000,
pp. 13666.
Friedmann, Y., Tolerance and Coercion in Islam, Cambridge 2003.
Frye, Richard, Feudalism in Sasanian and Early Islamic Iran, Jerusalem Studies
in Arabic and Islam, 9 (1987), pp. 1318.
The Golden Age of Persia:The Arabs in the East, London 1975.
The Political History of Iran under the Sasanians, in Yarshater, ed., The
Cambridge History of Iran, vol. III(1), pp. 11680.
Gafni, I. M., The Jews of Babylonia in the Talmudic Era:A Social and Cultural
History, Jerusalem 1990.
Gautier, P., Mtques, priques et paroiki:bilans et points dinterrogations, in
R. Lonis, ed., Ltranger dans le monde grec:Actes du colloque organis par
lInstitut dtudes Anciennes Nancy, Mai 1987, Nancy 1988, pp. 2346.
Gergel, T., ed., Alexander the Great: Selected Texts from Arrian, Curtius and
Plutarch, New York 2004.
Gibb, H. A. R., ArabByzantine Relations under the Umayyad Caliphate,
Dumbarton Oaks Papers, 12 (1958), pp. 21933.
The Fiscal Rescript of Umar II, Arabica, 2 (1955), pp. 116.
Gignoux, P., Courts and Courtiers, Encyclopedia Iranica, vol. VI, pp. 35664.
Gil, M., A History of Palestine 6341099, trans. E. Broido, New York 1992.
Glass, Gilding, and Grand Design: Art of Sasanian Iran (224642), exhibition
catalogue, Asia Society, New York 2007.
Goitein, S. D., Did Omar Prohibit the Stay of the Jews in Jerusalem?, repr. in his
Palestinian Jewry in Early Islamic and Crusader Times in the Light of the
Geniza Documents, ed. J. Hacker, Jerusalem 1980, pp. 3642.
A Mediterranean Society: The Jewish Communities of the Arab World as
Portrayed in the Documents of the Cairo Geniza, Berkeley 196793.
Bibliography
249
Goldman, B., The Later Pre-Islamic Riding Costume, Iranica Antiqua, 28 (1993),
pp. 20139.
Goldziher I., Muhammedanische Studien, Halle 188990; ed. and trans.
C. R. Barber and S. M. Stern as Muslim Studies, New York 1977.
Gordon, S., A World of Investiture, in S. Gordon, ed., Robes and Honor:The
Medieval World of Investiture, New York 2001, pp. 122.
Greatrex, G., Rome and Persia at War:502532, Leeds 1998.
Griffith, S. H., Images, Islam and Christian Icons, in P. Canivet and J. P. ReyCoquais, eds., La Syrie de Byzance lIslam VIIeVIIIe sicles: Actes du
Colloque international Lyon Maison de lOrient Mditerranen Paris
Institut du Monde Arabe 1115 Septembre 1990, Damascus 1992, pp. 12138.
Grignaschi, M., Quelques spcimens de la littrature sassanide conservs dans les
bibliothques dIstanbul, Journal Asiatique, 254 (1966), pp. 1135.
Gruen, E. S., Greek Pistis and Roman Fides, Athenaeum, 60 (1982),
pp. 5068.
Gutas, D., Greek Thought, Arabic Culture: The Graeco-Arabic Translation
Movement in Baghdad and Early Abba4sid Society (2nd4th/8th10th
centuries), London and New York 1998.
Gyselen, R., Le gographie administrative de lempire sassanide:les tmoignages
sigillographiques, Paris 1989.
Haldon, J., Economy and Administration: How did the Empire Work?, in
M. Maas, ed., The Cambridge Companion to the Age of Justinian, Cambridge
2005, pp. 2859.
Hassan, Z. M., Ahmad ibn Tu4lu4n, EI2, vol. I, pp. 2789.
Hasson, I., Les mawa4l dans larme musulmane sous les premiers umayyads,
Jerusalem Studies in Arabic and Islam, 14 (1991), pp. 176213.
Hawting, G. R., The First Dynasty of Islam, London and New York 2000.
Heather, P., Empires and Barbarians:The Fall of Rome and the Birth of Europe,
Oxford 2010.
Foedera and Foederati in the Fourth Century, in W. Pohl, ed., Kingdoms of
the Empire:The Integration of Barbarians in Late Antiquity, Leiden 1997,
pp. 5774.
The Goths, Oxford 1996.
Heffening, W., Das islamische Fremdenrecht bis zu den islamische-frnkischen
Staatsvertrgen, Hanover 1925 (repr. Osnabrck 1975).
Hennigen, P. C., The Birth of a Legal Institution:The Formation of the Waqf in
Third-Century AH Hanaf Legal Discourse, Leiden 2004.
Herman, G., The Sasanian Exilarchate in Babylon, Ph.D. thesis, the Hebrew
University, Jerusalem 2005 (Hebrew), forthcoming in English as A Prince
without a Kingdom, Texts and Studies in Ancient Judaism, Tbingen.
Hill, D., The Termination of Hostilities in the Early Arab Conquests, 634656
AD, London 1971.
Hinds, M. and H. Sakkout, A Letter from the Governor of Egypt to the King of
Nubia and Muqurra concerning EgyptianNubian Relations in 141/758, in
W. al-Qa4d3, ed., Studia Arabica and Islamica:Festschrift for Ihsa4n Abba4s,
Beirut 1981, pp. 20929.
250
Bibliography
Hinz, W., Darius und die Perser, Baden-Baden 19769.

Hoyland, R. G., Seeing Islam as Others Saw it, Studies in Late Antiquity and
Early Islam 13, Princeton 1997.
Hoyland, R. G., ed., Muslims and Others in Early Islamic Society, Variorum,
Aldershot 2004.
Hakim, A., Umar b. al-Khatta4b:Some Aspects of his Image as an Ideal Leader in
Early Islamic Traditions, Ph.D. thesis, Tel-Aviv University, 2002 (Hebrew).
Jokisch, B., Islamic Imperial Law, Berlin and New York 2007.
Jones, A. H. M., The Later Roman Empire 284602, Oxford 1964.
Kalfon-Stillman, Y., Arab Dress: A Short History from the Dawn of Islam to
Modern Times, ed. N. A. Stillman, Leiden 2000.
Kehne, Peter, H. Neumann, F. Starke, and H. Beck, International Treaties, in
Brills New Pauly, Antiquity volumes, ed. Cancik and Schneider.
Kehne, P., Treaties, Upholding of, in Brills New Pauly, Antiquity volumes, ed.
Cancik and Schneider.
Kennedy, H., The Prophet and the Age of the Caliphates:The Islamic Near East
from the Sixth to the Eleventh Century, London 1986.
Khazdan, A., Belt, in The Oxford Dictionary of Byzantium, Oxford 2005, vol.I,
p. 280.
Kazhdan, A. P. et al., eds., Oxford Dictionary of Byzantium, New York 1991.
Kister, M., Do Not assimilate Yourselves La4 tashabbahu4, Jerusalem Studies
in Arabic and Islam, 12 (1989), pp. 32153.
Lambton, A. K. S., Continuity and Change in Medieval Persia, Albany 1988.
Theory and Practice in Medieval Persian Government, London 1980.
Landau-Tasseron, E., Alliances among the Arabs, al-Qantara, 26 (2005),
pp.14174.
Alliances in Islam, in M. Bernards and J. Nawas, eds., Patronate and Patronage
in Early and Classical Islam, Leiden 2005, pp. 149.
The Cyclical Reform:A Study of the Mujaddid Tradition, Studia Islamica,
70 (1989), pp. 79113.
Sayf ibn Umar in Medieval and Modern Scholarship, Der Islam, 67 (1990),
pp. 126.
Lane, E. W., ArabicEnglish Lexicon, Cambridge 1984.
Lane-Pool, S., A History of Egypt in the Middle Ages, London 1914.
Lecker, M., The Constitution of Medina, Princeton 2004; see also Appendix
E:The Term Muwa4daa.
A Pre-Islamic Endowment Deed in Arabic Regarding al-Wahda in the Hija4z,
paper presented at the Huitimes Rencontres Sabeenes, Aix-en-Provence
2003, pp. 116; repr. in M. Lecker, People, Tribes and Society in Arabia
Around the Time of Muhammad, Variorum, Aldershot 2005.
Lev, Y., Persecutions and Conversion to Islam in Eleventh Century Egypt, Asian
and African Studies, 22 (1988), pp. 7391.
Levy, R., The Social Structure of Islam, Cambridge 1957.
Levy-Rubin, M., Ahmad ibn Tu4lu4n and his Attitude towards the Protected People
According to a Contemporary Samaritan Chronicle, in Y. Ben Aryeh etal.,eds.,
Chapters in the History of Eretz Israel and its Settlement:Researches Offered
to Y. Ben Porat, Jerusalem 2003, pp. 34560 (Hebrew).
Bibliography
251
Changes in the Settlement Pattern of Palestine Following the Arab Conquest,

in K. Holum and H. Lapin, eds., Shaping the East, forthcoming.
Shuru4t Umar and its Alternatives:The Legal Debate throughout the Eighth
and Ninth Centuries over the Status of the Dhimms, Jerusalem Studies in
Arabic and Islam, 30 (2005), pp. 170206.
Were the Jews Prohibited from Settling in Jerusalem following the Arab
Conquest?: On the Authenticity of al-Tabar
3s Jerusalem Surrender
Agreement, Jerusalem Studies in Arabic and Islam, 36 (2009), pp. 6381.
Levy-Rubin, M., and B. Z. Kedar, A Spanish Source on Mid-Ninth Century
Mar-Saba, in J. Patrich, ed., The Sabaite Heritage in the Orthodox Church
from the Fifth Century to the Present: Monastic Life, Liturgy, Theology,
Literature, Art, Archaeology, Orientalia Lovaniensia Analecta 98, Leuven
2001, pp. 6372.
Lewis, B., Egypt and Syria, in The Cambridge History of Islam, vol. I:P. M. Holt
et al., eds., The Central Islamic Lands, Cambridge 1970, pp. 175230.
Islam from the Prophet Muhammad to the Capture of Constantinople, vol. II,
New York 1974.
The Jews of Islam, London 1984.
Liddell, H. G., and R. Scott, GreekEnglish Lexicon, Oxford 1948.
Linder, A., The Jews in Roman Imperial Legislation, Detroit and Jerusalem
1987.
Lukonin, V. G., Political, Social and Administrative Institutions: Taxes and
Trade, in Yarshater, ed., The Cambridge History of Iran, vol. III(2),
pp.681746.
Maas, M., ed., The Cambridge Companion to the Age of Justinian, Cambridge
2005.
MacMullen, R., Social Mobility and the Theodosian Code, Journal of Roman
Studies, 54 (1964), pp. 4953.
Macuch, M., Ma4daya4n haza4r da4desta4n (Book of a Thousand Judgments),
Pahlavi Law-Book from the late Sasanian Period (First Half of the Seventh
Century), in Encyclopedia Iranica, online.
Das sasanidische Rechtsbuch Ma4takda4n i haza4r Da4tista4n (Teil II), Abhandlungen
fr die Kunde des Morgenlandes XLV, 1, Wiesbaden 1981.
Magnetti, D. L., The Function of the Oath in the Ancient Near Eastern
International Treaty, American Journal of International Law, 72 (1978),
pp. 81529.
Mango, M. M., Status and its Symbols, in C. A. Mango, ed., The Oxford History
of Byzantium, Oxford 2002, pp. 603.
Marlow, L., Hierarchy and Egalitarianism in Islamic Thought, Cambridge 1997.
Marsham, A., Rituals of Islamic Monarchy:Accession and Succession in the First
Muslim Empire, Edinburgh 2009.
Mayerson, P., Mavia, Queen of the Saracens, Israel Exploration Journal, 30
(1980), pp. 12331.
Metz, A., The Renaissance of Islam, trans. Salahuddin Khuda Bakhsh and
D. S. Margoliouth, Patna 1937.
Miller, D. A., Byzantine Treaties and Treaty-Making:5001025 AD, ByzantinoSlavica, 32 (1971), pp. 5676.
252
Bibliography
Miller, D. E., From Catalogue to Codes to Canon: The Rise of the Petition
to Umar among Legal Traditions Governing non-Muslims in Medieval
Islamicate Societies, Ph.D. thesis, University of Missouri-Kansas City, 2000.
Morony, M. G., Iraq after the Muslim Conquest, Princeton 1984.
Motzki, H., The Prophet and the Cat:On Dating Ma4liks Muwatta and Legal
Traditions, Jerusalem Studies in Arabic and Islam, 22 (1998), pp. 1883.
Murad, Hasan Q., Umar IIs View of the Patriarchal Caliphs, Hamdard Islamicus,
10 (1987), pp. 3156.
Was Umar II A True Umayyad?, Islamic Studies, 24 (1985), pp. 32548.
Neusner, J., A History of the Jews in Babylonia, vol. V: Later Sasanian Times,
Leiden 1970.
Nrr, D., Aspekte des rmischen Vlkerrechts die Bronztafel Alcantara, Munich
1989.
Noth, A., Abgrenzungsprobleme zwischen Muslimen und nicht-Muslimen:Die
Bedingungen Umars (a-uru4t al-umariyya) unter einem anderen Aspekt
gelesen, Jerusalem Studies in Arabic and Islam, 9 (1987), pp. 290315;
trans. M. Muelhaeusler in Hoyland, Muslims, pp. 10324.
Die literarische berlieferten Vertrge der Eroberungszeit als historische
Quellen fr die Behandlung der unterworfenen Nicht-Muslimen durch
ihre neuen muslimischen Oberherren, in T. Nagel et al., eds., Studien zum
Minderheitenproblem im Islam, vol. I, Bonn 1973, pp. 282314.
Quellenkritische Studien zu Themen, Formen und Tendenzen frhislamischer
Geschichtsberlieferung, Teil I:Themen und Formen, Bonner Orientalische
Studien 25, Bonn 1973.
Zum Verhltnis von kalifaler Zentralgewalt und Provinzen in umayyadischer
Zeit: Sulh-Anwa. Traditionen fr gypten und den Iraq, Die Welt des
Islams, 14 (1973), pp. 15062.
Payne-Smith, J., A Compendious Syriac Dictionary, Oxford 1903.
Perikhanian, A., The Book of a Thousand Judgments, New York 1980.
Iranian Society and Law, in Yarshater, ed., The Cambridge History of Iran,
vol. III(2), pp. 62780.
Phillipson, C., The International Law and Custom of Ancient Greece and Rome,
London 1911.
Pourshariati, P., Decline and Fall of the Sasanian Empire, London 2007.
al-Qa4d3, Wada4d, Madkhal ila4 dira4sat uhu4d al-sulh al-isla4miyya zaman al-futu4h,
in M. A. Bakhit and I. Abbas, eds., Proceedings of the Second Symposium
(4th conference) on the History of Bila4d al-Sha4m during the Early Muslim
Period up to 40 AH/640 AD, vol. II (Arabic), Amman 1987, pp. 193269.
Rajak, T., Was There a Roman Charter for the Jews?, Journal of Roman Studies,
74 (1984), pp. 10723.
Reinhold, M., On Status Symbols, Classical Journal, 64/7 (1969), pp. 3004.
Robinson, Chase F., Empire and Elites after the Muslim Conquest: The
Transformation of Northern Mesopotamia, Cambridge 2000.
Neck Sealing in Early Islam, Journal of the Economic and Social History of
the Orient, 48 (2005), pp. 40141.
Rose, J., Sasanian Splendor:The Appurtenances of Royalty, in S. Gordon, ed.,
Robes and Honor: The Medieval World of Investiture, New York 2001,
pp.3558.
Bibliography
253
Rubin, Z., Eastern Neighbours:Persia and the Sasanian Monarchy (224651),

in Jonathan Shepard, ed., Cambridge History of the Byzantine Empire, vol.I,
Cambridge 2008, pp. 13955.
Mavia, Queen of the Saracens, Cathedra, 47 (1988), pp. 2549 (Hebrew).
Nobility, Monarchy and Legitimation under the Later Sasanians, in John
Haldon and L. I. Conrad, eds., The Byzantine and Early Islamic Near East
VI:Elites Old and New in the Byzantine and Early Islamic Near East, Studies
in Late Antiquity and Early Islam 1, Princeton 2004, pp. 23573.
Rutgers, L. V., Roman Policy towards the Jews:Expulsions from the City of Rome
during the First Century CE, Classical Antiquity, 13 (1994), pp.5674.
Sadan, J., Le mobilier au proche-orient medieval, Leiden 1976.
Schiemann, G., Conubium, in Brills New Pauly, Antiquity vols., ed. Cancik and
Schneider.
Sezgin, F., Geschichte des arabischen Schrifttums, Leiden 196784.
Shahal, Y., Deditio per Pactionem, MA thesis, The Hebrew University of
Jerusalem, November 2005 (Hebrew).
Shahid, I., Byzantium and the Arabs in the Fourth Century, Washington 1984.
Byzantium and the Arabs in the Sixth Century, Washington 1995.
Shaked, S., From Iran to Islam: Notes on Some Themes in Transmission,
Jerusalem Studies in Arabic and Islam, 4 (1984), pp. 3167.
From Iran to Islam:On Symbols of Royalty, Jerusalem Studies in Arabic and
Islam, 7 (1986), pp. 7591.
No Talking during a Meal:Zoroastrian Themes in the Babylonian Talmud, in
C. Bakhos and R. Shayegan, eds., Talmud in its Iranian Context:Proceedings
of the Conference Talmud in its Iranian Context, UCLA Los Angeles 67th
May 2007, Tbingen forthcoming.
Shapira, H., The Deposition of Rabban Gamaliel:Between History and Legend,
Zion, 64 (1999), pp. 538 (Hebrew).
Sinclair, T. A., A History of Greek Political Thought, London 1951.
Sophocles, E. A., Greek Lexicon of the Roman and Byzantine Periods (from BC
146 to AD 1100), Cambridge 1914.
Sourdel, D., Questions de crmonial abba4sside, Rvue des tudes Islamiques,
28 (1960), pp. 12148.
Spaulding, J., Medieval Christian Nubia and the Islamic World:A Reconsideration
of the Baqt Treaty, International Journal of African Historical Studies, 28
(1995), pp. 57794.
Spler, B., Iran in Frh-islamischer Zeit, Wiesbaden 1952.
Stopler, M. W., Enterpreneurs and Empire:The Mura Archive, the Mura Firm
and the Persian Rule in Babylonia, Leiden 1985.
Strauss (Ashtor), E., The Social Isolation of Ahl al-Dhimma, in O. Komls, ed.,
tudes orientales la mmoire de Paul Hirschler, Budapest 1950, pp. 7394.
Tafazzoli, A., Sasanian Society, New York 2000.
Talbot-Rice, D., Art of the Byzantine Era, London 1963.
Talgam, R., The Stylistic Origins of Umayyad Sculpture as Shown in Khirbat
al-Mafjar and Mshatta, Ph.D. thesis, Hebrew University, Jerusalem 1996
(Hebrew, with English abstract).
Trimingham, J. S., Mawwya the First Christian Arab Queen, Theological Review
of the Near-East School of Theology, 1 (1978), pp. 310.
254
Bibliography
Tritton, A. S., The Caliphs and their Non-Muslim Subjects, London 1930.
Muslim Funeral Customs, Bulletin of the School of Oriental and African
Studies, 9 (1938), pp. 65361.
Tyan, E., Institutions du droit public, Paris 1954.
von Grnebaum, G. E., Islam:Essays in the Nature and Growth of a Cultural
Tradition, London 1961.
Ward, S., A Fragment from an Unknown Work by al-Tabar3 on the Tradition
Expel the Jews and the Christians from the Arabian Peninsula (and the
Lands of Islam), Bulletin of the School of Oriental and African Studies, 53
(1990), pp. 40720.
Watson, A., International Law in Archaic Rome, Baltimore and London 1993.
Weinfeld, M., Covenant Terminology in the Ancient Near East and its Influence
on the West, Journal of the American Oriental Society, 93 (1973),
pp.1909.
Westbrook, R., ed., A History of Ancient Near Eastern Law, Leiden and Boston
2003.
Whitehead, D., The Ideology of the Athenian Metic, Cambridge 1977.
Widengren, G., Feudalismus im alten Iran, Cologne and Oplanden 1967.
Quelques rapports entre Juifs et Iraniens lpoque des Parthes, Vetus
Testamentum, Supplements, 4 (Leiden 1957), pp. 197241.
Recherches sur le fodalisme iranien, Orientalia Suecana, 5 (1956), pp.
79182.
Some Remarks on Riding Costume and Articles of Dress among Iranian
Peoples in Antiquity, Arctica:Studia Ethnographica Upsaliensia, 11 (1956),
pp. 22876.
The Status of the Jews in the Sasanian Empire, Iranica Antiqua, 1 (1961),
pp.11762.
Le symbolisme de la ceinture, Iranica Antiqua, 3 (1968), pp. 13355.
Wiesehfer, J., Ancient Persia from 550 BC to 650 AD, London and New York
1996.
Wolfram, H., History of the Goths, Berlin and Los Angeles 1988.
E. Yarshater, ed., The Cambridge History of Iran, vol. III:The Seleucid, Parthian
and Sasanian Periods, Cambridge 1983.
Zakeri, M., Sa4sa4nid Soldiers in Early Muslim Society:The Origins of Ayya4ra4n
and Futuwwa, Wiesbaden 1995.
Zaman, M. Q., Religion and Politics under the Early Abba4sids:The Emergence
of the Proto-Sunn Elite, Leiden 1997.
Zaya4t, H., Sima4t al-nasa4ra4 wa-al-yahu4d f3 al-isla4m: al-sal3b wa-al-zunna4r,
wa-al-ama4ma wa-al-ghiya4r; shuru4t al-umariyya, al-Machriq, 43/2 (1949),
pp. 161252.
Index
aya4n, 129
Abba4d b. Ziya4d, 222
Abba4sid, 69, 75, 97, 102, 133, 135, 142,
145, 150, 153, 159, 179, 223, 229
governor of Egypt 9, 35, 55
Abdalla4h b. Ab3 Sad b. Ab3 Sarh, 8
Abdalla4h b. Ta4hir, 76
Abd al-Malik b. Marwa4n (r. 685705), 76,
100, 102, 133, 163
Abd al-Muttalib b. Ha4shim, 184
Abd al-Rahma4n b. Ghanam (d. 78/697),
171
Abd al-Rahma4n b. Tha4bit b. Thu4ba4n, 101
Abd al-Razza4q al-Sana4n3 (= Abd
al-Razza4q (126211 H)), 63,
64, 65, 67, 69, 75, 76, 89, 90, 91, 101,
149
Abu4 Bakr (r. 6324), 74
Abu4 Bakr al-Khalla4l (d. 923), 60
Abu4 Bakr b. Muhammad b. Amru b.
Hazm, 208
Abu4 l-Fath al-Sa4mir3, 105
Abu4 Muhammad Abd Alla4h b. Ahmad b.
Zabr al-Qa4d3 (255/870329/940)
Abu4 Ubayd [historian] (d. 224 H), 8, 33,
40, 63, 657, 69, 8991, 124, 199,
200
Abu4 Ubayda b. al-Jarra4h [general], 49,
715, 131, 158, 199
Abu4 Yu4suf b. Dha4s3 (Adha4s3), 106
Abu4 Yu4suf Yaqu4b (d. 789) (Kita4b
al-khara4j), 8, 40, 62, 6872, 74, 75,
7780, 824, 86, 87, 8991, 102, 103,
154, 158, 159, 166, 168, 2013, 206,

210
accommodation of Muslims, 9, 46, 47, 49,
78, 82, 83, 98, 125, 126, 142, 145,
146, 162, 167, 171, 172
Achaemenid, 144, 147, 152, 158
adab, 135
adha4n, 72, 75, 77, 78
A"dharbayja4n, 46
Adhra4t, 199
Adu4d al-Dawla, 146
Aetolia, 51, 194, 1809
agreement, seesurrender agreement; treaty
Ahasuerus, 152, 153
ahd (pl. uhu4d), 9, 18, 31, 33, 34, 36, 38,
40, 47, 56, 64, 65, 72, 80, 85, 174,
195 (seealsosurrender agreement)
ahd Ardash r b. Ba4bak, 138
ahl al-kita4b, 1214, 127
(seealsonon-Muslims)
ahl al-dhimma, 66, 71, 74, 81, 99, 123,
143 (seealsonon-Muslims)
Ahmad b. Tu4lu4n, seeIbn Tu4lu4n
al-Ahnaf, 37
al-Akhtal, 163
Alamanni, 16
Alexander the Great, 128, 220
Alexandria, treaty of, 50, 51, 53, 56
Al3 b. Ab3 Ta4lib, 66, 68
Al3 al-Rida4, 229
Allah (but Alla4h in quoted Arabic), 33, 37,
41, 49, 72, 77, 230 (seealsoGod)
alliance, seetreaty
255
256
Index
alwiya, seebanners
ama4n, 317, 40, 41, 44, 45, 57, 61, 64, 70,
73, 84, 86, 125, 171, 173, 189, 205
(seealsofides; melta4 d-qya4ma4; pistis)
ama4ma pl. ama4im, seeturban
amicitia, seetreaty
Amida, 13, 22, 30, 188
Am r al-Mumin n, seecaliph
a4mma, 129
Ammianus Marcellinus, 12, 13, 16, 27
Amorkesos, 19
Amr b. al-A"s, 8, 38
amsa4r al-muslim n, 63, 659, 75, 82, 83,
84, 101, 200
Amurru, 193
Anastasius, 19
Anastasius of Sinai (d. c. 700), 92
Ana4ta, 73, 188
Anta4bulus, 39
Antioch, 22, 23, 26, 27, 29, 30, 31, 186,
188
Antiochus III, 43, 50, 51, 180, 191, 194,
215
anwatan, 36, 636, 101, 164
Apamea, 31
apparel, seedress
aqd, 11, 18, 36 (seealsosurrender
agreement)
Arabian peninsula, 4, 18, 19, 20, 33, 38,
54, 55, 57, 64, 115, 133, 151
Arabic, 11, 34, 36, 49, 57, 78, 83,
100, 102, 105, 126, 136, 138, 139,
144, 1503, 155, 157, 167, 172, 211,
212
Arabs, 11, 16, 18, 19, 20, 32, 34, 47, 49,
50, 57, 63, 64, 76, 92, 94, 97, 142,
143, 151, 204, 207, 213, 214, 221,
222, 226
Bedouin ideal, 1303, 144, 150, 153
Ardab3l, 42, 46
Ardash3r I (22641 CE), 139, 140, 145
Arethas, 19
Ar3b b. Sad al-Qurtub3, 108
Armenia, 23, 45, 47, 48, 49, 120, 145, 195,
221
arms, 26, 28, 72, 77, 90, 95, 126, 150,
152, 158, 159, 167, 172, 230
(seealsodress; swords)
Arsacid, 145
Arukh, 156
asa4wira (Ar., Pers. asavara4n), seecavalry
asb, seeturban
Askalon, 196
aspa4r, 150 (seealsocavalry)
Aspebetos, 185
Attila, 17, 50
Aula4d, 47
Avesta; Avestan, 118, 137, 138, 224
Ayla, 33, 38, 191
al-Azd3 (Futu4h), 207
Aziras, 193
Baalbakk, 42
Babylon/Memphis (Manf), 51, 192
Babylonia, 154, 155, 217, 222
Ba4dgh3s, 45
Bahra4m III, 184
Bahra4m IV, 119
Bahrayn, 33, 38
al-Bala4dhur3, 8, 38, 40, 53, 74, 131, 151,
192, 198, 222
banner, 59, 72, 75, 135, 158, 201, 229, 232
(seealso processions)
Banu4 Taghlib, 92, 207
baqt, 9, 34, 35, 39, 41, 55, 179, 190
Bar Hebraeus, Gregory, 91, 96, 105
Bar-Kokhba revolt, 53
Barqa, 39
basilika (Byzantine law code), 121
Basra, 58, 61, 65
Batnae, 23
al-Bayhaq3, Ahmad b. al-Husayn
(9941066) (al-Sunan al-kubra), 60
Bedouin, 131
bell, 111 (seealso ghiya4r)
belt, 95, 120, 129, 133, 135, 147, 152,
154, 155, 156, 209, 210, 216, 230,
231, 232
cingulum, 156, 157
kamar, 120, 155, 156, 157, 231
mintaqa, 133, 148, 154, 155, 157, 210
Seealsogirdle
Benveniste, mile, 137
Beroea (Aleppo), 23
b a, seeprayer-house
Bihquba4dh, 42
biography, 92, 93
blood, 71, 81, 174, 176
Book of a Thousand Judgments, 119
Borrut, Antoine, 93
Bukhtishu4 b. Jurj3s, 163
Bundahishn 137, 140
Bushanj, 45
buttons, seeghiya4r
Index
Byzantine, 1, 4, 6, 10, 11, 13, 14, 18, 19,
20, 26, 31, 32, 39, 41, 42, 47, 50, 52,
58, 97, 132, 133, 135, 151, 164, 188,
189, 219, 221
administration, 31, 51, 52, 116, 118,
120, 126, 129, 130, 136, 162, 218
court dress, paraphernalia and ceremony,
128, 129, 138, 150, 153, 156, 157,
166, 221
social mobility, 1215, 168
caliph, caliphate, 71, 79, 95
Candidus, 30
Cappadocia, 23
Caracalla (Constitutio Antoniana), 115
carrion, 81, 174, 176
Catholicos, 120, 218
cavalry, 49, 115
Muslim, 37
Persian, 49, 137, 158
Seealsohorses; saddles
Chalcedonian, 53
Christensen, Arthur, 140
Christianity, Christians, 19, 34, 94, 95,
108, 110, 121, 127, 131, 136, 154,
155, 157, 163, 221
in Byzantine Empire, 31, 116, 117, 118,
120, 125, 166
Christian Arabs, 92, 163, 184, 185, 207
after Muslim conquest, 40, 514, 59, 62,
64, 69, 72, 73, 75, 76, 78, 79, 80, 83,
86, 8992, 171, 205
restrictions on, 99102, 104, 105, 106,
122, 124, 146, 158, 159, 171, 173,
202, 207, 210, 233
in Sasanian Empire, 45, 119, 120, 217
Chronicle of Sert, 23
Chronicon ad 1234, 23, 31, 35, 91
church, 22, 23, 31, 35, 40, 42, 45, 51, 52,
61, 639, 71, 738, 80, 82, 83, 85,
1015, 120, 126, 127, 129, 158, 160,
161, 166, 167, 171, 172, 175, 196,
197, 200, 201, 203, 204, 205, 207,
210, 219, 233
congregating in (ijtima4), 77, 122
Seealsoprayer-house
Cicero, 25
cingulum, seebelt
city, 8, 10, 13, 14, 18, 2133, 36, 39, 42,
43, 44, 46, 47, 51, 52, 53, 56, 57, 58,
637, 6972, 74, 75, 79, 82, 83, 84,
86, 90, 93, 96, 109, 113, 115, 122,
257
132, 136, 164, 165, 171, 173, 175,

176, 186, 188, 196, 216
class, 25, 26, 95, 96, 98, 100, 103, 117,
118, 128, 129, 130, 132, 136, 13750,
152, 154, 157, 163, 167, 168, 194,
214, 217, 224, 226
coexistence, 2, 70, 72, 75, 78, 83, 86
colour, seedress
Commander of the Faithful (Am r alMumin n), seecaliph
Conrad, Lawrence, 5
Constantine, 48, 85, 156, 183, 185
Constantinople, 13, 91, 97, 221
Constantius, 181
Continuatio, 106, 107, 110
conubium, seemarriage
conversion, 85, 119, 122, 166, 172
to Christianity, 19, 45, 117, 118, 119,
184, 185
to Islam, 54, 92, 96, 105, 122, 136, 142,
156, 207, 214, 219, 221, 223
to Judaism, 117, 123
to Zoroastrianism, 119
Cook, Michael, 5, 93
Cordova, 108
Corpus Iuris Civilis, 68, 121
Crone, Patricia, 5, 96, 124
cross, 42, 52, 59, 63, 71, 72, 73, 758, 82,
85, 91, 100, 101, 102, 105, 111, 126,
157, 158, 166, 172, 175, 180, 196,
199, 202, 203, 204, 210
Ctesiphon, 120
cushions, 145, 146, 227, 230 (seealso
seating)
Cyrus, bishop of Alexandria, 51, 53
Dab3l (Dw3n), 37, 53
Da4desta4n de4n g (Dd 38:11), 154, 155
Daja4im/Dujam, 19
Damascus (Dimashq), 23, 31, 40, 42, 65,
74, 85, 96, 196, 205
Covenant of, 190
governor of, 76, 100
Damra b. Hab3b, 149
Dara (Daras), 23, 50
Darband, 49
deditio, seesurrender treaty
defamation 31, 81, 124
of Christianity, 124, 166
of Islam, 71, 80, 124, 162, 166,
173, 219
devil, 103, 212
258
Index
dhimma, 33, 36, 41, 42, 172, 174, 175,

176 (seealso protection)
dhimm , 39, 58, 5971, 73, 75, 7786,
89, 96, 99109, 111, 124, 128, 129,
1436, 149, 153, 154, 161, 162, 173,
203, 211, 212 (seealso non-Muslims)
al-dhull wa-al-sagha4r, 95, 96, 139
dib r, 136, 138
dihqa4n, 96, 136, 138, 143, 147, 151, 156,
223
Diocletian, 128, 156
Dionysius of Tell-Mahre4 (d. 845), 74, 76,
91, 100, 190
diya, 81
dress, 6, 7, 76, 82, 88, 95, 98, 102, 105,
106, 126, 127, 129, 130, 131, 154,
156, 163, 167, 168, 172, 175, 199,
223, 229, 233
arms, 28, 72, 77, 90, 91, 95, 126, 150,
152, 158, 159, 167, 172, 230, 232
colours: black, 106, 107, 111, 135, 149,
157, 229, 230; blue, 106, 107, 111,
148, 149; gold, 13, 14, 17, 48, 52, 56,
128, 132, 146, 148, 150, 151, 152,
153, 155, 163, 188, 220, 230; green,
229; red, 38, 135, 148, 223; white,
105, 111, 148, 159, 229; yellow, 103,
104, 107, 148, 149, 229
Iranian dress, 96, 97, 98, 128, 132, 133,
135, 148, 158, 167, 168
khatam, seeseal
liba4s al-madha4lla, 95, 140, 147
luxurious dress, 96, 128, 131, 132, 133,
135, 139, 140, 147, 148, 153, 163
swords, 16, 126, 133, 135, 150, 152,
153, 156, 157, 167, 172, 199, 230,
232
turban (asb; ima4ma, pl. ama4im), 89,
92, 97, 101, 104, 126, 135, 146, 148,
157, 172, 207, 229, 230
Dujum [Zokomos], 19
Dupi-Tessub, 193
East, 1, 10, 11, 21, 23, 30, 32, 36, 39, 41,
50, 56, 57, 62, 114, 115, 118, 121,
124, 125, 130, 148, 153, 162, 164,
166, 180
Easter, 126, 158, 159, 167, 172
Edessa (Ruha4), 22, 29, 30, 39, 76, 84
Egypt, 9, 13, 33, 35, 36, 38, 39, 41,
42, 55, 65, 76, 93, 99, 101, 102, 110,
111, 153, 188, 193, 197,

213, 215
Eilath, 19
Elias of Nisibis, 110, 213
Emesa, 31
empire
Byzantine, 1, 6, 11, 20, 23, 32, 44, 50,
11622, 124, 130, 148, 156, 157, 164,
186, 189
Islamic, 111, 157, 214 (seealso
caliphate)
Roman, 12, 16, 18, 20, 21, 25, 44, 50,
115, 116
Sasanian, 11, 12, 47, 48, 116, 118, 119,
120, 121, 136, 148, 157, 221
epigamia (Gk.), seemarriage
epistle, 93, 94
Eretz Israel, 155
Esarhaddon, 193
Eusthantios, 15
Eutropius, 27
Eutychius, 31
Exilarch, 120, 156
Fadak, 33
Farzman, 14
Fa4timid, 104
Fattal, Antoine, 3, 6, 8, 59, 60, 61, 62, 79,
89, 96, 99, 108, 110, 124, 197, 203,
207, 211, 219
fides, 10, 11, 24, 28, 31, 32, 34, 36, 44, 57
(seealsoprotection, promise of)
Fihl, 52
Filast3n, 52, 71, 197
Firdaus3 (Sha4h-na4ma), 47, 127, 194, 224
foederati, 16, 17, 19, 25, 28, 182, 185
foedus (pl. foedera), 12, 13, 15, 16,
25, 28, 44, 47, 57, 182, 183, 185
(seealsosurrender agreement; treaty)
Forand, P., 56
forelocks, cutting of, 89, 90, 92, 95, 97,
101, 144, 153, 206 (seealsoghiya4r)
Fritigern, 17
funerals, 62, 83, 106, 109, 126, 127, 144,
158, 161, 162, 167
Gabboula, 23
Galatia, 23
Galerius, 12
Gaonic, 156
garb, garments, seedress
Index
259
Gaudarz, 47
Geniza documents, 123, 213
ger (Hebrew), seeresident-alien
Ghassa4nids, 19
al-Ghaza4l3, 161
Ghaznavid, 152
ghiya4r, 61, 69, 78, 83, 88, 89, 92, 958,
10011, 114, 127, 135, 143, 144, 147,
149, 163, 166, 167, 168, 207, 213,
226 (seealsobell; dress; forelocks,
clipping of; headgear; khila4f; seats;
zunna4r)
girdle, 89, 90, 92, 98, 101, 156, 157, 206
Jewish, 154, 157
Zoroastrian, 128, 154, 157
Seealsobelt; dress; zunna4r
G3v, 47
God, 33, 37, 41, 42, 74, 936, 106, 154,
155, 1714, 176, 222 (seealsoAllah)
Goitein, Shlomo D., 52, 99, 197
Goths, 1620, 47, 51, 185, 194, 196
guiding the Muslims, 46, 49, 61, 71, 74,
85, 192
Gutas, Dimitri, 142
Heather, Peter, 25
Heraclea, 43, 44
Heraclius, 15, 31
heretics, 116, 121, 124, 216
Hierapolis (Membij), 188
Hierapytna, 193
Hija4z, 61, 175, 203 (seealsoArabian
peninsula)
Hila4l al-Sa4b3 (9691056; Rusu4m da4r
al-khila4fa), 135, 150, 151, 157, 159,
230
Hims, 74, 196, 205
al-H3ra, 42, 67
Hisha4m b. Abd al-Malik, 207
homologia, seesurrender treaty
horses, 104, 105, 106, 110, 111, 131, 133,
139, 144, 150, 151, 152, 153, 167,
214
Hoyland, Robert, 5
Hudaybiyya agreement, 36
Huna bar Nathan, Rabbi, 231
Huns, 16, 17, 20, 47, 183, 194
al-Husayn, 38
Hu4shang, 138
Hab3b b. Maslama, 37, 39, 53

Hadad, 189
hadd (pl. hudu4d), 81, 175
(seealsopunishment)
Had3th (the whole body of tradition); a
had th (a specific had th), 4, 63, 64,
90, 146
hair, hairstyle, 97, 101, 126, 132, 146, 148,
153, 161, 167, 172, 175, 226, 228
(seealsoforelocks, clipping of; ghiya4r)
hajj, 202
Hajjia4ba4d, 48, 141
al-Hakam b. Umar al-Ruayn3, 90, 206
al-Ha4kim (r. 9961020), 99, 104, 108, 110,
111, 163
hal f, 115, 214
Haman, 152
Hamza al-Isfaha4n3, 19
hamya4na4, seegirdle
Hara4t, 45
Harra4n, 22
Ha4ru4n al-Rash3d, 68, 70, 76, 102, 103,
147, 163
Hasan al-Basr3 (d. 110 H), 67
headgear, 82, 89, 107, 139, 146, 147, 175,
230
Ibn Abba4s, 63, 64, 69

Ibn Abd al-Hakam, 8, 38, 55, 89, 90, 92,
93, 94
Ibn Abd Rabbih (860940; al-Iqd alfar d), 135
Ibn Ab3 Shayba (159235 H), 67
Ibn Asa4kir (Tar kh mad nat dimashq), 40,
59, 61, 72, 73, 76, 84, 89, 90, 91, 149,
201, 205, 206
Ibn Atham al-Ku4f3, 8
Ibn Bishr (Isha4q b. Bishr Abu4 Hudhayfa
al-Ha4shim3), 72
Ibn Diyya4s al-Nasra4n3 al-Qubt3, 39
Ibn Hanbal, 60
Ibn Hibba4n (884965), 60
Ibn al-Jawz3, 93, 104
Ibn Khaldu4n, 131, 132, 148, 153, 158,
160, 222, 232
Ibn al-Mualla4 (d. 286 H), 40
Ibn al-Muqaffa (d. c. 760 CE), 138, 139,
226
Ibn al-Murajja4, Abu4 al-Maa4l3 al-Musharraf
(c. 103040; Fada4il bayt al-maqdis
wa-al-khal l wa-fada4il al-sha4m), 61,
171
Ibn al-Nad3m, 138
260
Index
Ibn al-Naqqa4sh, 108, 207

Ibn Qayyim al-Jawziyya (Ahka4m ahl aldhimma), 59, 90, 108, 112, 127, 171
Ibn Taghr3 Bird3, 108
Ibn Tu4lu4n, Ahmad, 109, 110, 160, 161
Ibn al-Washsha4 = Muhammad b. Isha4q
al-Washsha4 (Kita4b zarf al-zurfa4), 149
Ibn al-Wa4sit3 (Gha4z3 b. al-Wa4sit3), 84, 85
Ibn Zabr (870940 CE), 60, 89, 90, 91,
206
al-Ibsh3h3, 207
3d al-Adha4; 3d al-Fitr, 159
identity, 116, 120, 121, 146, 154, 157, 228
ideology, 6, 7, 88, 92, 93, 94, 97, 112, 137,
141, 163, 166, 167, 168
idol (wathan), 106, 107, 109, 212
(seealsoshayta4n, shayta4n)
al-Ijl3, Abu4 Dulaf (d. 840), 132
ika4f, seesaddles
al-Ikhsh3d, Muhammad (r. 934), 99, 108,
110, 211
ima4m, 55, 59
ima4ma (pl. ama4im), seedress
Imru4 al-Qays, 19, 184
inheritance, 117, 118, 124, 137, 142, 162,
166
Iotaba, 19
Iran, 1, 4, 7, 45, 47, 48, 50, 57, 129, 142,
155, 179, 221
Iranian classes/estates, 118, 137, 138,
140, 141, 146, 150, 151, 152, 153,
156, 161, 163, 217
Iranian customs, 51, 56, 127, 128, 133,
135, 143, 144, 145, 147, 148, 149,
150, 157, 160, 221, 228
SeealsoPersians
Isfaha4n, 45, 47
ius gentium, 10, 115
Iya4d b. Ghanm, 84
Iya4s b. Qab3sa, 73
iza4r, 104, 148 (seealsoghiya4r)
Ja4hiliyya, ja4hilI customs, 161
al-Ja4hiz (-pseudo) (Kita4b al-ta4j), 135, 145,
165
al-Jahshiya4r3 (d. 331 H; Kita4b al-wuzara4),
147, 211
Jamsheed [Yama], 137
al-Jarba4, 38
Jar3r b. al-Khatafa4, 163
al-Jarra4h b. Abdalla4h al-Hakam3, 39
Jawhar, 104
al-Jaz3ra, 65, 84
Jerusalem (Il3ya4), 23, 31, 35, 42, 50, 52, 53,
65, 131, 191, 196, 197, 215, 221
Church of the Ascension, 76
Church of the Eleona, 76
Church of the Holy Sepulchre, 218
Dome of the Rock, 76
Mount of Olives, 76
Temple Mount, 76
Jews, 1, 51, 52, 53, 76, 94, 127, 153, 157,
197
autonomy, 116, 215
Byzantine period, 6, 31, 116, 117, 118,
121, 166, 216
circumcision, 116
conversion to Judaism, 117, 123
holding public office, 108, 117, 118,
120, 124, 216
inheritance, 117, 124
marriage, 117, 124
ownership of slaves, 117, 123, 124
Patriarch, 116, 117
privileges, 116, 163
religio licita, 122
restrictions, 52, 99, 101, 105, 111, 117,
124, 146, 215, 217, 233
Roman period, 116, 194
Sabbath, 116, 118, 119, 156
Sasanian rule, 119, 120, 154, 155
synagogues, 109, 117, 118, 122, 218
testimony, 117
jiwa4r, 32, 33, 34, 54, 189
(seealsoprotection)
jizya, 40, 42, 43, 44, 49, 64, 71, 74,
84, 101, 102, 173, 175, 203, 231
(seealsotaxes)
John of Nikiu, 51
Jokisch, Benjamin, 68, 121, 125, 162
Josephus, 191, 215
Joshua the Stylite, 13, 14, 31
Jovian (r. 3634), 12, 13, 22
judge, 41, 81, 117, 137, 150, 175
Julian (r. 3613), 22, 27, 28
Jund al-Urdunn, 107, 108
jurists, 41, 56, 61, 63, 67, 68, 69, 74, 75,
80, 114, 123, 166, 201, 218, 219, 233
(seealsoshar a)
Jurja4n, 35, 37, 45, 49, 56
Justin II (r. 56578), 48
Justinian (r. 52765) (Corpus Iuris Civilis),
12, 13, 19, 23, 118, 121, 129, 156,
185, 216, 220, 221
Index
Kaba-yi-Zardusht, 48
Ka4ba-yi-Zardusht inscription
(Naqsh-i-Rustam), 12, 48,
141, 156
ka4fir, seenon-Muslims
al-Kalb3, Hisha4m (737819 or 821 CE;
Jamhara), 207
Kalfon-Stillman, Yedida, 132, 133
kamar, seegirdle
kan sa, seeprayer-house
al-Ka4sa4n3, 207
ka4tib (pl. kutta4b), 96, 147, 150
(seealsodib r)
ka4tib al-jund, 136
ka4tib al-khara4j, 136
Kava4dh I (r. 488531), 13, 195
Kava4dh II (r. 628), 15
Kay-Ka4vu4s, 194, 195
al-Kawa4thil, 73, 202
Kay-Khusro, 47
Kha4lid [transmitter], 207
Kha4lid b. al-Ja4biya, 131
Kha4lid b. al-Wal3d, 40, 73
kha4ssa, 129
khatam, seeseal
Khaybar, 33
khila4f, 61, 101 (seealsoghiya4r)
khila, 135 (seealsodress)
Khirbat al-Mafjar, 133
Khura4sa4n, 49, 65
Khusro Anu4sh3rwa4n (r. 53179), 12, 23,
29, 30, 37, 53, 119, 136, 138, 188,
189
khutba, 94
Kinda, 19
kisra4 (pl. aka4sira), 131, 132
kita4b, 36, 37, 38, 40, 53
kita4b ama4n, seesurrender agreement
Kita4b al-ta4j f s rat Anu4sh rwa4n, 138
Ku4fa, 58, 66
Kunduhar, 222
kunya, 78, 83, 126, 142, 144, 146, 149,
150, 167, 172
kurs , 135
kust g, 154, 155, 1457 (seealsogirdle)
la4 tashabbahu4, 127, 143, 162
law, 43, 45, 68, 84, 99, 106, 115, 125, 128,
192, 193
Byzantine law, 6, 11625, 162, 163, 166,
167, 216, 218, 219
international law, 10, 33
261
Muslim law, 3, 60, 61, 64, 68, 69, 79,

80, 81, 91, 96, 1215, 162, 166, 167,
174, 175, 179, 203
profession, 117, 118
Sasanian law, 11825, 139
Seealsoius gentium; shar a
Letter of Tansar, 138, 139, 140, 144, 145
lights, showing of, 72, 77, 78, 83, 126,
127, 157, 158, 159, 172, 186
longue dure, 2
Lucius Verus, 18
Ludd, 42, 197
Maa4n, 202
Maghrib, 111, 163
Magnesia, battle of (190 BCE), 43
Ma4h Bahra4dha4n, 45, 47, 193
Ma4h D3na4r, 45, 47, 193
al-Mahd3, 207
Mahmu4d (Seljuk sultan), 110
Makhu4l al-Sha4m3 (d. 113/731), 71, 73, 80,
81, 83, 85
Malalas, John (mid-sixth century), 21, 188
Malchus, 19
Ma4lik, 93
Ma4lik b. Anas (Muwatta), 93
Ma4lik3, 93
Mamersides, 28
al-Mamu4n (d. 206/821), 72, 132, 229
Mamikonean house, 48, 195
Mamlu4k, 68, 112
al-Mansu4r (r. 7545), 102
Mansu4r b. Gha4lib, 93
Manu4shihr, 138, 147
Maqna4, 33, 38
al-Maqr3z3 (d. 1442), 35, 41, 55, 56, 104
Marcellinus Comes, 13
Marcus Aurelius, 18
marriage, 17, 115, 117, 119, 139, 188,
216, 226
between Muslim and non-Muslim, 80,
123, 174, 203
Martyropolis, 23
Marw, 36
Marwa4n II, 76
Marwaru4dh, 36, 45, 47, 49, 53
marzuba4n/marzba4n, 36, 37, 48, 49, 53,
120, 145, 147
Masru4k b. Abd al-Rahma4n b. Ghanam, 90
al-Masu4d3, 145, 147, 160
Mauritius, 181
Mavia, 18, 19, 185
262
Index
mawkib, seeprocession
mawla4 (pl. mawa4l ), 115, 226
status of, 142, 214
Ma4zandara4n, 47
Mazdakite movement, 141
Mazdean, 137
Medina, 65, 66, 93
Megas, 29
Melkite, 31
melta4, seesurrender agreement
melta4 d-qya4ma4, seesurrender treaty
Menander the Guardsman (Protector), 14,
15, 20, 45
metoikos (Gk.), seeresident alien
Michael the Syrian, 76, 91, 101, 210
minority, 1, 3, 62, 68, 114, 115, 116, 120
mintaqa (pl. mana4tiq), seegirdle
misr (pl. amsa4r), 64, 65, 66, 67, 69, 86
(seealsocity)
Misr, seeEgypt
monasteries, 42, 45, 76, 122, 171, 207
monks, 42, 101, 102, 122, 154, 156, 157,
171, 204
Monophysite, 1
Mordechai, 152
Morony, Michael, 5
mos maiorum (Latin), 116
Moses of Chorene, 48
mosque, 66, 75, 76, 85, 161, 200
(seealsoprayer-house)
MS Kprl 1608, 138
Mua4wiya b. Ab3 Sufya4n, 131
Mua4wiya b. Qays, 19
Muhammad, 33, 44, 79, 80, 94, 104, 164,
168, 173, 184
Muhammad b. Kab al-Quraz3 (d. 10819
H/72637), 61
Muhammad b. Tughluq, 227
Muhammad b. Umayr (d. 999), 229
Muhammad b. Zayd, 55
Muharram, 104, 105
al-Muizz (r. 95375), 104, 108, 110
al-Muqaddas3, 110, 111
Mu4qa4n, 45, 47
al-Muqtad3 (r. 1091), 110
al-Muqtadir (r. 90832), 99, 104, 108, 110,
151, 211
Muqurra, 9, 35, 55
Mursilis, 193
mushrikun
4 , 94, 122 (seealsonon-Muslims)
al-Mustansir (r. 1086), 110
al-Mutasim (r. 83342 CE), 132, 145, 157
al-Mutawakkil (r. 84761), 78, 87, 96, 99,

100, 102111, 123, 143, 146, 149,
154, 161, 168, 210, 212
Nabatean, 18
Najra4n, 33, 38, 53, 54, 55
al-Nak3b, 73, 202
Nama4rah, 18
Naqsh-i-Rustam, 12
na4qu4s (pl. nawa4q s), 62, 63, 64, 77, 78, 85,
91, 101, 103, 160, 166, 180, 200
Narseh, 12, 184
Nauro4z, 138
Netanel, 108
N3mru4z, 47
Nisibis, 13, 22, 50
noise, 161
Nomalius (?), 19
non-Muslims, 18, 34, 36, 58, 59, 68, 69,
99, 109, 113, 121, 124, 125, 130, 142,
1649, 177, 208, 215, 226
apostasy, 123
conversion, 96
defamation of Islam, 124, 162
differentiation of, 3, 6, 61, 62, 83, 86,
88, 89, 90, 91, 98, 100, 101, 114, 126,
127, 143, 144, 146, 148, 149, 152,
154, 155, 157, 206, 213
holding of public office, 59, 95, 96, 100,
101, 102, 103, 123, 137, 162, 163,
211
humiliation of, 3, 7, 948, 103, 104,
105, 106, 114, 142, 150, 161
law of inheritance, 124, 162
marriage to Muslims, 123, 124
ownership of slaves, 122, 123
protection of, 33, 123, 142 (seealsoahl
al-dhimma; dhimm )
subjugation of, 6, 114, 126, 128, 129,
142, 145, 147, 148, 153, 159, 161,
162, 163
testimony, 124, 162
tolerance of, 59
Nonnosus, 19
Noth, Albrecht, 3, 9, 36, 60, 61, 62, 70, 77,
95, 114, 126, 127, 180, 199, 203, 214
Nubia, 8, 9, 34, 35, 39, 41, 55
nusha4tiruhum, 84, 85
oath, 10, 11, 13, 16, 17, 22, 24, 28, 29, 30,
31, 34, 35, 41, 42, 44, 48, 188, 190
Ohrmazd, 148
Index
Pacatus, 19
pact, 9, 10, 11, 24, 26, 37, 47, 48, 195
pactio/pactum, 12, 24, 25, 26, 34, 36,
44, 188, 196
Sha4fi3s pact, 58, 79, 84, 97, 1736
Seealsosurrender agreement; treaty
Pact of Umar, seeShuru4t Umar
pagans, 94, 116, 121, 122, 124, 216
Pahlavi, 136, 138, 139, 144, 226
Pa4iku4l3 inscription, 48
Palaestinas, 19
Palm Sunday, 72, 73, 74, 78, 82, 103, 105,
126, 158, 167, 172
Paphlagonia, 23
Patriarch, Patriarchate, 31, 52, 53, 100,
116, 117, 118, 177, 190, 218
peregrinus (Latin), seeresident-alien
Persian, Persians, 5, 1215, 19, 203,
2531, 45, 4851, 97, 120, 138, 142,
143, 144, 146, 153, 155, 165, 167,
182, 184, 186, 188, 198, 220, 221,
230, 232
royal/court dress, 128, 133, 135, 147,
148, 152, 156, 167, 209
SeealsoIran
Peter the Patrician (c. 50064), 12
Petrus, metropolitan of Damascus, 204
Petrus of Maioumas, 204
Pharos, 43, 191, 215
Philip the Arab, 12
Philip of Macedon, 128
Phillipson, Coleman, 24
Photius (Bibliotheca), 19
physicians, 136, 138, 145, 163
pigs, 59, 72, 83, 219
prohibition of, 63, 64, 69, 72, 75, 81,
85, 124, 126, 166, 172, 174, 176, 203,
204, 205
slaughter by Abd al-Malik, 76, 100
Pirisabora, 27
P3ru4z (r. 45984), 48
pistis, 24, 31, 32, 36, 43, 44, 57, 190
(seealsoama4n; fides)
polis (pl. poleis), 43, 191, 193, 215
Polybius, 31
polytheism, 62, 82, 124, 126
(seealsomushriku4n)
pommel, seeghiya4r
pork, 66, 77, 78, 81, 201
Pourshariati, Parvaneh, 48, 49, 141
prayer-house, 8, 61, 64, 67, 71, 73, 101,
122, 124, 164, 199, 201
263
building of, 63, 66, 69, 111

destruction of, 66, 74, 75, 77, 107
Seealsochurch; mosque; synagogue
processions, 115, 157, 162, 166
Christian, 59, 62, 64, 72, 73, 83, 103,
105, 106, 126, 127, 157, 158, 159,
167, 172, 180
Muslim, 157, 158, 159, 230, 232
Persian retinue, 146, 158
Procopius (Wars), 13, 29, 185
prohibition, 20, 54, 69, 76, 77, 78, 80, 83,
84, 90, 91, 97, 102, 106, 108, 111,
117, 118, 123, 126, 130, 137, 146,
149, 162, 166, 203, 206, 207, 210,
227
building prayer-houses, 69, 77, 101, 180
carrying arms, 126, 150, 152, 167
conducting processions, 62, 82, 105,
106, 126, 127, 157, 158, 167, 180,
204
pigs and wine, 69, 81, 109, 126, 201,
203, 204
raising voices, 62, 91, 101, 102, 107,
126, 127, 157, 158, 160, 161, 167
showing lights, 126, 127, 157, 158, 159,
167
using clappers (nawa4q s), 62, 78,
91, 126, 127, 157, 158, 159, 167, 180
proskynesis, 128
protection, 8, 10, 27, 32, 33, 40, 41, 42,
45, 49, 54, 65, 71, 74, 75, 79, 81, 86,
115, 117, 118, 122, 126, 162, 174,
176, 189
promise of, 24, 28, 34, 84
(seealsoama4n; fides; melta4 d-qya4ma4;
pistis)
See also dhimma; jiwa4r
Pseudo-Zachariah, 22
public sphere, 64, 135, 162, 167
punishment, 19, 30, 66, 67, 81, 121, 219
(seealsodiya)
qaba4 (Per. kapa4h), 89, 90, 91, 95,
101, 133, 147, 148, 206, 209
(seealsodress)
Qadarite, 93
qadhf, seedefamation
qa4d , 40, 159 (seealsojudge)
al-Qa4d3, Wada4d, 9, 10, 39, 40, 56
qalansuwa, seeheadgear
Qa4l3qala4, 56
al-Qalqashand3, 108, 139
264
Index
Qarq3siyya, 74
Qasr al-Hayr al-Gharb3, 133
Qasr Ibr3m, 9
Qays, 19, 20
Qu4mis, 45, 47
Qumran, 190
Qura4n, Qura4nic, 32, 33, 34, 78, 83, 94,
95, 123, 126, 164
Qusayr Amra, 133
Rab3 al-Awwal, 79
al-Ra4d3 (r. 93440 CE), 105
Rajab, 54
Ramla, 106, 212
al-Raqqa, 180
rasa4il, seeepistle
Ra4shidu4n, 74, 75, 79, 89, 205
Rawwa4fa, 18
al-Rayy, 45, 47
regulations, seerestrictions
resident alien, 115, 117, 118, 121, 128
restitutio, 25, 44
restrictions, 1, 2, 4, 6, 60, 61, 64, 6670,
75, 77, 78, 82, 83, 86, 88, 95, 98, 99,
103, 104, 109, 110, 111, 125, 128,
143, 163, 165, 166, 169, 171, 207,
212, 213
al-Ha4kim, 99, 104, 108, 211
al-Mansu4r, 102
al-Mutawakkil, 99, 100, 10311, 143,
146, 148, 161, 168, 210
pre-Islamic, 114, 115, 117, 120, 121,
125, 166, 217
Umar II, 89, 91, 96, 102, 112, 199
al-Wa4thiq, 105
Rhodos, 193
riha4la, seesaddles
Robinson, Chase, 5
Romans, 10, 1219, 21, 22, 27, 28, 29, 44,
50, 51, 57, 192
Rome, republican, 15, 25, 43, 47, 50, 180,
192, 194, 214
Ruha4, seeEdessa
Rustam, 47, 194, 195
Sabbath, 116, 118, 119
saddles (sirj, riha4la, ika4f), 82, 89,
90, 91, 92, 97, 98, 101, 103, 126, 129,
150, 151, 152, 153, 167, 172, 175
Safar, 54, 104
safe-conduct, seeama4n
Sa3d b. Abd al-Rahman b. Hasan

b. Tha4bit, 207
Salah al-D3n, 111
Sa4lih b. Al3, 102
Sa4lim b. Abdalla4h, 208
Samosata, 76
al-Sarakhs3 (d. 483 H; Sharh), 66, 67, 68
Sar3 b. Musarraf, 61
sar r, 135
Sarmatians, 16
Sasanians, 12, 21, 22, 25, 26, 27, 29, 44,
49, 125, 132, 155, 158, 183
classes/estates/society, 118, 121, 123,
128, 130, 13641, 163, 167, 168, 220,
227
social (aristocratic) ethos, 95, 127, 129,
131, 1415, 222
status symbols, 130, 132, 133, 135, 148,
150, 156, 157, 159, 221
treatment of minorities, 118, 119, 120,
166, 217
Sayf b. Umar, 52, 197
Schacht, Joseph, 32
scribe, 40, 47, 96, 106, 118, 120, 130, 136,
137, 138, 163, 167 (seealsoka4tib)
Scythia, 16, 17
seal (khatam), 8, 15, 16, 20, 37, 40,
41, 47, 48, 78, 83, 126, 132, 135,
148, 152, 153, 165, 172, 194, 230
(seealsodress; ghiya4r)
seats, seating, 78, 82, 83, 98, 107, 126,
127, 142, 145, 146, 160, 162, 167,
172, 175, 227 (seealsocushions;
ghiya4r)
Sebeos, 30, 48
Sefire III (c. 750 BCE), 193
Seleucids, 43, 193
Seljuk, 110
Sergiopolis, 30
Severus b. al-Muqaffa, 105
al-Sha4fi3 (150204/767820 Kita4b alumm), 40, 69, 78, 79, 81, 83, 87, 88,
124, 146, 166
Shahal, Yariv, 25
Shahunshah (Great King), 128
al-Sha4m, 61, 65, 71, 84, 85, 111, 180, 189,
196, 202
Sha4pu4r I (r. 240/24370/73 CE), 12, 22,
141
Sha4pu4r III, 119
shar a, 9, 49, 84, 176 (seealsolaw)
Index
Shawwa4l, 104
al-Shayba4n3, Muhammad b. al-Hasan
(13287 H; Kita4b al-siyar), 66, 67,
68, 69
shayta4n, pl. shaya4t n, seedevil
Sh3ite (adj.), 89
Sh3ra4z, 110, 111
Shoshinduxt (daughter of Jewish Exilarch),
120
Shurahb3l b. Hasana, 196
Shuru4t Umar, 3, 111, 142
Abu4 Yusu4fs version, 70, 75, 77, 78,
82, 83, 166, 168 (seealsoAbu4 Yu4suf
Yaqu4b)
alternative versions, 6, 62, 69, 70, 98,
166
canonization of, 6, 60, 166, 168
dating and formation, 4, 603, 135, 162,
163, 168
enforcement of, 6, 99111, 214
nusha4t iruhum version, 84, 85
origins, 2, 113, 128, 162
provenance, 113, 122, 125, 129, 130,
138
al-Sha4fi3s version, 79, 80, 82, 83, 84,
109, 166, 168, 1735 (seealsoalSha4fi3)
stipulations of, 78, 82, 83, 84, 86, 100,
109, 122, 124, 129, 144, 167
versions of, 3, 59, 60, 61, 85, 98, 122,
171, 172, 203
sirj, seesaddles
silk, 47, 101, 128, 131, 132, 133, 139, 148,
206, 221 (seealsodress)
Singara, 13, 22
s ra, seebiography
sirwa4l, 133 (seealsodress; ghiya4r)
Sista4n, 195
sitr, 135
slaves, slavery, 23, 24, 55, 56, 103, 104,
115, 117, 119, 122, 123, 124, 140,
142, 147, 148, 150, 162, 166, 172,
175, 176, 198, 214, 216, 219
societas, seetreaty
society, 7, 79, 166, 194
Arab, 4, 130
Byzantine, 116, 118, 124, 129, 130, 157,
168
Iranian, 4, 7, 118, 128, 129, 130, 136,
137, 1405, 148, 149, 163, 167, 168,
221
265
Islamicate, 1, 2, 4, 5, 69, 79, 81, 86,

88, 96, 97, 98, 113, 114, 121, 124,
125, 127, 129, 130, 135, 136, 137,
141, 142, 143, 147, 149, 150,
162, 164, 167, 168, 169, 201, 223
Seealsoaya4n; amma; kha4ssa
Sophronius, 131, 177
Sozomenos, 18, 19
spa4badh, 145, 147
Spain, 204, 217
spies, 80, 93, 126, 131, 174
spondai, seetreaty
status, 15, 7, 11, 34, 45, 53, 60, 62, 63,
64, 65, 73, 81, 11319, 121, 122, 123,
125, 129, 130, 136, 137, 138, 1405,
147, 153, 154, 1637, 177, 195, 214,
215, 217
status symbol, 96, 97, 128, 130, 132, 133,
135, 136, 142, 147, 148, 150, 151,
152, 157, 167, 168, 220, 223
stirrups, seeghiya4r
al-Subk3, Taq3 al-D3n, 68
subordination, 6, 86, 121, 124, 159, 160
Sufya4n al-Thawr3 (97161/71678),
61, 90
Sulayma4n b. Abd al-Malik (r. 71517), 133
sulh, sulhan, seesurrender agreement
sunna, 65
Supiluliumas, 193
Sura, 23, 30, 198
su4ra, 94, 95
Su4ra 9, 33
surrender agreement, 2, 4, 5, 6, 9, 10, 21,
25, 26, 30, 32, 35, 38, 425, 579, 99,
114, 116, 122, 164, 166, 188
conditions of, 28, 29, 36
deditio; deditio in fidem; deditio per
pactionem, 24, 25, 27, 43, 44
detailed agreements, 8, 506, 165
homologia, 24, 29, 36, 44
melta4 (d-qya4ma4), 24, 31, 34, 35, 36
negotiations, 21, 24, 26, 27, 57, 165
sah fa, 18
sulh, sulhan, 36, 38, 61, 637, 6975,
78, 82, 84, 86, 93, 126, 158, 164
stipulations of, 11, 37, 44, 51, 79, 80
verba deditionis, 24
Seealsotreaty
swords, 16, 30, 51, 126, 135, 150, 152,
153, 157, 167, 172, 199, 230, 232
(seealsoarms; dress)
266
Index
synagogue, 63, 66, 67, 71, 73, 77, 104,

109, 117, 118, 122, 162, 166, 200,
201, 218, 219
tabaqa, seeclass
al-Tabar3, 8, 36, 40, 52, 67, 68, 104, 105,
107, 131, 193, 214
Tabarista4n, 49, 138, 145, 147
Tabarista4n wa-J3lj3la4n, 45
Tabu4k, 202
al-Ta4if, 38, 65
tafs r, 96
Talha b. Musarraf (d. c. 11213/7301),
61, 90
Talmud, Babylonian, 190
Berachot 60b, 155
Shabbat 49b, 156, 231, 232
Zevahim 19a, 231
Tanu4kh, Tanu4khids, 92, 207
taxes, 8, 37, 44, 49, 52, 70, 79, 94, 113,
115, 116, 119, 136, 143, 164, 198,
207, 219, 233 (seealsojizya)
taylasa4n, 89, 90, 110, 206 (seealsodress)
Tella, 186
Tertullian, 116
Thamu4dians, 18
Theodora, 220
Theodosius I, 17, 120
Law Code of, 121
Theodosius II, 17, 184
Theophanes, 76, 109, 204
Theophilus of Edessa, 91
Thomas Artsruni, 48
Tiberias, 31, 52, 196, 198, 205
Tifl3s, 193, 197
tira4z, 107, 133, 148, 163, 220
(seealsodress)
al-Tirmidh3 (Abwa4b al-liba4s), 149
Trajan (r. 98117 CE), 21, 22, 26
treaty, 8, 9, 34, 53, 54, 55, 56, 60, 61, 72,
80, 84, 85, 184, 185, 189, 192, 197,
199
amicitia, 11, 25
with Barbarians, 16, 18, 19, 50, 185
foedus aequum, 15
foedus iniquum, 13
with Goths, 16, 17, 47, 51, 185
international, 11, 193
protection of deity, 10, 42
with Romans, 1220, 22, 25, 29, 50, 51,
120, 153, 180, 182, 185, 274
with Sasanians, 12, 13, 14, 15, 22, 26,

27, 29, 44, 47, 50, 52, 120, 153, 182,
274
societas, 25
surrender treaty, 37, 425, 72, 73, 77, 86
terminology, 11, 17, 24
vassal treaty, 36, 42, 459, 80, 193
Seealsosurrender agreement
Tritton, Arthur S., 60, 61, 62, 79, 110, 154,
203
Tu4lu4nid dynasty, 109
turban, seedress
al-Turtu4sh3 (Sira4j al-mulu4k), 59, 171
Tu4s, 47
al-Tu4s3, Niza4m al-Mulk, 150, 230
Tyan, Emile, 33
Uba4da, 52
Ubayd Alla4h b. Ab3 Jafar, 38
Ugarit, 215
Umar b. al-Khatta4b (r. 63444 CE), 2, 35,
39, 60, 67, 74, 75, 76, 79, 84, 85, 89,
91, 95, 127, 131, 143, 149, 171, 172,
178, 199, 203, 209, 211, 221, 222,
232
Umar b. Abd al-Az3z (r. 71720 CE), 39,
62, 65, 67, 75, 76, 90, 95, 96, 97, 98,
127, 203, 206
epistles, 92, 93, 97
ghiya4r edict, 8992, 94, 1013, 107,
109, 148, 168
ideology, 6, 88, 924, 112, 168, 207
Umar b. al-Wal3d, 208
Umayyad, 36, 55, 75, 89, 93, 132, 133,
142, 158, 205, 206
Uthma4n b. Affa4n (r. 64456), 39, 43, 74,
76, 100
Vahan 48
Valarsh, 195
Valens, 16, 18
Valentian I (r. 36475 CE), 16, 17, 216
Valerian, 22
Vilmar, 105
Vulso, Gnaeus Manilius, 43
wal (pl. awliya4), 123
Wal3d II (r. 7434), 229
al-Wal3d b. Nu4h, 61
al-Wa4thiq (r. 8427), 103, 105
waz r, 153, 159, 161
Index
We4kart [Waygird], 138
Widengren, Geo, 47, 120, 155, 194
wine, 30, 47, 59, 64, 66, 72, 75,
81, 213
restriction of, 63, 69, 81, 83,
109, 124, 126, 166, 174, 176, 201,
203
Xenophon (431355 BCE; Cyropaedia),
150, 152, 158, 160, 227
Yahya4 b. A"dam, 33, 53
Yahya4 b. Hamza (d. 176 H), 40
al-Yaman, 545, 65
al-Yaqu4b3 (d. 897), 8, 196, 198
Ya4qu4t (al-Muqtada), 193, 207
Yarmu4k, battle of, 74
Yazdgird I (r. 399421), 11920,
231
Yazdgird II (r. 43857), 119
Yaz3d I (r. 6803), 133
Yaz3d b. Abdalla4h al-Hadram3, 39
267
Yaz3d b. Abd al-Malik (r. 71517), 207

Yaz3d b. Ab3 Sufya4n, 131
Za4bulista4n, 195
Zachariah of Mitylene, 14
Zayd b. Aslam, 38
Zinjirili, 189
Zoroaster, Zoroastrianism, 120, 137
Zoroastrians, 118, 121, 154, 157, 162,
220, 231
conversion from religion, 119
inheritance, 119
law, 1215
marriage, 119
slave, 119
Zosimus, 13
zunna4r (pl. zana4n r; Gk. zone, zonarion,
Syr. zu4na4ra4), 82, 90, 91, 95, 97, 98,
103, 107, 111, 126, 154, 155, 157,
167, 172, 175, 209, 230 (seealsobelt;
ghiya4r; girdle)
Zura4ra, 68

(Cambridge Studies in Islamic Civilization) Milka Levy-Rubin-Non-Muslims in The Early Islamic Empire - From Surrender To Coexistence-Cambridge University Press (2011)

Uploaded by

Copyright:

Available Formats

(Cambridge Studies in Islamic Civilization) Milka Levy-Rubin-Non-Muslims in The Early Islamic Empire - From Surrender To Coexistence-Cambridge University Press (2011)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Cambridge Studies in Islamic Civilization) Milka Levy-Rubin-Non-Muslims in The Early Islamic Empire - From Surrender To Coexistence-Cambridge University Press (2011)

Uploaded by

Copyright:

Available Formats

Non-Muslims in the Early Islamic Empire